| Pages:
1
2 |
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
Cool electronic devices you can make for thr lab but are uncommon.
So, other then
-hotplates/stirrers
-tube furnaces
-kiln furnaces
What are some fun electrochemical design projects?
I thought
1) mechanical stirrer
2) ball mill. Interesting project due to the fact that you need to change velocity depending on container size
3) aspirator pump: mechanical pump to cycle fluid to an aspirator, vacuum gauge monitoring connected to a very damp pid, cooling system to keep the
water at 4c or so
4)coolant cycle pump (common but a good one would use peltiers in a pid to do 0 01c resolution at 1kW cooling capacity. Expect approx 2k cost
5) melting point tester. Started towork on this may detail project
|
|
|
SWIM
National Hazard
Posts: 970
Registered: 3-9-2017
Member Is Offline
|
|
Aspirator pumps and ball mills are actually pretty common on here.
Mechanical stirrers and condenser coolers less so, but there's been work on those recently by Jjay and Melgar respectively, and several others.
Might be good to review the threads on those items and see what others have tried, and what problems still need to be whipped.
I would be interested to hear of an example where a 0.01C accuracy on the cooling water flow would be useful.
|
|
|
Ubya
International Hazard
Posts: 1232
Registered: 23-11-2017
Location: Rome-Italy
Member Is Offline
Mood: I'm a maddo scientisto!!!
|
|
1kW of cooling power using peltiers, that will be fun (considering inefficiencies $ 2k thrown away)
---------------------------------------------------------------------
feel free to correct my grammar, or any mistakes i make
---------------------------------------------------------------------
|
|
|
DavidJR
National Hazard
Posts: 908
Registered: 1-1-2018
Location: Scotland
Member Is Offline
Mood: Tired
|
|
Melting point apparatus would be good.
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
Quote: Originally posted by Ubya  | | 1kW of cooling power using peltiers, that will be fun (considering inefficiencies $ 2k thrown away) |
It's actually not that bad. 2k$ would be really high end with swagelok parts and stuff. Peltiers have a bad rep but they are actually really nice for
stability, and you would be surprised how usable and convenient something in the 1kW region is in the laboratory. Once you use a machine that you can
adjust the PID loops of to your process you will probably not wanna go back to clunky compressors lol, or you will want a dual stage system with the
compressors cooling the peltiers.
I was thinking that high temperature stability of the coolant water would mean a better vacuum stability for the pump at its maximum low pressure.
[Edited on 10-5-2018 by coppercone]
|
|
|
VSEPR_VOID
National Hazard
Posts: 719
Registered: 1-9-2017
Member Is Offline
Mood: Fullerenes
|
|
Centrifuge would be a nice touch and drying oven/desiccator is useful
Within cells interlinked
Within cells interlinked
Within cells interlinked
|
|
|
Sulaiman
International Hazard
Posts: 3558
Registered: 8-2-2015
Location: 3rd rock from the sun
Member Is Offline
|
|
| Quote: Originally posted by coppercone |
1) mechanical stirrer
2) ball mill. Interesting project due to the fact that you need to change velocity depending on container size
3) aspirator pump: mechanical pump to cycle fluid to an aspirator, vacuum gauge monitoring connected to a very damp pid, cooling system to keep the
water at 4c or so
4)coolant cycle pump (common but a good one would use peltiers in a pid to do 0 01c resolution at 1kW cooling capacity. Expect approx 2k cost
5) melting point tester. Started towork on this may detail project
|
The electronics for 1,2,3 & 4 are mostly motor speed controllers,
readily available as cheap ready made modules via eBay etc.
#5 can be done with a phase controller (triac light dimmer) module + thermometer
or a REX-C100 with thermocouple.
so
the 'hard' parts for all of the suggestions are actually the mechanical parts.
Each of the suggestions is already commercially available,
so the main motivation for d.i.y. is cost
- using whatever is 'in stock' or available regionally will be used.
So each of us will come up with a slightly different implementation.
The important thing is to give enough general information for others to recreate their own version based on their availability of parts.
=============================================
I made a small dual TEC-based water cooler
- on its own suitable for 50 to 100 Watts cooling power, nearer to 50 than 100 
A useful device by itself for small scale (10/14) distillations etc.
With the addition of a hot water 'radiator' suitable for larger scale (24/29).
IF I wanted more cooling power
(1 kW is quite a large still ... 1.8l water or 4.3l ether per hour)
I would first investigate passive coolng - large air-cooled surfaces,
if active cooling is required the a commercial/domestic water chiller or refrigeration unit would be MUCH cheaper and efficient.
If you search SM you will find attempts at all of your ideas.
EDIT: I re-read my post and it seems rather negative/de-motivating - sorry.
Each of us that attempts any of the above, and reports their results, adds to a general body of knowledge.
Even failures - preferably with reasons - is useful knowledge.
So, ignore the tone of my grumpy mood post and go full steam ahead.
BUT, please see what others have done first
[Edited on 11-5-2018 by Sulaiman]
CAUTION : Hobby Chemist, not Professional or even Amateur
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
A good melting point tester will use a four wire rtd and have very well controlled temperature change rates so the rtd doesnot permamently drift. The
electronics also have to be low noise and temperature stable.
You will find the sensor and sensor interface electronics (even supposedly good ones like from honeywell) have fairly shitty specifications.
I would shoot for stability of better then 100 parts per million per degree celcius, 100ppm repeatability and 500ppm accuracy. (calibration is
difficult and would actually require many samples from test sources to be linear hence the loose 500ppm specification.
You would actually want two melting point testers, i would limit coil rtds to controlled ramp rates at 150c or so and a seperate tester tester meant
for higher temperature operation that uses a rtd which takes a beating and is less accurate or a less accurate thermocouple.
There is also the issue of oxidation and the thermal emissivity changing with temp. The higher temp one would desire an argon blanket. I experimented
with aluminum that had a heating cartiridge in it with a groove for the heater and i was not impressed.
I guess the nice thing about thermal electronics is that you can get away with relative butchery on the power electronics side (i.e. a triac) so long
you have good control and sensor electronics.
Good peltier control is kinda difficult too because you need to ramp them slowly and provide a somewhat filtered signal to prevent thermal stress
damage .
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
And if you want a defacto way to catagorize a tec, you basically make a heat load, which is a well insulated coldplate that has either heater strips
stuck onto it or heating cartriges.. Then you cyxle water between the tec chiller and the heat load.
You put a fixed wattage on the heat load and measure the temperature of whats dissipating your tec power (probobly air but you can use a water or
freon heat echanger) and the temperature of your water.
Then you say the tec system is capable of some particular temperature differential at some particular flow with some number of watts heating the
system.
This is how the high end ones are catagorized. You want heavy insulation on the whole system and decent amount of turbulance in your two heat
exchangers.
Then you can do further catagorizations like changing the heatload power to see the responce characteristic.
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
I also think that using precisely chilled water would benefit fractioning with active cooling, but maybe the scale required for this to be useful is
too large, to keep the number of theoretical plates consistant during the whole process.
I dont think it would be too useful for distilation condensation control though.. Am i wrong?
Of course you could make a device that attaches to the distiller to eliminate the need for water flow, or to control reflux.. But i feel that a useful
water less system is toodifficult for a research chemist because we have a variety of different condesners we like touse for reflux depending on whats
going on.. I think this is more suited for a known process.
If you have excellent lab scafolding to support a heavy object in odd positions then having some standard reflux column might be an aleight option..
You can even use the same sensors used to detect water on car windsheilds to maintain the position of your reflux line by gently varying the cooling
power..
And maybe you can make thermal transfer bushing sets to accomodate different glassware sizes..
But, i do have trouble imagining a simple device that can matcha reflux condenser without coolant moving through the interior coil. It would need to
be long. Dont feel something useful is practical to make by yourself without serious effort unless yku like small scales.
Biggest benefit is that you dont have to worry about a water leam destroying/stressing your setup or causing an explosion, erruption etc.
But it would be suceptible to corrosion and damage from leaks and when you disassemble glassware with solvents in it.
Guess you can calculate the interior area of a fredrich column or reflux condesner and guesstimate how long a linear condesner cooled by tecs you
would need for the same cooling area. And thermally coupling tecs mounted to a heatsink to a piece of glass without a prohibitively thick interface
medium like grease is a serious pain in the ass. You can use special low melting point solders to mate certain metals with glass but i believe you
would get stress build up on a decent surface area that would be close or exceede the capacity of glass.
I would like to design these kinds of things towards 24/40 setup sizes, i feel that they are practical and fun.
I think you enter a whole different ball game with microchemistry like 14/20 though when there are not alot of watts to deal with.
Another device i had a bone to pick with for a long time is a kugelrohr. I feel that they are ridiculosuly overpriced.
[Edited on 11-5-2018 by coppercone]
[Edited on 11-5-2018 by coppercone]
|
|
|
Twospoons
International Hazard
Posts: 1282
Registered: 26-7-2004
Location: Middle Earth
Member Is Offline
Mood: A trace of hope...
|
|
| Quote: Originally posted by coppercone | | Quote: Originally posted by Ubya | | 1kW of cooling power using peltiers, that will be fun (considering inefficiencies $ 2k thrown away) |
It's actually not that bad. 2k$ would be really high end with swagelok parts and stuff. Peltiers have a bad rep but they are actually really nice for
stability, and you would be surprised how usable and convenient something in the 1kW region is in the laboratory.
[Edited on 10-5-2018 by coppercone] |
There are other ways to achieve accurate control at those power levels, using a compressor driven heat pump. Look up "hot gas bypass". This is a
system which uses a bypass valve to feed some of the hot condenser side gas into the evaporator coil. I built a controller for a system like this for
refrigerated shipping containers - it would happily hold a container at 1C with 0.1C accuracy using a 7kW refrigeration compressor capable of -25C.
The other method is to have a reservoir of very cold fluid (water or brine), chilled by compressor, and using a mixing valve to control the cooling
loop temperature.
These methods seem far more practical than setting up 1kW of peltiers with their terrible COP.
Helicopter: "helico" -> spiral, "pter" -> with wings
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
if you use water cooling and a cooling tower a 1kW chiller is a rackmount piece of equipment that fits in a 19 inch rack and is about 5 inches tall
and 1.5 feet deep. this is with the pump, power supply, control electronics, internal reservoir tank, plumbing etc
Otherwise,. for air cooling, you are looking at something like two 10x10x4 inch heat sinks with decent fans (might want to break it up into 2 separate
modules so you don't get the pressure drop along the fins, plus the before mentioned items
hs -> peltier -> heat exchanger (with plumbing for coolant loop)-> peltier -> hs
with the sideway view being like
||||||||||||||||||||||
||||||||||||||||||||||
||||||||||||||||||||||
==========
||||||||||||||||||||||
||||||||||||||||||||||
||||||||||||||||||||||
The idea is that heat is being removed from the top and bottom to the air, or to liquid cooled plates (rather then heat sinks).. you can push air
through the whole thing if you put it in a box with the channels formed by the heat sink fins facing the face of the fan. Otherwise you need two fans
and alot of room around the thing (Unless you need real uniformity, which is kinda pointless given that your cooling a fluid stream, its retarded to
do it this way, just because mounting it mechanically would be ridiculous.. would be really complicated vent arrangement. This method is extremely
compact. With the method I described, you can also put two fans, one sucking one blowing, on both ends of the channel, to account for pressure drop
without needing super fans or complicated duct system/.
You don't want to design something that has air intake on the back and the side and blows out air both sides unless its really necessary.......
If your fluid stream going through the main exchanger is really slow, the temperature differential MIGHT have some kind of effect.
These can do 0.01c control easily. 0.001c with very good insulation. And if you make it with the right parts, the coolant cycle pump and heat sink
fans will ware out several times before there is any ware on the main heat exchange unit.
How fast is the response of a hot gas bypass unit? Like what kind of thermal time constants can it deal with? And how is the longevity of the mixing
valve / etc with repeated temperature oscillations (like from temperature cycling).. thats where these guys excel at. Shipping container is a very
static load..
but I don't want this thread to get derailed by peltiers. I will be honest though I don't really like mechanical compressors they just fucking annoy
me for whatever reason. And they sound like shit. I would like to hear about them though, I was always curious about higher end laboratory
refrigeration systems.
[Edited on 11-5-2018 by coppercone]
[Edited on 11-5-2018 by coppercone]
|
|
|
diddi
National Hazard
Posts: 723
Registered: 23-9-2014
Location: Victoria, Australia
Member Is Offline
Mood: Fluorescent
|
|
i am tinkering with a recirculating peltier cooling system atm. just need a couple of days on the mill to get it how i want. if it works i will post
more info
Beginning construction of periodic table display
|
|
|
wg48
National Hazard
Posts: 821
Registered: 21-11-2015
Member Is Offline
Mood: No Mood
|
|
Below is what you need for a 1kW @ 30C delta. You would also need a 2kW DC supply and somthing to put it all in. It would keep a small lab cosy in
winter. A 100W system would be more practical.
|
|
|
Panache
International Hazard
Posts: 1290
Registered: 18-10-2007
Member Is Offline
Mood: Instead of being my deliverance, she had a resemblance to a Kat named Frankenstein
|
|
For a water chiller just pickup a water cooler from the side of the road and jerry it into a recirculating water chiller. i have a few of these, a
small pump easily fits inside them as they have excess space in the middle so they are tall enough. i also converted one into a water bowl for my dog,
keeps her water at 4C, which likely is completely unesseccaery as she is happy as hell to use the toilet but i choose to believe in summer she feels
extra love because of my efforts.
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
You want two heatsinks, to make a sandwich on your heat exchanger. You won't get terrible better performance using an all copper expensive to braze
heatsink...
That heat exchanger you listed looks kinda odd, you want something like this
http://www.lytron.com/Cold-Plates
https://www.amstechnologies-webshop.com/cp15g05-aluminum-col...
For the power supplies, its a good idea to find arrangements of your peltiers that can use something like a telecom 48V power supply made by lambda,
meanwell, etc. Then you control it with a H bridge.
If you want 2kW of power, and you have two sides, you can independantly power the two different peltier sides, if you want redundancy incase of a
short circuit.
Using air for above 500W is a little crazy. I have not seen it done. Traditionally this seems to be done by using water to cool the peltiers and
directing said water to a cooling tower or large radiator outside the building (this is typically part of facilties infrastructure for certain labs).
http://www.meanwell.com/meanwell_products.html
Those will be OK since you don't have some kinda crazy reactive load or whatever. Then you just chop them with a simple H bridge circuit.
[Edited on 12-5-2018 by coppercone]
[Edited on 12-5-2018 by coppercone]
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
| Quote: Originally posted by Panache | | For a water chiller just pickup a water cooler from the side of the road and jerry it into a recirculating water chiller. i have a few of these, a
small pump easily fits inside them as they have excess space in the middle so they are tall enough. i also converted one into a water bowl for my dog,
keeps her water at 4C, which likely is completely unesseccaery as she is happy as hell to use the toilet but i choose to believe in summer she feels
extra love because of my efforts. |
What kind of heat exchanger do you use for a corrosive system?
I have a nice lytron one, that I can hook up into the freon line of a compressor, but I doubt it would like cooling whatever corrosive trash is sucked
in through an aspirator, so I would need to make a secondary loop regardless of what the primary loop is cooled by, and a secondary recirculation
pump.
Glass coils?
How hard is it to wind a glass coil without a glass lathe? I have oxy or air / hydrogen or acetylene torches
Easiest would be to use stainless tubing
https://www.grainger.com/product/3ACU2?cm_mmc=PPC:+Google+PL...
I have my doubts about using copper. With all the effort that goes into building this thing, I would hate to have to look at nasty gross ass
chemically damaged copper. I would like it to be spotless like a nuclear reactor even if its in a old 5 gallon cat litter bucket.
Are there any common lab particulars you can think of that stainless steel 316 will react with?
Br2, Cl2, HCl, HNO3 come to mind quickly.
Also, I have a nice motor drive that uses 90V and has a field winding. It's actually meant for pumps.
Can you suggest a pump that would work well for a lab aspirator ?(my focus is solid state devices and precision instrumentation I don't know much
about driving pumps, specifying them, motor drivers, etc)
I would like to setup a loop to control the vacuum via feedback of vacuum pressure.
I would also like to know how much heat in a pump is generated by the pump assembly, and how much by the electric motor, and if the heat conducted to
the fluid can be reduced by separating the two by a long shaft or belt drive.
https://www.ebay.com/itm/316-Stainless-Steel-Round-Tube-OD-1...
[Edited on 12-5-2018 by coppercone]
|
|
|
Sulaiman
International Hazard
Posts: 3558
Registered: 8-2-2015
Location: 3rd rock from the sun
Member Is Offline
|
|
I hope that you do make a cost effective water cooler,
and contribute some more information,
there have been a few TEC-based discussions,
latest: http://www.sciencemadness.org/talk/viewthread.php?tid=80597
and I'm having a dabble myself: http://www.sciencemadness.org/talk/viewthread.php?tid=64759&...
CAUTION : Hobby Chemist, not Professional or even Amateur
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
another biggie for those peltiers is to provide proper torque to mate them with the heat sink for the thermal compound to work well.
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
But for 316 stainless, I looked at the list:
Potential problems include:
Aluminum Fluoride D
Aluminum Hydroxide
Ammonium Phosphate, Dibasic
Aluminum Chloride
Aluminum Salts
Ammonium Fluoride
Aniline Hydrochloride
Antimony Trichloride
Aqua Regia (80% HCL, 20% HNO
Aromatic Hydrocarbons ?????????
Bleach (chlorine, 12.5%)
Bromine
Calcium Hypochlorite, 30%
Chloric Acid
Chlorine Dioxide
Chlorine Water
Chlorine, anhydrous liquid
Clorox® bleach (sodium hypochlorite, 5.25%)
Copper Chloride
Copper Fluoborate
Ethyl Sulfate
Ferric Chloride
Ferrous Chloride
Fluorosilicic Acid
Hydrobromic Acid (including dilute)
Hydrofluoric Acid (Unsure of minimum concentration)
Hydrogen Cyanide
Hydrogen Sulfide
Hypochlorous Acid
Iodine
Iodine Solution
Lead Sulfamate
Magnesium Chloride
Melamine
Mercuric Chloride, dilute
HCl
Nickle Chloride
Ginger Oil (?????)
Concentrated Oxyalic Acid
Phosgene
Phosphoric Acid Concentrated
Phosphorus Oxychloride
Resorcinal
Silver Bromide
Silver Chloride
Sodium Benzoate
Sodium Bisulfate, 10%
Sodium Bromide
Sodium Cyanate
Sodium Sulfide
Stannic Chloride
Stannic Chloride
Sulfur Chloride
Sulfuryl Chloride
Tartaric Acid
Tetraphosphoric Acid
Tin Ammonium Chloride
Tin Salts
Trichloroacetic Acid
But I mean, so long you use a trap condenser,. and dispose of the old water after your done, can the stuff listed as corrosive harm the steel in the
concentrations that would form from lab scale vacuum use?
|
|
|
Panache
International Hazard
Posts: 1290
Registered: 18-10-2007
Member Is Offline
Mood: Instead of being my deliverance, she had a resemblance to a Kat named Frankenstein
|
|
Coppercone ive pm'ed you however if you are wanting to loop back for pressure regulation I've had vac pumps (graphite vane-types) controlled via a
VSD, this however is for systems that are producing some non-condensable gases, otherwise you need a finely controllable atmospheric bleed still
somewhere between system and pump, like a needle valve.
Once running you could program the vsd but obviously its not going to adjust your needle valve.
Standard on all vacuum systems is a cold trap, basically a large bore U-tube sitting in either liquid nitrogen or dry ice, but no-one ever uses these
in a lab or home environment. Cheaper using wash bottles specific to your needs (ie caustic for acid non- condensable etc).
And im surprised the list is so short for 316.
btw that lytron site is quite good
[Edited on 13-5-2018 by Panache]
[Edited on 13-5-2018 by Panache]
[Edited on 13-5-2018 by Panache]
|
|
|
Sulaiman
International Hazard
Posts: 3558
Registered: 8-2-2015
Location: 3rd rock from the sun
Member Is Offline
|
|
| Quote: Originally posted by coppercone | another biggie for those peltiers is to provide proper torque to mate them with the heat sink for the thermal compound to work well.
|
I just glued heatsink:TEC:WaterBlock:TEC:Heatsink with araldite epoxy resin 
Because the temperature changes and differentials of temperature at each boundary are quite small.
If the araldite fails I'll report back here a.s.a.p. but it seems OK so far.
============================================
The most obvious heat exchangers:
https://www.ebay.co.uk/itm/230mm-Graham-Condenser-24-29-Join...
https://www.ebay.co.uk/itm/410mm-24-40-Graham-Condenser-Join...
and I got one of these from my secret santa, it is efficient,
https://www.ebay.co.uk/itm/200mm-24-29-Glass-Reflux-Graham-C...
then route the hot water output via an old/used or cheap central heating radiator to remove most of the heat,
for higher efficiency add axial fans to force air through the radiator.
Then use your active cooling, reservoir and pump.
Why pump 100's Watts via TECs when only 10's Watts are required ?
Not very compact but cheap, reliable and efficient.
ANY large surface area water container can be used to passively cool the hot water from the glassware; copper tubing, automobile radiator, air-con
radiators etc.
The radiator can be distant from your glassware, or, outside your lab.
EDIT: I have only had one lash-up test during winter so far,
I hope to start measurements on this arrangement in a few months from now during summer, because as ambient temperature increases, the radiator
becomes less effective... T.B.D.
[Edited on 13-5-2018 by Sulaiman]
CAUTION : Hobby Chemist, not Professional or even Amateur
|
|
|
Panache
International Hazard
Posts: 1290
Registered: 18-10-2007
Member Is Offline
Mood: Instead of being my deliverance, she had a resemblance to a Kat named Frankenstein
|
|
Graham or linear condensers are best for vacuum distillations because they provide no back pressure (actually never really thought about it but in
vacuum systems technically would this be referred to as forward pressure?)), however they are horribly inefficient and are susceptible to 'walling';
where a fluid flow down the centre of the tube escapes heat exchange. Adding something that disrupts this flow is the same as using a nonlinear
condensor.
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
| Quote: Originally posted by Sulaiman | | Quote: Originally posted by coppercone | another biggie for those peltiers is to provide proper torque to mate them with the heat sink for the thermal compound to work well.
|
I just glued heatsink:TEC:WaterBlock:TEC:Heatsink with araldite epoxy resin
Because the temperature changes and differentials of temperature at each boundary are quite small.
If the araldite fails I'll report back here a.s.a.p. but it seems OK so far.
============================================
The most obvious heat exchangers:
https://www.ebay.co.uk/itm/230mm-Graham-Condenser-24-29-Join...
https://www.ebay.co.uk/itm/410mm-24-40-Graham-Condenser-Join...
and I got one of these from my secret santa, it is efficient,
https://www.ebay.co.uk/itm/200mm-24-29-Glass-Reflux-Graham-C...
then route the hot water output via an old/used or cheap central heating radiator to remove most of the heat,
for higher efficiency add axial fans to force air through the radiator.
Then use your active cooling, reservoir and pump.
Why pump 100's Watts via TECs when only 10's Watts are required ?
Not very compact but cheap, reliable and efficient.
ANY large surface area water container can be used to passively cool the hot water from the glassware; copper tubing, automobile radiator, air-con
radiators etc.
The radiator can be distant from your glassware, or, outside your lab.
EDIT: I have only had one lash-up test during winter so far,
I hope to start measurements on this arrangement in a few months from now during summer, because as ambient temperature increases, the radiator
becomes less effective... T.B.D.
[Edited on 13-5-2018 by Sulaiman] |
The problem with epoxy is that the thermocoefficent of expansion of the aluminum or other heat sink material is different then the ceramic of the tile
so it turns into a reliability problem with cracked tiles and open circuits. It may take a while.
You also want to cycle the temperature slowly, its gonna depend on thermal time constants, but I don't remember what is considered reliable rate of
change.
[Edited on 13-5-2018 by coppercone]
|
|
|
DavidJR
National Hazard
Posts: 908
Registered: 1-1-2018
Location: Scotland
Member Is Offline
Mood: Tired
|
|
| Quote: Originally posted by coppercone |
The problem with epoxy is that the thermocoefficent of expansion of the aluminum or other heat sink material is different then the ceramic of the tile
so it turns into a reliability problem with cracked tiles and open circuits. It may take a while.
|
Indeed - you can get special thermally conductive adhesives which are designed for sticking heatsinks onto chips etc, which I think are silicone based
as they feel rubbery when cured. That would allow for a little bit of movement so I would expect that to be more reliable than epoxy resin. Also,
obviously araldite isn't specifically designed to be thermally conductive, so thermal performance is probably worse, but considering how thin the
layer is, that may not be a problem.
|
|
|
| Pages:
1
2 |
![[*]](images/xpblue/default_icon.gif){kind=icon}
