Sciencemadness Discussion Board

Building an open source hot plate stirrer - feature requests!

mux - 11-12-2016 at 04:13



* Quick aside: I understand why these things are $500+ and people still buy them, I understand the concept of professional/educational lab budgets and needing to buy everything in one place, etc. etc.

** this post is formatted weirdly

aga - 11-12-2016 at 05:38

I got mine about 2 years ago for around €100 (new) and it still works fine, despite being very badly treated.

If/when the electronics fail (dissolve !) i'll probably rig up something to add a few features, such as remote control.

Perhaps a replacement/upgrade board for a cheap hotplate/stirrer would be easier than building the whole thing from scratch.

gdflp - 11-12-2016 at 10:01

Interesting ideas you have, though this has been a prospect on the forum for a while(DIY hot plate). None have succeeded AFAIK with a final product that I would consider superior to a good commercial product made by a company like Corning or IKA.

If you do truly succeed with designing a hot plate with the stated features and price point, I think that quite a few members on here will be interested, especially those who are just starting out.

Out of curiosity, which hot plates have you taken apart? New-ish hot plates on the cheaper end, or older hot plates from Corning/IKA/etc? Many members on here, including myself, use these older hot plates, available quite cheap used, and I've never heard of an issue with the build quality before.
Quote: Originally posted by mux  
Why use high emissivity ceramic tops? Just because you can then boast 1500W heating capacity? *

There are several reasons that ceramic tops are used on commercial hot plates. One is simply chemical resistance, as these tops are substantially more resistant to chemical attack than the aluminum tops(though notably susceptible to attack by heated strong bases).

They do allow substantially higher surface temperatures to be attained as well, this is a place where I think your design is lacking. I'd be hard pressed to believe there isn't a reason that there aren't any commercial hot plates with aluminum tops that spec surface temps of 500°C without good reason, you may run into substantial structural issues if nothing else. Typically, aluminum variants max out in the neighborhood of 300°C. Unfortunately, going with a ceramic plate will increase the cost of the build as well, not to mention sourcing a suitable material.

Ceramic also has a substantially higher heat capacity than aluminum, which results in a more stable surface temperature; this is very important in some situations.

Texium - 11-12-2016 at 10:17

Yeah ceramic hot plates are superior to aluminum ones for their chemical resistance alone. I would never buy an aluminum hotplate simply because I know that if I had a corrosive mixture boil over it would likely be ruined instantly. The ceramic top of my old 1970s Thermolyne has had to hold off hot conc. sulfuric acid on multiple occasions.

If such a hot plate as described was available (with a ceramic top), I would certainly buy one. Although I love my old relic, I have been dreaming of having thermocouple control for a while now, I just haven't been able to justify the expense.

Sulaiman - 11-12-2016 at 11:59

I think that if you make a quality build then adding features in electronics is relatively cheap,
and once you have control & monitoring for power & stirring
then you may as well add IoT, usb, WiFi, device net .......
I doubt that pople will buy a hotplate stirrer from a small manufacturer based on alledged/unproven quality alone
you need better bells and whistles than the competition
voice command / speech synthesis / braile interface etc.


poly-phase rotating fields is not enough
- you need a 'centralising' force on the stirbar, easily done with magnets, not as easy as you may hope using external coils.
(I am dabbling in this area, if you have success please keep us updated as I'm considering stirring for a hot air bath)




[Edited on 11-12-2016 by Sulaiman]

m1tanker78 - 11-12-2016 at 12:06

I set out to do basically the same thing some months ago out of necessity. I got tired of having good heating without stirring or stirring with laughable heating. I cobbled together a brushless dc motor (with magnetic rotary encoder), programmable logic brain, bluetooth communication, bare-bones android app to monitor and control functions, disposable Al plate/sheet top, Al sheet shrouds for various flasks and a digital heat gun providing the hot air bath. I hacked the heat gun so that the FPGA can read the current setting and send button presses to make adjustments to temperature and fan speed. There were 4 ports set aside for temperature measurement.

One major feature I began work on but didn't have time to develop was a stall detector. I can't tell you how many times I left a slurry stirring only to come back and see that the insolubles had settled and the stir bar was simply rotating about its longitudinal axis -- sometimes just bouncing around in the flask. The stirrer has to be slowed or stopped in order for the stir bar to reengage. Doing so is trivial but detecting when the stir bar disengages is not so trivial in my experience.

The bluetooth link was quite buggy sometimes but that's what I had on hand at the time. A wifi link with an external antenna would be sweet.


mux - 11-12-2016 at 13:01

Quote: Originally posted by aga  
(...)
Perhaps a replacement/upgrade board for a cheap hotplate/stirrer would be easier than building the whole thing from scratch.


This doesn't really solve a bunch of issues, like badly designed mica/nichrome heaters. And not unimportantly: cheap ebay hot plates can just go away or be totally different 6 months on. It's hard to design against an ever-changing platform. Don't get me wrong though, I'd love to reduce the scope of the project.

----

I see there are some misconceptions about aluminum tops. Aluminum can withstand 500C, it just can't withstand 700C nichrome wire and it has a fairly high coefficient of thermal expansion, which means you get issues at high temperature gradients. This is fixable with good design and different types of heaters. I've done a high-temperature aluminum project before for work. All of this isn't to say I'm a know-it-all and all manufacturers are all wrong, just that the temperature argument doesn't pass the sniff test. Relatively pure alloys, non-precipitation hardened, annealed, should work fine up to 500ish C.

Also, 'ceramic' tops aren't always ceramic. A lot of them are aluminum or steel with a ceramic surface (generally alumina). For chemical inertness, alumina enamel should suffice - it's the exact same stuff used for fume hood resurfacing and hot plate resurfacing. This should be exactly as inert to everything but the nastiest of chemicals (e.g. HF - although that can be fixed with some PTFE tape).
Quote: Originally posted by gdflp  

Ceramic also has a substantially higher heat capacity than aluminum, which results in a more stable surface temperature; this is very important in some situations.


That's a slight physics misconception, it's not dependent on heat capacity but on the quotient of heat flux on conductivity. Higher conductivity or lower heat flux = better distribution. Capacity only makes a hot plate more dangerous (by retaining heat after the heater is turned off).

----
Quote: Originally posted by gdflp  
Interesting ideas you have, though this has been a prospect on the forum for a while(DIY hot plate). None have succeeded AFAIK with a final product that I would consider superior to a good commercial product made by a company like Corning or IKA.


I'm very aware of this. IMO there's a distinct lack of proper engineering going on in most DIY hot plate projects. Also documentation. The objective here is to provide a good foundation to work from for other people, or possibly even a decent product (although I'm sure iteration 1 won't please everyone).
Quote: Originally posted by gdflp  
Out of curiosity, which hot plates have you taken apart? New-ish hot plates on the cheaper end, or older hot plates from Corning/IKA/etc? Many members on here, including myself, use these older hot plates, available quite cheap used, and I've never heard of an issue with the build quality before.


It's super hard to judge build quality without taking stuff apart. I've most recently taken apart a set of IKA hot plates, before that... Weidinger, a Dutch branded one (forgot exactly which) and a cheap chinese one. These all used mica heaters (a big no-no, although I understand why, it's cheap and provides high temperature at good isotropy), had horrible electronics (yes, even the IKA, it was just a big pile of poop inside) and so on. On the outside, only the chinese one had any sort of hint that it wasn't great. Mechanical stability, looks, etc. are generally great. Just don't open them.

----
Quote: Originally posted by Sulaiman  

I doubt that pople will buy a hotplate stirrer from a small manufacturer based on alledged/unproven quality alone
you need better bells and whistles than the competition
voice command / speech synthesis / braile interface etc.


Of course. I'm not (primarily) making a product, I'm making a foundational, open-source design. Even if I produce nothing but one for myself, other people can copy and improve this instead of having to reinvent the wheel or making design mistakes.

Quote: Originally posted by Sulaiman  

poly-phase rotating fields is not enough
- you need a 'centralising' force on the stirbar, easily done with magnets, not as easy as you may hope using external coils.
(I am dabbling in this area, if you have success please keep us updated as I'm considering stirring for a hot air bath)


Very good point. I was actually thinking about this over the afternoon - how do they do this in magnetic stirrers? I think the answer is in one hot plate that I tore down: they slightly angle the solenoids inward (or outward) to make an effective magnetic trap. I'm definitely making a standalone stirrer first before attempting the hot plate-stirrer combo, so I can test and optimize this.

------------
Quote: Originally posted by m1tanker78  
(...)

One major feature I began work on but didn't have time to develop was a stall detector. I can't tell you how many times I left a slurry stirring only to come back and see that the insolubles had settled and the stir bar was simply rotating about its longitudinal axis -- sometimes just bouncing around in the flask. The stirrer has to be slowed or stopped in order for the stir bar to reengage. Doing so is trivial but detecting when the stir bar disengages is not so trivial in my experience.

The bluetooth link was quite buggy sometimes but that's what I had on hand at the time. A wifi link with an external antenna would be sweet.


The 'best' way to detect stall is the same as in electric motors: measure the phase angle of current and/or continuously measure current. Ideally, in a 3-phase (6-pole) motor, you just see sinusoidal current as you apply voltage over the poles. If the stir bar (rotor) stalls, one position will consume a lot of current while the 'motor' position perpendicular to that consumes almost none. By either seeing a wildly varying phase angle or big changes in current consumption over the magnetic field rotation period, you can determine that the stir bar is either stalled or not keeping up. That info can also be used to save power (and thus unintentional heating) - just turn down the current until the stir bar stalls.

Do you have a link to your project? Have you documented it somewhere? I'd love to see as many (failed and successful) projects as possible, I too wouldn't like to reinvent the wheel too much and/or fall in the same traps as others did before.

[Edited on 11-12-2016 by mux]

Fulmen - 11-12-2016 at 13:36

For the stirrer I suspect one could use a step-motor driver. These are dirt cheap and could be controlled by something as simple as a astable multivibrator.

Dr.Bob - 11-12-2016 at 16:12

I've used a lot of hotplates. For my money, I have seen more long lasting Corning ones than any other, by a large margin. I have used IKA ones, they are nice when new, but seem to fail quickly, and it has been hard to get parts for them or get them repaired. The Corning ones are pretty good, and last forever. The best electronic ones I have seen are made by Variomag, and they stir well, and have no moving parts, and stir well.

HeYBrO - 11-12-2016 at 17:04

Those are some interesting ideas mux! a member has already "made" a hotplate stirrer if you're curious to see: https://www.sciencemadness.org/whisper/viewthread.php?tid=23... . Obviously it doesn't meet your requirements but maybe it is useful none the less.

[Edited on 12-12-2016 by HeYBrO]

mux - 12-12-2016 at 00:25

Quote: Originally posted by Fulmen  
For the stirrer I suspect one could use a step-motor driver. These are dirt cheap and could be controlled by something as simple as a astable multivibrator.


Any kind of mechanical stirrer motor is not acceptable; that's an easy point of failure in a machine that could regularly get over boiling temperatures in the main unit.

That would also require something else than rare earth magnets, as their Curie temperatures are too low. And that gets nasty quick. That's why solid state seems like the absolute way to go: no moving parts, high temperature tolerance, high reliability, infinite variability (practically).

Fulmen - 12-12-2016 at 02:17

Mux: Sorry for not being clear, I was thinking of a stepper driver for controlling 4 electromagnets. You basically end up with a 4-step motor.

mux - 12-12-2016 at 03:12

Ah, excuse me. Yes, of course, the only problem being that a 4-quadrant control would not have a deterministic rotation direction, so the 'better' mixing methods (back and forth mixing, vibration mixing) wouldn't be easy. Although that can just be fixed by using a second stepper driver...

That's more of an implementation question, though. Plenty of microcontrollers have phase shifted PWM outputs, some up to 12, so I can just as well use one of those and build up discrete coil drivers. I've got plenty of experience with that. The actual choice of parts comes down to (software) complexity and cost. That's a bit further down the road.

Fulmen - 12-12-2016 at 03:43

A stepper driver has full control over the direction, that's the beauty of it. And even a 3$ A4988 driver can microstep (usually 16 steps), giving a 5.6° step angle. That should provide a much smoother operation than many commercial stirrers. I would strongly recommend that you start with this, as more people are interested in (and capable of) building a stirrer than a complete hotplate.

I started looking into this as I desperately need a good stirrer (and a stirrer hot plate now that my old ICA has died). But I can't design electromagnets, and didn't feel like spending hours winding coils by trial and error.

Sulaiman - 12-12-2016 at 04:12

I work in an electronics repair company - anything industrial
repair cost is often not important - resuming production is,
I can usually replace or have rewound transformers and motors,
compatible belts, bearings etc. are easily found
failure of a custom coil or a read-protected micro would mean a write-off.
Reliability is good, so is repairability,
where possible, it would be good if you could use 'standard' parts that are likely to be available in a few years.
A robust, repairable hotplate stirrer could become a classic.
To the point where I would be more likely to buy a well documented kit than a well made product,
because your target market is likely to want to modify/enhance the kit model.
... saves a fortune on assembly costs too :D

. add a hole tapped for retort clamp-stand rods
. if a spot has precise temperature control it could be used for melting point determination
. a removable wind shield that also keeps heat around flasks, beakers etc. would be nice


A small added benefit would be to design the product for postage ... as small and light as possible.


__________________________________________________
A related question: for my mag-stirrer experimenting I bought these
stirbars.jpg - 7kB
I found difficulty getting them to spin stably under various conditions,
has anyone tried this shape ?
are they more or less stable than the round rod type ?

mux - 12-12-2016 at 04:51

Quote: Originally posted by Fulmen  
A stepper driver has full control over the direction, that's the beauty of it. And even a 3$ A4988 driver can microstep (usually 16 steps), giving a 5.6° step angle. That should provide a much smoother operation than many commercial stirrers. I would strongly recommend that you start with this, as more people are interested in (and capable of) building a stirrer than a complete hotplate.

I started looking into this as I desperately need a good stirrer (and a stirrer hot plate now that my old ICA has died). But I can't design electromagnets, and didn't feel like spending hours winding coils by trial and error.


I'm aware of these little chips; have used many of them for 3D printers and such. Unfortunately, you're not dealing with a stepper motor here. Even stepper motors, when they slip excessively (e.g. when you're not putting enough current through the coils) will be able to reverse spontaneously! A stir bar + solenoids is essentially a very high slip motor, so I'm fairly sure that although stirring low viscosity solutions would be fine, large stir bars or very viscous/powder compounds will not go well with a 4-pole stator. I could be wrong of course, but I do know a thing or two about electric machines and a 6-pole solution would be preferable in this situation. Also for centering the stir bar and avoiding spinout.

But of course: the proof of the pudding is in the eating. I'll begin designing a strirrer and testing that concept. I have a couple different roundbottom and erlenmeyer flasks to test with, although I'll have to look into stir bars (I see that there are many fancy shapes).

--------------------------
Quote: Originally posted by Sulaiman  
(...)
where possible, it would be good if you could use 'standard' parts that are likely to be available in a few years.
A robust, repairable hotplate stirrer could become a classic.


Absolutely, 100%. I'm a tiny electronic design company, I can't afford getting custom made stuff anyway. Aside from mechanical parts that have to be milled or 3D printed (which IMO are bearable costs considering typical retail prices of these devices - I'll keep it to 2.5D operations), all the electronics can be made using 100% multi-supplier parts.
Quote: Originally posted by Sulaiman  

To the point where I would be more likely to buy a well documented kit than a well made product,
because your target market is likely to want to modify/enhance the kit model.
... saves a fortune on assembly costs too :D


This is why I'm specifically making this an open source project. I don't necessarily want to make this into a business. It'll either be a kit or a limited release. Also, postage can get pretty bad overseas (I'm in Europe).
Quote: Originally posted by Sulaiman  

. add a hole tapped for retort clamp-stand rods
. if a spot has precise temperature control it could be used for melting point determination
. a removable wind shield that also keeps heat around flasks, beakers etc. would be nice
.A small added benefit would be to design the product for postage ... as small and light as possible.


All great suggestions. Given that I'm starting with the stirrer, that one will almost certainly have a standardized lab stand connection. Not sure if a tapped hole is the best way to go, depends a lot on the weight I suppose.

Instead of a wind shield, I'll probably do what I do with my 3D printer: have a little rubber or silica mat with several cutouts that you can drape over the surface. Basically anything low-emissivity and low conductance will work here. Proper wind shields have the problem of not necessarily fitting on any apparatus, and then it becomes an annoying thing to store.

Postage is going to be an issue either way, because I don't see this ever weighing less than 0.5kg or fitting in a 1" thick box, i.e. it'll definitely fall outside standard sizes for bubble envelopes. For Europe this still isn't a problem (I live in the Netherlands, postage within the EU is cheap enough), but for the US that may mean it'll have to go through a secondary reseller or just be a design that people build for themselves. Again, I'm not building this to be sold in the hundreds or thousands, I don't really take the logistics too seriously at this point.
Quote: Originally posted by Sulaiman  

A related question: for my mag-stirrer experimenting I bought these

I found difficulty getting them to spin stably under various conditions,
has anyone tried this shape ?
are they more or less stable than the round rod type ?


Same question here. Can somebody enlighten us about different rod shapes? I've seen some weird ones in catalogues; little ringed rods, plus-shaped cross-sections, etc. What should I be testing with?

Related: I'll be doing a big Fisher Scientific chemical order this week, do they have anything special stir-bar related that I should buy? They're having an end of year sale with super cheap nitric acid etc., so I'm like a kid in a candy store right now.

Fulmen - 12-12-2016 at 05:47

Quote: Originally posted by mux  
Even stepper motors, when they slip excessively ... will be able to reverse spontaneously

Ummm... A 4 pole "stepper" does create a directional rotating field. With excessive slip one could end up with a harmonic situation where it acts like a 2-pole, but is this really realistic? And wouldn't this be possible with any number of poles using a common straight stir-bar?

But as you say, the proof is in the eating. And don't let perfection be the enemy of good, unless you use a mechanically linked stirrer I think there is always some risk of slippage or spinout.

mux - 12-12-2016 at 07:02

These questions are probably a bit premature, and focus a bit too much on implementation details that honestly I don't even know if they're relevant anyway. It could well be that I don't even have enough space to put in more than 4 poles after doing the calculations on required field strength. I'll come back to this when the design process is a bit further along.

I'd like to recenter the discussion on big picture stuff first: what features are missing, which are bogus and is that a spider on the wall?

Sulaiman - 12-12-2016 at 07:31

AFAIK stir bars rotate synchronously with the applied rotating magnetic field
ANY slip causes chaos.
(there is a phase delay rotation angle as torque is increased)

a mechanical rotating permanent magnet dipole
or three-phase (6-pole) appear to me to be the only viable choices,
theoretically 3 coils could provide a controlable rotating field
what I want is something like 3 Helmholtz coils,
producing maximum field strength and gradienfs near the centre.... tbd

mux - 12-12-2016 at 07:48

Yes, that was my initial idea as well. Exactly.

I've just done some back of the envelope calculations on attainable field strength, and it's not too great. It's looking like I can reasonably expect to generate something in the order of 100mT - maybe 150mT - with 2A drive current, about 8W power consumption and a suitably small solenoid (about 1cm3 winding volume). That's about 1/3rd the typical field strength from a neodymium magnet as found in DIY stirrers. I mean, it's in the right type of order of magnitude, but it's clearly going to be a challenge.

I'm going to build a little calculation/iteration design tool to optimize this and do some FEM modeling to see how I can get the shape of the magnetic field just right.

Bert - 12-12-2016 at 07:57

My favorite stirring hot plate uses a quartz heating element under a glass top. 16 years, plenty of spills and other mistakes, and only a few scratches on that quartz glass.

Bibby Sterilin. I have not seen another like it, and got it new for about USA $250.00 delivered. Heck of a deal. It is set up with a socket and removeable jumpers for remote control of heating, so a PID or such can be used. The stirring is controlled only from the unit. Does have a socket for a support rod, horizontal and out the back, rather than vertical and on top.

http://www.bibby-scientific.com/

Yes, it would be nice to monitor and adjust the stirring and temperature from an iphone.

I suspect that by the time one adds all the nice features desired, this is not going to be cheaper. But possibly having your equipment user designed/customized would make it worthwhile regardless of any savings?

mux - 12-12-2016 at 09:19

There is no additional hardware cost to the connectivity features. As a rough guide, the major costs are, from most expensive to cheapest:
- Electrical safety certification ($1000ish in-all, or $2500+ if I want an official registration) one-time
- Milling parts ($10-15 per unit)
- Heater ($10 per unit)
- Electronics (<$10)
- Assembly ($4-6 per unit)
- Power supply ($5ish, not sure yet)

For reference; an ESP8266 module costs $2 and provides wi-fi functionality. Add another dollar for an FCC/EU certified one. Adding in a thermocouple connector and readout is less than a dollar. These kinds of features are cheap, but they do tend to become a hurdle to completion as software complexity balloons. So you do want to keep those in check, even though they don't add significant dollar amounts to the final product.

It's a hard requirement to stay on budget. If I go over budget, features get scrapped starting from the least requested features. All pricing will be done from multiple suppliers.

If this turns out to be sufficiently popular (and again, I'm not aiming for that), a certified reference design will explicitly NOT be customizable. Customize it, do (almost) anything differently from the prescribed design, and you void any and all guarantees of electrical safety. If this stays niche enough, you can usually get away with pretty much everything, you just can't blame anyone if you blow your fuses or if your house burns down.

Fulmen - 12-12-2016 at 11:37

Quote: Originally posted by Sulaiman  
AFAIK stir bars rotate synchronously with the applied rotating magnetic field

*thinking* You're right, it must run synchronous. But that applies to mechanically driven permanent magnets as well, doesn't it? My IKAMAG has a permanent magnet, and it's not immune to overspeed.

Quote: Originally posted by mux  

I've just done some back of the envelope calculations on attainable field strength, and it's not too great ... That's about 1/3rd the typical field strength from a neodymium magnet as found in DIY stirrers.

That's about where I stopped myself, I just didn't know enough about magnets and required field strengths to build a well designed unit.

However, the small commercial units I've seen didn't have particularly strong coils. Even my old IKAMAG has a fairly weak magnet compared to modern neodymium magnets. And while not perfect for any and all application they seems to work OK for most uses.

careysub - 12-12-2016 at 11:45

Quote: Originally posted by zts16  
Yeah ceramic hot plates are superior to aluminum ones for their chemical resistance alone. I would never buy an aluminum hotplate simply because I know that if I had a corrosive mixture boil over it would likely be ruined instantly.


OTOH, with a home-build aluminum top stirrer the aluminum plate could be cheaply replaced after such an incident, about $5.

mux - 12-12-2016 at 11:52

Quote: Originally posted by Fulmen  

That's about where I stopped myself, I just didn't know enough about magnets and required field strengths to build a well designed unit.

However, the small commercial units I've seen didn't have particularly strong coils. Even my old IKAMAG has a fairly weak magnet compared to modern neodymium magnets. And while not perfect for any and all application they seems to work OK for most uses.


I've just purchased a few comparatively monstrous stir bars (11x54mm) from eBay (location said UK, so they should arrive soon), I'll be experimenting with that in one of my Erlenmeyer flasks with some... yoghurt or something. That should be pretty much the worst-case situation.

I have a bunch of A4988s (on a 3D printer dev board) as well, I'll wind a couple coils with 28AWG magnet wire and see how that works. That should give me a good experimental setup to independently test required field strength, distance from the stir bar, distance between magnets, etc. Expect a video on this somewhere in the next 2 weeks.

Bert - 12-12-2016 at 12:29

Quote: Originally posted by careysub  
Quote: Originally posted by zts16  
Yeah ceramic hot plates are superior to aluminum ones for their chemical resistance alone. I would never buy an aluminum hotplate simply because I know that if I had a corrosive mixture boil over it would likely be ruined instantly.


OTOH, with a home-build aluminum top stirrer the aluminum plate could be cheaply replaced after such an incident, about $5.


What is the price for a 300mm square top for an IR hot plate?

BTW, 900 watts. The response time to a change in heat setting on this is FAST compared to the ceramic block units with internal resistance elements- kind of like the difference between cooking stove top with gas range and the slower to react electric ones.

(Can you tell I like these?)

1786-1.jpg - 161kB

Fulmen - 12-12-2016 at 13:16

mux: Sounds great, you do appear to be the right man for this job.

Let me know if I can help. I can do metal fabrication (lathe, mill, heat treatment), calculations, CAD/blueprints etc.

[Edited on 12-12-16 by Fulmen]

aga - 12-12-2016 at 14:33

Is this Idiocy in Action or an attempt to build 1 (one) Super hotplate/stirrer ?

Sale Price, Cost, Profit !

If none of these matter, groovy.

If so, dead horse. No way to compete with the mass-producers by making something pretty much the same as they alread make on a vast scale.

In the end, they heat and stir, which is what people want.

Edit:

Definite Lathe and Mill jealousy. Gotta get me those machines.

More edit:

More the Lathe. Can do basic milling with a lathe.

[Edited on 12-12-2016 by aga]

Fulmen - 12-12-2016 at 19:24

Honestly, I don't understand how people can get by without a decent machine shop.

careysub - 12-12-2016 at 19:34

Quote: Originally posted by Fulmen  
Honestly, I don't understand how people can get by without a decent machine shop.


Creativity, hand tools, light power tools, knowing how to source ready-made dimensioned metal pieces when needed (infrequently, and rarely do more than two dimensions need to be exact) and knowing how to make things out of composites - fiberglass, carbon fiber, wood, epoxy, and doing casting/molding with various materials (polymers, sodium silicate and vermiculite/perlite, etc.).

You'd be surprised what I can do with a metal file.

Being addicted to precision machining can stifle creativity and learning.

Along these lines it is interesting to study the machines made in the early industrial revolution before the availability of precision machining. The early grasshopper steam (or beam) engines for example were made without any precision machined parts (though the piston and cylinder had to be ground to a close fit). All the parts had large tolerances, and were worked to fit when necessary.

The Wikipedia page does not discuss this interesting and important aspect of the designs origin though:
https://en.wikipedia.org/wiki/Grasshopper_beam_engine

[Edited on 13-12-2016 by careysub]

careysub - 12-12-2016 at 19:45

Quote: Originally posted by aga  
Is this Idiocy in Action or an attempt to build 1 (one) Super hotplate/stirrer ?

Sale Price, Cost, Profit !

If none of these matter, groovy.

If so, dead horse. No way to compete with the mass-producers by making something pretty much the same as they alread make on a vast scale.

In the end, they heat and stir, which is what people want.
...


The key is making something does exactly what you want it to do, rather than matching up to the broad demands of a mass market.

I make telescopes because what I make, you cannot buy.

It is possible to make a piece of equipment that meets your specific needs for a fraction of what a commercial offering that would be required to do the same thing (but lots of stuff you don't need) would cost.

mux - 13-12-2016 at 01:32

Quote: Originally posted by Fulmen  
mux: Sounds great, you do appear to be the right man for this job.

Let me know if I can help. I can do metal fabrication (lathe, mill, heat treatment), calculations, CAD/blueprints etc.

[Edited on 12-12-16 by Fulmen]


There's a decent chance shipping to the US won't be possible (at an acceptable price). If the design works and a kit is designed, US 'customers' (=forum members) will need to be served from within the US, as far as they're not making the units themselves. In that case: yes.

Otherwise, I have plenty of robots to do the work for me. I don't have a lathe, but otherwise I'm fully equipped.

Quote: Originally posted by aga  

Sale Price, Cost, Profit !

If none of these matter, groovy.

If so, dead horse. No way to compete with the mass-producers by making something pretty much the same as they alread make on a vast scale.
[Edited on 12-12-2016 by aga]


This is unfortunately all too often a gross misunderstanding of how to attack a problem. If you don't set a budget, if you don't limit your design space, projects get out of hand quick. Feature creep and delay is the natural state of things.

I don't make a budget and define sale price and margins because I intend to sell, I do this to limit myself and the scope of the project to a degree that makes sense.

For personal projects, nothing matters. You just want a thing that's for you and you alone, and you take full responsibility for whatever happens. But as soon as you build something with the intent to improve the world, you HAVE to take into account the limits of your project scope and you HAVE to systematically attack the problem.

Hopefully that better explains why I'm putting so much emphasis on the design process and not necessarily the implementation details.

j_sum1 - 13-12-2016 at 02:22

I don't have a stirrer hotplate. It is (perpetually) on the list of things to get. If this gets off the ground then I would happily buy a kit. And if the design and quality is good (as seems to be the goal here) then I am fine with paying a bit of shipping.

I would probably never use wifi or bluetooth but good temperature control would be great. And the ideas floated for stirring sound fantastic.

Fulmen - 13-12-2016 at 04:34

Quote: Originally posted by mux  
There's a decent chance shipping to the US won't be possible (at an acceptable price).

I wasn't really thinking of manufacturing kits, but as I am very interested in the stirrer I could at least build some prototypes for R&D/testing.

In fact I was contacted by an acquaintance who sells brewing supplies regarding the possibility of making stirrers units. If this produces any results I might be able to sell some through him.

Dr.Bob - 13-12-2016 at 05:32

The oval stirbars are only for round bottom flasks, they must be a smaller radius than the flask in order to stir well, so that they spin on the center only. That is why most rbfs have a scratched area right on the center bottom. The straight rods ones work fine for most beakers and erlenmeyers, the hexagonal ones simply have a slightly more surface area but there is no real difference. But having a center pivot ring (larger diameter in the center) simply lifts the rod off of the flask, except at the center, creating a bearing of sorts to rotate on, that helps some, mostly when stirring a suspension or slurry.

There are also many various odd ones, crosses, discs, windmills, and such, some of them work well, but many are hard to drive without a good stirrer.

If anyone is looking for a used stirring hotplate, I do have a few that I have finally found and finished testing, almost all are Corning models. I will post some photos soon, when I get a moment of time. I am trying to repair some of the broken ones as well, but also have some that could be parts donors for people who want to play with them. So if you want to play with a broken one, I can sell them for cheap, mostly just the shipping costs. Just send me a u2u if you are interested, but I may be slow to answer, as this is a very busy period for me, so please be patient with me.

mux - 13-12-2016 at 06:00

Would be nice to have a couple of random stir plates to learn from. I'd be happy to take them off your hands. I'll send a U2U.

Elemental Phosphorus - 14-12-2016 at 15:09

I would be interested in buying a used hotplate stirrer. What would you charge?

mux - 15-12-2016 at 01:34

We're not nearly there yet, first it has to be designed, then it has to work, then we can think about selling :)

Fulmen - 15-12-2016 at 11:17

I'm assuming he was talking about Dr. Bob's offer, something that should be taken in private.

mux - 16-12-2016 at 13:12

Well, I guess you can chalk that up to my growing problem of misinterpreting forum posts. Nevermind.

Tiny status update: I've set up a little experiment with an Allegro A4983 stepper board (custom 3d printer control board I built a long time ago) and four 300-winding solenoids. The results so far are... let's just say that this could be an entirely new, exciting product called the Mixture Basher!

https://www.youtube.com/watch?v=6touhF9ISHE

I don't have a proper stir bar handy at the moment, and these hard disk magnets don't really have such well-defined magnetic fields. That really exposes some centering issues. Even if I can get it to properly rotate, it's a bit stop-start.

Also, this is already with ~300 turns per cm. I can maybe pull off twice as much, but that's about the max. Input power is 10W now, and although it's easily strong enough to stir low-viscosity liquids, it's not going to do anything beyond the consistency of cordial.

I mean, you can't expect much progress to be made in a week, but it's already clear that this is going to need a ton more experimenting.

Fulmen - 16-12-2016 at 13:41

Did you see this post: http://www.sciencemadness.org/talk/viewthread.php?tid=3056&a...

mux - 16-12-2016 at 13:47

I've been browsing and cataloging various different projects, but hadn't encountered that thread yet. Thanks!

Fulmen - 16-12-2016 at 15:15

From the looks of it you might need to focus the field more, how were the magnets arranged?

eesakiwi - 17-12-2016 at 03:12

I 'focused the field' of two hard drive magnets by placing two cylindrical magnets on them where I wanted the field to be. They were neo magnets as well.

I expect they will use the field from the flat hard drive magnets to make them stronger. Now the field that gets up past the plate is in two spots and quite strong.

My commericailly made stirrer uses just a normal fan motor, generic sort, like from a microwave oven (easy to get) and it uses a wire wound rheostat to control the revs/speed of it. Does not make sense to me how it works but it does.
The magnets are 1/4 inch dia by 1 inch, mounted vertically in a 1 1/2 inch dia by 2 inch long Aluminium slug/cylinder that's mounted on the fan motors shaft.
It has no hotplate.

mux - 17-12-2016 at 12:53

Quote: Originally posted by Fulmen  
From the looks of it you might need to focus the field more, how were the magnets arranged?


20mm from the center, no iron core, perpendicular to the surface. No real attempt to focus or direct the field in any way.

I have to stress, again, this thread is now just a little work log of progression on the project. Don't expect good results in every post I make :)
Quote: Originally posted by eesakiwi  
I 'focused the field' of two hard drive magnets by placing two cylindrical magnets on them where I wanted the field to be. They were neo magnets as well.


I'm going to try to use straight inductors first; they have a ferrite core and a decent amount of windings, would save me a lot of manual winding. The only downside would be that the hot plate would be slightly magnetic even when you turn off stirring.
Quote: Originally posted by eesakiwi  

The magnets are 1/4 inch dia by 1 inch, mounted vertically in a 1 1/2 inch dia by 2 inch long Aluminium slug/cylinder that's mounted on the fan motors shaft.
It has no hotplate.


A big reason why this wouldn't fly in my design is the fact that there is a (500C) hot plate on top of the stirrer unit. Any kind of motor would perish pretty quickly in that sort of environment. Unless you seal and cool it, which costs money, adds complexity, etc. etc. yada yada

[Edited on 17-12-2016 by mux]

Fulmen - 17-12-2016 at 18:14

Quote: Originally posted by mux  
Don't expect good results in every post I make :)


Unacceptable! I DEMAND progress :P

Without an iron core your results are much more understandable, the magnetic permeability of air is far to low to produce an effective magnet.

mux - 23-12-2016 at 05:58

Alright, another non-update unfortunately.

I ordered various items to test with more than a week ago (from the UK, so I expected quick delivery). Unfortunately, nothing has arrived so far and I'm now entering the holidays and all that. Long story short; probably no updates until halfway through January.

The good news: I have experimented with various ferrite-cored inductors, and those do give me well-defined rotation as well as very good torque. I'm pretty sure that as soon as I have some big stir bars, I'll be able to verify that it works well enough and get going on a proper design.

Fulmen - 23-12-2016 at 07:25

That's encouraging news, keep up the good work. Have you tried the configuration that kmno4 used or simply 4 separate magnets? If so you can expect a further increase in torque.

Hotplate Stirrer Design Project

Photonic - 24-12-2016 at 01:56

I think this is a cool project, however, I think you're probably way low on the actual cost, although I hope you can do it. R&D is expensive!

I'm happy to lend a hand in what I can do. I've designed thermocouples, data loggers, and power supplies. My latest one was on the MAX31856 which is quite honestly one of the chips I'd recommend for your project. It's low cost comparatively, and digital. It also accepts external cold junction sensors which is quite ideal for a project like yours. 0.007*C resolution too! (Yes, that's 0.007!)

I know you're thinking solid state stirrer but you have to think about power consumption as well. Instead of stepper motors check out some brushless motors especially with the advent of camera gimbal systems. They can rotate very slowly, and can carry some torque!

On the WiFi side you also have to consider emissions tests which will add quite a bit to your project. One of my big concerns is the safety with making a mains connected kit. Most "kits" don't consume these levels of power so you have to factor that into your end project. What you might be able to do though is separate the project into modular blocks. This way you can keep the high power side isolated from the others and then you can sell the other half of it without having to require safety testing, emissions, and so forth as long as it's sold in kit form.

I can CAD, program embedded systems, and might be able to assist with some machining as well.

With all that said good luck!!! I'll be keeping an eye on this project and wish you the absolute best. I look forward to seeing your prototypes and how this progresses. Sounds like everyone here including myself is more than willing to help out in any way that we can!

mux - 24-12-2016 at 02:50

I suppose we have almost perfectly overlapping competences, so that would make for a great schematic/mechanical design/BOM checking buddy. I'll happily use your expertise for this project!

As for the cost side of things; It's pretty simple: that's a requirement, so I won't go over it. I've back-of-the-enveloped it to work and I'll just tweak the details to actually make it work if some kind of cost setback arises. If it doesn't work then the project just crashes, because I honestly don't see good value in a $200+ unproven, random-guy-made-it hotplate stirrer. I'd make one for myself obviously, but that's it.

As for the stirrer: Solid state does significantly up the power requirements, which may be the biggest issue. Unfortunately I don't see this working at all with a moving solution, as
1) I cannot get second suppliers for any motors that would reasonably fit and have suitable voltage requirements, torque and connection schemes. Not having a second supplier is fine for one-offs, but kills a series-produced project.
2) It's a very easy point of failure, and due to the nature of the project I'm pretty sure that some single-supplier motor will not be available by the time your stirrer fails, nor will I be around to help you (I intend to, but from a user perspective this project should be robust enough to allow for me to disappear)
3) How do I guarantee safety in flammable/explosive atmospheres? I can't pot a motor. Brushed motors are immediately disqualified, meaning I need to throw in a frequency drive as well. It's a bit of a rabbit hole. Of course you should buy intrinsically safe equipment if you do this, but let's be honest. People are going to generate H2 near this thing, and that has an explosive mix at like 5-95% concentrations.

For such a small project where I don't have the luxury of ordering specific motors with a reasonably far-out backorder option (e.g. 10 years), it's a much less maintainable option as well as just having mechanical issues. Motors are almost definitely also going to be more expensive. Of course I know that power consumption, consequently heat dissipation as well as magnetic field strength is going to be poorer with solid state, but there are a lot of compelling reasons for me to try solid state first.

Fulmen: I'm still using straight up magnets, now with a ferrite core, no further concentration or field shaping efforts. Ideally I would just use off the shelf inductors as much as possible (because that would mean no custom parts necessary, which is awesome).

Edit: forgot about the wifi emissions thing. Yes, that's definitely the biggest cost. I do power supply design for a living, so I can handle the AC safety issues, but I don't have an anechoic chamber or proper spectrum analyzers/antennae/etc. to do emissions testing. I do have a testing facility I can use, but that would run about €1000 for a full test suite. Maybe I can call in a favor or something, we'll see. Depending on actual cost, I might opt for a pre-certified wi-fi module instead.

[Edited on 24-12-2016 by mux]

Melgar - 24-12-2016 at 02:54

How practical would it be to drive the stirrer with a belt? That way, you could use any old cheap, high-speed DC motor to drive it, just by using a high gear ratio, and place the motor to the side, where it won't be in any danger of overheating or being affected by the magnetic field of the permanent magnet? I mean, if the goal is low cost, you can't do much better than that.

Sulaiman - 24-12-2016 at 03:07

very practical.

I have repaired famous brand multi-position hotplate-stirrers that used rubber belts, one motor to turn up to 8 stirrers.

Fulmen - 24-12-2016 at 04:20

Mux: I see your point, although as kmno4 showed it can be done quite simple. At lest test this configuration before making a decision, I have a suspicion it could improve performance noticeably. Wether you use this for increased torque or reduced power consumption is up to you, but it just might put this in the range where a simple stepper driver would work. I know I'm fixated on these, but when one looks at cost, simplicity and availability it's really hard to ignore that solution.

mux - 24-12-2016 at 13:27

Quote: Originally posted by Fulmen  
(...) it might put this in the range where a simple stepper driver would work. I know I'm fixated on these, but when one looks at cost, simplicity and availability it's really hard to ignore that solution.


Well, it's already well within the range of stepper drivers, because I'm already getting sufficient torque without the pointy bits with A4983's. So you're already vindicated. However, I have to stress that the proof of concept stage is very far removed from appliance-like 'it just works'. My comments aren't meant to shoot down helpful suggestions, I just want to immediately name pros and cons so people know what I'm aiming for. For instance:
Quote: Originally posted by Melgar  
How practical would it be to drive the stirrer with a belt? That way, you could use any old cheap, high-speed DC motor to drive it, just by using a high gear ratio, and place the motor to the side, where it won't be in any danger of overheating or being affected by the magnetic field of the permanent magnet? I mean, if the goal is low cost, you can't do much better than that.


That belt is more temperature sensitive than the motor, but that's not the sensitivity issue here! I should clarify more because the motor suggestion keeps coming back in one way or another.


Motors can be really cheap - under a dollar a piece including shipping from eBay. Incredible, really. Unfortunately, you get what you pay for. Cheap motors have either bushings or really loose-tolerance bearings, which means as they rotate, they'll just rattle themselves loose over time. As torque scales with current, at high current (and thus high torque, e.g. at low RPM) you'll generally burn through the enamel of the windings. Brushed motors spark and erode over time. Basically: it'll work just fine for a bit, but it's bad design practice to fit a high reliability instrument with this kind of motor.

Also, these motors generally have plastic backs (low melting temperature) and low Curie temperature laminations. What do you think the sprocket or belt guide is made of? Nylon, crimped onto the shaft. Even if it's actively cooled, things can still easily get up to the Tg of plastics (70-80C) when the hot plate is at 500C. After a few hundred cycles, that plastic will slip on the shaft. Or you need active cooling, meaning a fan, meaning the fan will eventually fail because that's what fans do, also meaning you get dust inside everywhere.

So what do you need? You need a motor with decent torque at low speed, a decent speed range, that keeps cool at low speeds, no brushes that wear out or really long-lived brushes, a proper thrust bearing in the back and a decent tolerance regular ball bearing in the front and no overly temperature. These, unfortunately, are not cheap. Think ~$10, maybe even more at the quantities we're using here. And you'll likely have to order it custom (there are standardized ways to designate motors, so you can ostensibly make it second supplier that way).

Four or six ferrite-cored inductors cost ~$0.50. The driver chip is another $2. No thermal issues (Tc of ferrite is well over the melting point of the solder used to connect them up), no longevity issues, no supply issues and certainly cheaper than a quality motor. And it can be manufactured in an automated way, no hand assembly of moving mechanical parts. An electrical engineer's wet dream.

Melgar - 24-12-2016 at 13:51

Quote: Originally posted by Sulaiman  
very practical.

I have repaired famous brand multi-position hotplate-stirrers that used rubber belts, one motor to turn up to 8 stirrers.

Interesting. Considering that ink jet printers typically have at least three DC motors in them, and several belts as well, the only thing that would need to be purchased would be the second wheel, and most likely, some bearings. Skateboard bearings are usually very high quality for the price, so assuming someone can find a discarded ink jet printer, (in my experience, not hard to find at all) the rest would be easy to source. And, like you say, you can run multiple stirrers from a single motor.

Melgar - 24-12-2016 at 14:49

Quote: Originally posted by mux  

That belt is more temperature sensitive than the motor, but that's not the sensitivity issue here! I should clarify more because the motor suggestion keeps coming back in one way or another.

Well, considering that every magnetic stirrer I've ever seen or used has been motor-driven, that's to be expected. The belt and second wheel could be easily shielded with, say, aluminum, possibly coated with furnace cement or plaster, if that became an issue. Then you'd have only the shaft and magnet protruding through the shielding. The shaft could be stainless steel, which has really low thermal conductivity for a metal, and is probably the most commonly-used material for shafts at this scale anyway.
Quote:
Motors can be really cheap - under a dollar a piece including shipping from eBay. Incredible, really. Unfortunately, you get what you pay for. Cheap motors have either bushings or really loose-tolerance bearings, which means as they rotate, they'll just rattle themselves loose over time. As torque scales with current, at high current (and thus high torque, e.g. at low RPM) you'll generally burn through the enamel of the windings. Brushed motors spark and erode over time. Basically: it'll work just fine for a bit, but it's bad design practice to fit a high reliability instrument with this kind of motor.

You know, sometimes I feel like an idiot on these forums, since a lot of y'all were chemistry majors and I was a mechanical engineering major. But you're in my wheelhouse now, and seem to have totally missed my point regarding motors. That being, you can rotate a shaft at lower speeds by driving it with a belt, using a very small wheel on the motor, and a much larger wheel on the shaft you're trying to rotate. That way, the motor can spin at higher speed and lower torque, thus avoiding the sort of coil burnout you're referring to.

Quote:
Also, these motors generally have plastic backs (low melting temperature) and low Curie temperature laminations. What do you think the sprocket or belt guide is made of? Nylon, crimped onto the shaft. Even if it's actively cooled, things can still easily get up to the Tg of plastics (70-80C) when the hot plate is at 500C. After a few hundred cycles, that plastic will slip on the shaft. Or you need active cooling, meaning a fan, meaning the fan will eventually fail because that's what fans do, also meaning you get dust inside everywhere.

And everything will eventually die with the heat death of the universe. I don't see your point. And you apparently failed to see mine too, since one of the other major advantages of using a belt drive would be that you can drive the magnet from a short distance away, where temperatures are much lower.

Quote:
So what do you need? You need a motor with decent torque at low speed, a decent speed range, that keeps cool at low speeds, no brushes that wear out or really long-lived brushes, a proper thrust bearing in the back and a decent tolerance regular ball bearing in the front and no overly temperature. These, unfortunately, are not cheap. Think ~$10, maybe even more at the quantities we're using here. And you'll likely have to order it custom (there are standardized ways to designate motors, so you can ostensibly make it second supplier that way).

Uh, no. I can virtually always find DC motors for free when I look for them, and in all the projects I've ever built, I've never had one burn out that I didn't run with the expressed purpose of burning it out. (ie, measuring what factors are indicative of a motor about to burn out.) And even these motors were surprisingly difficult to burn out, even when packed in fiberglass insulation, with a mixture of sandblasting grit and sugar poured into their cases. We've come a long way since the days of Michael Faraday, and even cheap DC motors can be surprisingly resilient.
Quote:
Four or six ferrite-cored inductors cost ~$0.50. The driver chip is another $2. No thermal issues (Tc of ferrite is well over the melting point of the solder used to connect them up), no longevity issues, no supply issues and certainly cheaper than a quality motor. And it can be manufactured in an automated way, no hand assembly of moving mechanical parts.

How many transistors do you think are on that chip? What happens if one of them fails? How likely is it that you did your calculations are wrong and the voltage spike from shutting off the inductors fries your board? Electrical failures are at least as common as mechanical ones, and much more difficult to see coming.
Quote:
An electrical engineer's wet dream.

Hence, the reason that mechanical devices tend to be designed by mechanical engineers.

mux - 25-12-2016 at 00:53

It seems like you're missing my point, possibly on purpose (because I refute your refutations in the text itself!). This topic is not to bicker at each other, this thread is here to help each other primarily.

But, to go into the as-yet partially unaddressed issues. First of all - salvaged parts are not an option here; this is a uniform design, I can't depend on people having specific salvaged parts.

Second - I'm not disputing that motors can work. They clearly can, because a large part of stirrers already use them. It's just that I don't see sufficiently cheap motors doing the job. Anything that would work is pretty expensive - too expensive to fit this project's budget, given that I am shooting for a certain degree of reliability and performance. We're not talking qualitative statements here, we're purely talking about quantitatively limiting design space.

Look inside any reasonable quality hotplate stirrer and look at the motor inside. It's not a cheap 6V DC no-name motor. It's a brand name motor with proper bearings and fairly high power rating. You don't find those for nearly the budget I'm on. And going with a cheaper motor means you need to add other stuff to account for the points on which the cheaper motors have skimped, or adding little milled motor holders to account for differences in sizing of motors as suppliers change. This, again, is not going to be cheap.

Anything can be engineered, you can solve any nail-problem with any hammer-solution, it's just that finding the optimum is the true art.

I've looked into buying motors for this job. I have not found anything. Can you find something better? I'm specifically looking for:

- Sub-$4 complete assembly
- At least two long-term suppliers
- near enough to 0.1Nm torque at 10-2000RPM (0.05Nm is okay, 0.01 is too weak)
- Hard-brushed or non-brushed (i.e. 5000+ h lifespan)
- Can be side-loaded and axially loaded (order single digit N)
- Has a datasheet

[Edited on 25-12-2016 by mux]

[Edited on 25-12-2016 by mux]

Fulmen - 25-12-2016 at 01:35

I'm with mux here, for a build like this the motor-less design is perfect. The drivers are a few dollars and the inductors can be wound using very simple equipment. Once you start dealing with motors and belts you quickly end up with a more complex support structure as well.

Mux: You could try simply bolting the coils to a steel plate (turning them into two horse shoe magnets), even without the "arrows" this should increase field strength. Using steel bolts in stead of ferrite cores shouldn't affect the performance as long as you design the coils for it.

[Edited on 25-12-16 by Fulmen]

Sulaiman - 25-12-2016 at 05:29

if possible do not use thick steel plates or large steel nuts/bolts as part of an a.c. magnetic circuit as eddy-current heating will be a problem (3000 rpm= 50 Hz etc.)
air, powdered or thin strip steel etc are prefered.

Melgar - 7-1-2017 at 14:20

Quote: Originally posted by mux  
First of all - salvaged parts are not an option here; this is a uniform design, I can't depend on people having specific salvaged parts.

You could design it using the most commonly-used parts available, that can be found either off-the-shelf, or in salvaged equipment. That would seem to be the best solution, as we don't want to assume that nothing that we design will ever break or need replacing. We could then compile a list of sources for parts that can be used to replace anything in the design.
Quote:
Second - I'm not disputing that motors can work. They clearly can, because a large part of stirrers already use them. It's just that I don't see sufficiently cheap motors doing the job. Anything that would work is pretty expensive - too expensive to fit this project's budget, given that I am shooting for a certain degree of reliability and performance. We're not talking qualitative statements here, we're purely talking about quantitatively limiting design space.

Then how come solid-state stirrers are always orders of magnitude more expensive than motor-driven ones, and seen to be aimed at applications where motor-driven stirrers wouldn't work as well? This is the part of your point that made no sense at all to me. It's trivial to thermally isolate the motor from the hotplate, so you DON'T need a motor that can withstand all of the thermal stresses that you're talking about. And most magnetic stirrers that I've seen have had fans mounted directly on their own shafts as a means of circulating air past the motor. That, combined with the fact that heat tends to rise, and the motor is typically located below the plate, is usually all the protection a motor needs.
Quote:
Look inside any reasonable quality hotplate stirrer and look at the motor inside. It's not a cheap 6V DC no-name motor. It's a brand name motor with proper bearings and fairly high power rating. You don't find those for nearly the budget I'm on. And going with a cheaper motor means you need to add other stuff to account for the points on which the cheaper motors have skimped, or adding little milled motor holders to account for differences in sizing of motors as suppliers change. This, again, is not going to be cheap.

This might be because a lot of scientific equipment was made in the US or Western Europe until fairly recently, and was often overdesigned. Not that that's a problem, but if you open up any of the newer Chinese-made devices, their motors are nothing special. Also, for holding the motor in place, it's pretty common practice to have mounting holes on the front face of the motor, so that its diameter and length can be variable, thus allowing the engineers to swap out the motor for a different one if it proves to be inadequate, without having to redesign the whole product. This also provides the motor with a heat sink, if it needs one. So this problem you speak of has already been solved in industry, you just didn't realize it. For more information, see here:

http://reprap.org/wiki/NEMA_Motor

This is just for stepper motors, which are more than adequate to meet a hotplate's needs, (an but will require a controller chip for fancy stuff like oscillating stirring. If you just want a DC motor that can work at low speeds without burning out
Quote:
Anything can be engineered, you can solve any nail-problem with any hammer-solution, it's just that finding the optimum is the true art.

I totally agree, which is why I think your "wet dream" comment is pretty telling of what kind of hammer you're holding. In this situation though, I think it's important to not depart too far from the tried-and-true solutions that are already out there, since then you're far more likely to run into an unforeseen problem that it's beyond your means to solve.
Quote:
I've looked into buying motors for this job. I have not found anything. Can you find something better? I'm specifically looking for:

- Sub-$4 complete assembly
- At least two long-term suppliers
- near enough to 0.1Nm torque at 10-2000RPM (0.05Nm is okay, 0.01 is too weak)
- Hard-brushed or non-brushed (i.e. 5000+ h lifespan)
- Can be side-loaded and axially loaded (order single digit N)
- Has a datasheet

Two of these specification requirements seem absurd to me. As long as you get a motor with a standard mount, there's no need to lock in to any specific supplier. Also, 10-2000 rpm? 10 RPM is one rotation every six seconds. Why on earth would you need something that slow? In any case, your best bet is probably a brushless DC motor, or maybe even a stepper motor, since they're fairly cheap, commonly found in appliances in case someone wants to build one themselves for even less money, can either be controlled by an inexpensive IC or an Arduino if someone wants to get fancy, they don't have brushes (as the name implies, and neither does a stepper), a brushless DC motor has a huge speed range at near-constant torque, and you can even do fancy stuff like rotate back and forth, presumably using an Arduino for controlling it.

Your $4 for the entire assembly requirement seem a bit unrealistic to me. First of all, unless your time is worth less than $1 an hour, winding your own ferrite cores is going to be tedious as hell. And even if your components can handle the voltage spikes from them, are you 100% sure that every one of them can handle those spikes for 5000 hours? Are you taking the PCB cost into consideration?

I'm quite certain that an alnico magnet, a standard-mount brushless DC motor, and the controller, could be had for under $10, but I can't imagine you're taking every cost into consideration if you think you can build an induction one for that cheaply.

Sulaiman - 8-1-2017 at 05:36

If you are looking for a motor that has proven long life in a high temperature environment,
with integral heatshield and enough power, yet simple to operate with no sparks,
I suggest an electric oven fan, e.g. £13 retail eBay UK, bulk must be really cheap.
and guaranteed alternate suppliers for the forseeable futire.


Fan_Motor.jpg - 472kB

mux - 8-1-2017 at 08:09

Quote: Originally posted by Melgar  

I'm quite certain that an alnico magnet, a standard-mount brushless DC motor, and the controller, could be had for under $10, but I can't imagine you're taking every cost into consideration if you think you can build an induction one for that cheaply.


What I'm trying to get at here is that I'm more than 'pretty sure' that it won't, because I did some preliminary component sourcing and so far got the opposite. As much as I appreciate your input (and I do, very much so, this kind of stuff needs to be looked at critically, don't get me wrong when I counter your arguments), at this point I need concrete sources, with full cost analysis, not (educated) guesses. Without a hard counter to what I'm building now, I will continue down this path until I either succeed or fail.

In fact, a significant part of the additional cost in engineering a motor-driven solution is in assembly - my time. An inductive drive costs so little because everything is cheap: I can buy inductors off the shelf for $0.20/pc which are essentially pre-wound ferrite cores and the mounting complexity is all contained in the PCB.

Sulaiman again confirms my suspicion that any kind of motor that would comply with my mechanical and environmental constraints would be significantly over budget. Note that those 13 GBP do not include mounting, electrical drive and the like.

---------------------

In other news, I'm back from my holiday, so I'll be slowly chipping away at this project. During the holiday I couldn't resist building a very preliminary mechanical and electrical design for a stirrer-only project:





What you're seeing here is an exploded view of the mechanical design as well as the (again, super-preliminary, this-wont-actually-work) electronics. The mechanical design is an exploded view, showing the overall shape (bit like a hockey puck). It's made of a top aluminum block which houses the inductors (which sit on their own round-ish PCB), a middle aluminum disc which mounts the second PCB and acts as a heatsink for the driver and a bottom plastic cover because.. well, wifi is screened pretty effectively by aluminum otherwise. Power entry is through the black connector and a little button is on the front. Mounting the stirrer can be with either standard 60mm clamps or with an M10 thread in the back (also fairly standard among retort supplies).

The PCB is shaped... weirdly. This is mostly for testing purposes. Really, it's 3 PCBs all stitched together in such a way that they can be easily broken apart. On the left is the inductor mounting PCB, on the right is the main electronics (dominated by the big wi-fi module, some power management and the driver chip) and in the bottom is a USB to serial PCB to program the ESP8266. This can all be assembled in one go and tested, and then before mounting it in the enclosure it's broken apart into the separate PCBs.

This design is NOT meant to be any kind of idea of how the final product would look or work. I mainly just wanted to have some kind of baseline as to how big everything would be and how the costing would look. So as of now, these are just pretty pictures to look at.

[Edited on 8-1-2017 by mux]

Sulaiman - 8-1-2017 at 09:37

As a repair enginner, the first things I saw were;

. electrolytic capacitors in a hot unventilated enclosure, at least use solid/polymer capacitors

. no way to replace the capacitors - could you use lands that protrude beyond the capacitor leads ... then they will be easier to replace.
(to get old ones off, grip with pliers etc. and twist cw/ccw/cw/ccw... starting gently, the body will come away without damaging the tracks, allowing de-soldering of the remaining bits of leads)

the same applies to the lands for the indutors.

[Edited on 8-1-2017 by Sulaiman]

mux - 8-1-2017 at 09:59

Note that this is just a stirrer, this is not designed to become warm at all. The electrolytics are designed to be safe life.

Most round lands you see are actually through holes, but my PCB rendering program for some reason doesn't render them as such. So yeah, the inductors and caps are in fact through hole components :)

Sulaiman - 8-1-2017 at 11:02

Your magnetic circuits are very open,
if the pcb ends of the electromagnets were magnetically joined, even with moderate airgaps (e.g. 1.6mm)
your electrical power dissipation in the coils would be 1/4 of that if not magnetically connected.

if air gap = x2, then A.t = x2 so I2.R = x4

and the aluminium surrounding the coils/inductors/electromagnets will act like shorted-turns.

[Edited on 8-1-2017 by Sulaiman]

mux - 8-1-2017 at 13:13

That is an excellent point. Hmm... Obviously, there will be a sizable air gap to the stir bar anyway, but the way it is now, it's a woefully inadequate magnetic circuit. Can't believe I haven't thought about that in this design. I'm stupid sometimes.

The aluminum won't really do much, all of the interaction is purely magnetic (i.e. eddy currents). Because of the low permittivity, the dissipative effects will not be too distinct from air. We're not dealing with sufficiently high-frequency magnetic fields here, worst case at 3krpm it's doing 50Hz. Here, I'll show you:

https://www.youtube.com/watch?v=m9rnNJsbNa0

wg48 - 8-1-2017 at 16:30

Quote: Originally posted by mux  
That is an excellent point. Hmm... Obviously, there will be a sizable air gap to the stir bar anyway, but the way it is now, it's a woefully inadequate magnetic circuit. Can't believe I haven't thought about that in this design. I'm stupid sometimes.

The aluminum won't really do much, all of the interaction is purely magnetic (i.e. eddy currents). Because of the low permittivity, the dissipative effects will not be too distinct from air. We're not dealing with sufficiently high-frequency magnetic fields here, worst case at 3krpm it's doing 50Hz. Here, I'll show you:

https://www.youtube.com/watch?v=m9rnNJsbNa0


Interesting video. I am curious about the parameters of your set up.

How many turns on the coils and what is the peak current and voltage?

Was the voltage applied to the coils a squarewave and what frequency? It looks a lot slower than 50Hz.

Did you try the same with water in the beaker and perhaps cooking oil? When the distance between the coils and the stir bar is increased (in water or cooking oil) what is the distance when it looses lock and at what freqency?

How thick is your aluminium sheet and by how much does it reduce the maxium height to remain locked at a given frequency?

mux - 9-1-2017 at 02:04

These are just silly prototype experiments. Yes, I'm currently using this to stir my cordial, it works just fine with liquids, can even liberate a stir bar from some loose soluble powder. It's a magnetic stirrer like any other.

I'm currently using an A4983 stepper driver at 200mA through some random inductors, unknown number of turns and magnetic field strength. Its maximum rotation speed is 600rpm atm because of software limitations.

Like any other stirrer, it's only effective pretty close to the stirrer. Go beyond 2-3cm (1in) and it barely stirs anymore. The sheet you see is 2mm. That's about the minimum thickness to avoid the magnet from sticking to the ferrite cores.

mux - 9-1-2017 at 06:02

I've got another question for the seasoned stirrer people around here. I suppose everyone.

Please excuse the lack of appropriate glassware (I seem to have misplaced my big meth-making 5L roundbottom) as well as photography gear. I swear I'm a professional!

https://www.youtube.com/watch?v=2lI-i23f_Gc

In this video I'm inducing spinout/decoupling at 2500rpm. After the stir bar spins out, it pretty much won't ever properly start again. It just vibrates in place. This is with a monstrously big (54mm) stir bar, so that may have something to do with it, but even my normal-sized stir bar won't neatly start spinning if I just immediately start it up at 2500rpm. The speed has to be ramped up slowly, and even then (as you can see a bit further in the video) it sometimes destabilizes briefly.

Now, when ramping up the speed slowly and eventually plateauing, it's stable and will work for hours (well, only tested that once, but it seems rock solid).

I've worked with a grand total of one magnetic stirrer before in my life, and I do remember getting it to work in some glassware used to be annoying to impossible, while it would reliably work in roundbottoms and such. Is this just a fact of life, or do I have serious centering issues?

Sulaiman - 9-1-2017 at 06:32

if you use magnetic stirbars they must rotate synchronously with the rotating magnetic field,
as torque load on the spinbar increases, it will lag the applied field by an ever increasing angle, theoretical maximum 90 degrees, then fly away.

to get it spinning again you need some random chaos to get the spinbar moving close enough to synchronous with the rotating field that it can be 'captured'
or slow down the rotating field to a near stop then ramp up again, but not so far or fast this time :)

that's my understanding so far.


This does raise the potential for very high speed rotating magnetic fields coupling to non-ferrous conductive disks
(ptfe coated aluminium, pure platinum, catalytic disks ...)

EDIT: it just occured to me, when you do experiments again,
put LEDs (e.g. red/green in one package for +/- current) near the respective electromagnets,
which indicate electromagnet voltage, or preferably current (magnetic field proportional to current).
That way a video could be analysed frame-by-frame, or in slo-mo to see if the above is true.
It would look 'cool' with LED magnetic field rotation indicators too :cool:

[Edited on 9-1-2017 by Sulaiman]

JJay - 9-1-2017 at 06:33

Just FYI, a 5L is not a big "meth-making round bottom." I know there's a cultural disconnect here, but a 22L is a big meth-making round bottom. A 5L would probably be used for purifying solvents.

Not to harsh your buzz, but while some means of automating control might be appreciated, you would do well to follow industry standards instead of simply putting your stirrer on the Internet. What if hackers break into my stirrer and cause a fire, etc...
and don't presume to lecture me on Internet security with a false sense of superiority; WiFi in a stirrer is just an unnecessary and undesirable gimmick.

Sulaiman - 9-1-2017 at 07:16

As an inexperienced hobby chemist, one thing that I would like is
to be able to set up an experiment (especially distillation, reflux etc.) and once apparently stable,
be able to leave it un-attended until I get either a 'FINISHED' signal, or an 'ALARM' signal.

The most important ALARM signal, and a minimum requirement,
would be an isolated relay contact that is closed ONLY when the equipment is working 'properly'
In industry this is the E-STOP, EMERGENCY STOP signal line
that can be daisy-chained through all monitored nodes,
if any equipment detects a fault, or has no power, the E-STOP line will not pass the current required to enable operations down the line to continue.
The E-STOP line commonly operates on a 24 Vdc line, (but not neccessarily)
the end of line usually has a 24 V relay (or optocoupler etc.) that operates when all intermediate nodes (equipment with isolated relay contacts, E_STOP switches, breakable wire barriers ....) are operating 'nominally', to produce an enable signal for processes/power to continue.

Each node should monitor the incomming E-STOP signal,
if not present that constitutes a fault so the heater-stirrer should take appropriate action
(e.g. immediate cessation of heating, stirring continued for a time ...)

Ideally I would like the heater-stirrer to detect out-of range temperature and loss of spinbar rotation as faults.

[Edited on 9-1-2017 by Sulaiman]

[Edited on 9-1-2017 by Sulaiman]

mux - 9-1-2017 at 07:16

Sulaiman: Sure, I'll see what i can do w.r.t. the LED idea. That's not a bad idea.

JJay: I'm just joking around ;-) As for the Wi-Fi comment:

Well, you seem fairly determined that you don't want it, so... don't buy this stirrer :P. The reality is that I don't see how people hacking into your stirrer could fuck up anything, especially if it's an encrypted connection and the stirrer has a password protected interface. Let's be real here: anyone determined and close enough to hack your stirrer can just... mess with the knobs on a physical stirrer as well.

Wi-fi connectivity is different from an open connection to the internet. If your local network is compromised to the point where they can mess with your stirrer, they can also mess with your other personal electronics. The stirrer, again, seems to me to be way down the list of things to worry about. I'm aware of the IoT security hazard, but in this case any true hazard introduced by the stirrer is just as much introduced by having WiFi or a cellphone in the first place. The specific comment a bout a stirrer causing a fire is a design defect by itself, that should never happen under any circumstance. This seems like the weakest possible argument you can have against wifi connectivity. If you have explosive, flammable or delicate chemicals in your lab, it's already in a fumehood with proper fire protocol in place. It's not like you're keeping this stirrer under your bed. Same goes for a wi-fi connected hot plate.

On the other hand, the convenience granted by this type of connection is very significant, at least in the use case I'm designing it for. I don't do chemistry fulltime in a lab. My 'lab' is a small corner of my garage or electronics workspace and my experiments can take hours, even overnight to finish. I'm not babysitting that. Robots should do that.

And in order to program that robot, there should be an interface that requires the least possible additional effort from the user and creator. Physical interfaces (e.g. a little touch screen or graphical interface) are complex, cumbersome and expensive. A web interface offers practically infinite compatibility, can be easily (re)designed and updated to fit any use case and can have much higher information content and programmability. THAT is why I'm convinced wifi to be core to this design.

JJay - 9-1-2017 at 07:31

Nah, you just think idiots will buy your stirrer if it has WiFi. We both know it would be better if it could be part of an integrated solution that could work in places with heavy electromagnetic interference such as most any heavy industrial environment; even USB would be vastly superior to WiFI. If you want WiFi, you could connect it to a controller that has WiFi. You're incredibly confused here; you're actually suggesting that allowing hackers to control your machinery and industrial processes without authorization is no worse than letting them read your cell phone contacts.

That's not the point though; the point is, WiFi is inferior. I could use a vanilla USB interface connected to a controller that has WiFi if I wanted it.

People, can we really take this guy seriously? I think he's a fraud.

Sulaiman - 9-1-2017 at 07:46

Wi-Fi is neccessary because inside the stirbar thermal and spatial sensors, and LED bars that display the temperature and spin speed whilst spinning.
like the propellor-clock https://www.youtube.com/results?search_query=propellor+clock

:cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool::cool:

I want one

P.S. as far as I can tell, a huge market for heater-stirrers is the 'vaping' community ... money no object, iphone interface mandatory.
as are elegant lines and sexy displays, precise speed and temperature not important.

[Edited on 9-1-2017 by Sulaiman]

JJay - 9-1-2017 at 07:57

Pay attention to Sulaiman; he's clearly a smart guy who knows what he's talking about.

mux - 9-1-2017 at 09:28

Quote: Originally posted by JJay  
Nah, you just think idiots will buy your stirrer if it has WiFi. We both know it would be better if it could be part of an integrated solution that could work in places with heavy electromagnetic interference such as most any heavy industrial environment; even USB would be vastly superior to WiFI. If you want WiFi, you could connect it to a controller that has WiFi. You're incredibly confused here; you're actually suggesting that allowing hackers to control your machinery and industrial processes without authorization is no worse than letting them read your cell phone contacts.

That's not the point though; the point is, WiFi is inferior. I could use a vanilla USB interface connected to a controller that has WiFi if I wanted it.

People, can we really take this guy seriously? I think he's a fraud.


I really don't follow you here anymore. Do you care to elaborate a bit more on this post? What are your exact issues?

Melgar - 9-1-2017 at 10:23

Ok, apologies if any of my criticisms weren't particularly constructive, I'll try keep the tone as constructive as possible from now on, if that helps. I'm just happy as long as you've learned some new things about motors, especially if you plan to use a stepper controller to drive it. As far as the possibility of falling back to using a motor, I did a quick search, and there are a lot of brushless DC fan motors being sold for a few bucks on eBay and Alibaba. These are basically the same type of motor as stepper motors, but typically with an average of six steps per rotation, compared to a lot more (usually) for stepper motors. Their controllers also tend to be a lot simpler, at the expense of not being able to fine-tune position as well as with a stepper, but for a stirrer, that'd seem to be an acceptable trade-off. The fan motors usually have the controller built-in, and many can be controlled on a third wire, with either a PWM signal or an analog DC control signal. Fan motors seem ideal, because they're designed to have a long life, and their speed range would be approximately what you'd want at the power that you'd want. For example:

https://www.alibaba.com/product-detail/12V-DC-BLDC-electric-...

Another:

https://www.alibaba.com/product-detail/Dc-Brushless-Fan-Moto...

As far as magnets, I got ten alnico 5 bar magnets (1.5 inches long, 3/8 inch square cross section) for like $11.50 on eBay, and the manufacturer sent me a brochure with my order, stating their manufacturing capabilities, so it wasn't just some underpriced surplus. Considering the size of the magnet in my last stirrer, I was worried these ones would be too small, but they're plenty powerful, and have no trouble turning a stir bar from 1-2 inches away. I also found a 600W hot plate at Family Dollar for $10, that's about 10 inches square, and about an inch and a half deep. It's also teflon-lined aluminum. I figured for baths, I could melt wax in it and use that for my fluid, then I don't have to worry about spilling the bath after it's cooled off. My plan was to basically control and monitor everything with an Arduino, that way the electronics are cheap and I can program in actions for any failure modes I can anticipate.

My project idea was conceived independently of yours, and the stirrer is kind of a minor part of it. Right now I'm just using a motor I pulled out of an inkjet printer and haven't had any problems with it. If it does break, I have three more.

mux - 9-2-2017 at 02:37

Well, I was meaning to post almost 2 weeks ago with the completed PCB, but I decided to wait until I had a fully assembled unit. Boy, was that a mistake. Anyway, the money shot:



That photo is dark for a reason. My milling machine really doesn't like the 4mm mill bit, so it took me a few tries before I at least halfway succeeded at milling the aluminum parts. Had to file and sand the sides by hand, which doesn't look good.

It works though, and I'll post a more comprehensive update including a video on my youtube probably before the end of the weekend.

mux - 16-2-2017 at 00:16

Royally late, but here's the video:

https://youtu.be/Kn-eLq7-Fbw

Sulaiman - 16-2-2017 at 04:20

Impressive video,
I especially liked the multi-pcb design being testable before cutting.
Most of all, YOU DID IT ! well done.

Although it works better than I expected,
I am certain that your 'electromagnets' are far from optimal,
try a simple experiment;
break the top ferrite ends off of the inductors, keeping the bottom ends and central core intact,
see if the stirring 'strength' increases.
I'm not suggesting this as a working configuration, just a test.
If you have spare inductors :D

Watching your progress with admiration ......

P.S. since the majority of the flux path is air, adding tiny gaps in ferrite makes little difference,
i.e. you can superglue together any ferrite core configuration that you fancy, or just glue together any broken pieces.
(except the core within the coil, best to be continuous rather than contiguous)

[Edited on 16-2-2017 by Sulaiman]