Sciencemadness Discussion Board

Pulse Discharge Circuit Help

Centimeter - 13-4-2009 at 12:45

I am building a pulse discharge circuit and I was hoping to get some advice on designing a nice safe and effective circuit. I am but a chemist and this circuit is for use in a reaction cell, so please understand that I am not an expert in high voltage high capacitance circuit design. As such, any help you can give me in laymen’s would be much appreciated.

I have an internally rectified transformer with an input of 115V-60Hz and an output of 9000 VDC/146 MICROAMP. Here is a spec sheet of the transformer.

http://6130.iso-group.com/NSN/6130-00-217-7716

I have twenty four 2KVDC 1uF mica paper pulse discharge capacitors with max peak currents of 500amps. The capacitors are Cornell Dubilier KV* series and they were salvaged from surplus exploding bridge-wire blasting machines. I will be combining these in a variety of ways to modulate voltage and capacitance.

Since the transformer is internally rectified, do I need to incorporate any more diodes to prevent the capacitors from discharging into the transformer? If so, what kind of diode would be best for my purposes?

I need to be able to make rather accurate calculations (5-10%) of what voltage and amperage I’m discharging. Any recommendations on how to best do this taking into account that I will be using a “hit the two electrodes together as fast as I can” switch? Or perhaps someone can recommend an equally cheap yet more efficient switching method?

Finally, does anyone have a recommendation for a cheap DIY method of varying the voltage of the transformer?


[Edited on 13-4-2009 by Centimeter]

woelen - 13-4-2009 at 13:22

Varying the voltage of the transformer is not that difficult. Cheapest is to take a series connection of normal low-voltage transformers, or a transformer with many side-taps at which different AC-voltages can be connected. You then can connect your HV-transformer to those different AC-voltages and this allows you to obtain many different voltages. Be sure to NEVER connect it to more than 115 V AC.

If you are willing to spend some more money, then you could take a variac transformer. These come in two flavors, the so-called "auto transformers", which only have one voltage output, which is not galvanically disconnected from the mains power supply and a true transformer, which is galvanically separated from the mains power and which is more safe to work with (but also much more expensive). Variacs are expensive devices anyway.

If you build up a circuit with a series connection of your capacitors, then do not forget to add parallel resistors to these capacitors. These resistors serve two purposes. They are bleeder resistors, which slowly discharge the device when it is powered off (and may safe your life!!!). They also serve the purpose of equalizing the voltage over all capacitors. Without bleeder resistors you may have very uneven voltage distribution over your capacitors. Imagine that one of them has a small but not neglectible leak current, while the others don't have that (or have much lower leakage). Then slowly, that leaking capacitor will have only a very low voltage across it, while the other capacitors see more voltage and this may lead to overstressing the other capacitors. With your 145 uA charge current, it would be good to have 500 MOhm to 1 GOhm bleeder resistors in parallel to your capacitors, giving a 1.5 .. 3 uA leakage current and a discharge time of a few tens of minutes. Use 1.5 kV per capacitor, to give you some safety headroom.

Also keep in mind that the kind of transformer you use usually supplies higher voltage at zero load. This may be dangerous, so always attach a dummy load to it, which draws around 10 uA (1 GOhm is suitable) current.

Regard your HV-transformer, together with its dummy load as a complete system and I would like to advice another HV-diode (or better: set of HV-diodes) as a rectifier, just for added safety. I do not know the transformer and its properties, so just add such a diode, the cost should not be a problem at all.

All required components can be purchased on eBay. High voltage resistors must really be rated for the voltage they are used for. Crappy or low-rated resistors may lead to discharge around the resistor (looks spectacular, but is not what you like :o ).

Also keep in mind that at voltages of 10 kV or so you will have quite a lot of corona discharge, which may be very annoying. It leads to loss of power and formation of ozone. You can hear the hissing and soft crackling around the circuitry. Always use rounded solderings, never have sharp needle-like pins anywhere in the circuit. Sharp needle-like pins lead to excessive corona discharge.

And finally a word of waring. You write you are not experienced in HV-electronics. Keep in mind that a SINGLE error may be fatal and may kill you are may make you a handicapped man for the rest of your life. The goddess of chemistry can bite but the god of HV can kill with a single slap. Actually, for me this is the reason that I quit doing HV experiments. My wife does not want it anymore after I had a very nasty shock from a 2 kV neon transformer :o :(. She really is afraid of the stuff, much more so than from all my chemistry-related things, simply because of the fact that a single error (yes, even a single touch at the wrong place) can kill me.

I built a high voltage power supply for my own experiments. Reading the page may help you getting ideas of how you want to solve your problems:

http://woelen.homescience.net/science/physics/misc/larger-ps...



[Edited on 13-4-09 by woelen]

Centimeter - 13-4-2009 at 14:06

Thank you very much for your reply woelen. Although I am not an expert in these kinds of circuits, I am aware of their danger and methods of avoiding such incidents (A very scary one at that! I can totally understand your wife’s bias). I usually conduct all experiments while wearing gloves and sitting on a bucket and I always have my friend who’s an EMT present just in case. I hope that nothing bad happens, but in this case I am willing to risk it—I have a very good experiment in mind! Perhaps someone has some additional suggestions on ways to improve the safety of the experiment?

Would you be able to recommend an exact type of diode (i.e. specifications) to incorporate into my circuit? Assuming that I will be combing the capacitors in series and parallel ranging from voltages of 2KV to 24KV, what kind of resistor would you recommend I purchase? Also, my experiment hinges on the ability to determine the properties of the discharge. What kind of analytical tools would be appropriate for doing this?

Jor - 13-4-2009 at 14:17

Wouldn't wearing thick rubber boots and thick rubber gloves greatly reduce the risk? AFAIK their resistance is almost infinite,a and thus those voltages can't hurt you. At least at the point when you touch the wire.
I have heard that although you would not be harmed yet, you ARE charged, and touching something (like the ground) would still kill you? Is this true? How can you safely release the charge if you have been in contact with the circuit?

Out of interest, how did that shock feel Wilco? Did you have to go to the hospital? Suffered burns?

It's not worth messing with that HV unless you know EXACTLY what you're doing it seems. No room for error.

dann2 - 13-4-2009 at 14:33


Hello,
Due to the nature of the circuits you will be working with, when you get a shock (hopefully not) alot of damage will have occured in a split second. This is not the type of circuitry that you will 'stick' to etc. The notion of having a buddy there to help sounds almost scary (but is no harm at the same time). The buddy would need to have a reaction time of micro seconds.
Having a large light/indicator (group of Neon bulbs behind clear red cover) connected in such a way that when there is any voltage present it will give you a very clear obvious indication of this voltage presence and thus warn you to keep clear and not do any adjustments etc untill the circuit has fully discharged.
Sitting on a bucket does not sound great. You could end up kicking that bucket (pun intended).
Dann2

Twospoons - 13-4-2009 at 14:48

Bear in mind, just ONE of those caps, fully charged, is capable of killing you. I have several suggestions for you.
1/ it would be worthwhile buying a meter and measuring the capacitance of each cap. You say these have been used in a fast discharge circuit and so may have suffered some damage. Knowing the actual value of each cap will help you set up your banks to avoid massive imbalances.
2/ buy a sheet of polycarbonate (not acrylic - it shatters) to cover everything with when in use. It makes accidental contact much more unlikely, and will contain the shrapnel, should something break.
3/ build a "chicken stick" - a plastic stick with a nail on the end with which you can short each capacitor before handling. Be aware that these caps will exhibit "dielectric recovery" and regain some charge after being shorted. Bleeder resistors are essential.
4/ Neon "telltale" lights are useful as a warning that caps are charged - theres no other way to tell without touching.

If you do shock yourself, and your heart goes into fibrillation, cardiac massage will not save you. You need a defibrillator for that.

Gloves are all very well, but be aware that high voltages can track across a surface much further than they will arc in air. I've seen a 25kV arc go right around an A4 size sheet of mylar - almost 20cm.

Sedit - 13-4-2009 at 15:49

Since you are working with such high voltages you want to put an aircore RF choke in line with the transformer to protect it from resonace that will set it self up in your lines producing high frequency RF that can blow out the insulation in your transformer. Iv lost alot of OBIT's because of not using an RF choke.

Mr. Wizard - 13-4-2009 at 17:57

Your capacitor bank sounds very interesting. As others have mentioned you are working with dangerous stuff here. To the comment about wearing rubber boots, that would only protect you if you were trying to keep the voltage from going through you, and your feet going to ground. The actual danger exists between the two electrodes of the capacitor, and neither has to be grounded. The current could flow from one arm to another, through your chest and heart. Game Over dude.

The voltage source you talk about, capable of 146 micro-amps at 9 KV, would see a discharged HV capacitor as a dead short for quite a while, until the voltage across the cap had risen quite a bit. A micro amp is 1/1000 of a milli-amp , which is 1/1000 of an amp. Your source will deliver 146 millionths of an amp. It will take quite a while to charge up a cap at that rate. Since current I=Voltage/Resistance we can figure out the effective resistance of your source. R= 9000/146x10^-6. This works out to about 62 Million ohms , 62 MegOhm using this in the formula. We can then calculate the time for the 1 micro Farad cap to 63% of 9000 Volts.
http://en.wikipedia.org/wiki/RC_delay
http://www.tpub.com/neets/book2/3d.htm

This would be RC or 62,000,000 x 1/1,000,000 or about 62 seconds to charge up one cap to .63 x9000 volts. Even your relatively harmless source could quickly build up to a very dangerous charge. The big danger with caps is they can dump huge current into you, almost instantly.

One rule is to always keep one hand in your pocket, preferably your left hand, where your heart is located. This avoids any 'across the chest' current flow, which will turn your beating heart into a quivering lump.

I hope I did the math right :-)

Centimeter - 13-4-2009 at 20:16

Would these diodes be appropriate for my circuit? Assuming of course that I don’t go above ~18kv when making the capacitor bank?

http://cgi.ebay.com/4-x-0-1A20kV100nS-High-Voltage-Diode-HV-...

These seem too short for such high voltages. If so, what are some ways to prevent arching across the diode?

Resistors of 500MOhm seem to be quite expensive. I was thinking of making my own out of these:

http://www.tedss.com/item.asp?id=2021007216

I figure 20 in series would give me ~500Mohm for quite cheap. Lots of soldering though… Would these be appropriate for my purposes? Do I need to have one between every single capacitor? I forgot to mention that the capacitors were joined together in twos when they were manufactured. Thus in actuality I have twelve 2kv and 2uF capacitors. Would a 500mohm resistor still be an appropriate choice?

It would seem that my transformer is kind of a bad choice for this circuit. Would Woelen’s power supply be the best choice for my purposes? I was hoping for higher voltages (>20KV) if at all possible…

Any thoughts on cheap DIY switches and analytical tools?

While I appreciate the warnings, please know that I fully understand that I can easily kill myself doing this and will be taking every precaution that I am capable of taking in preventing an incident. On that note, please keep the safety recommendations coming. Before doing any serious repetitive experiments, I will be sure to post pictures of the set up for safety approval.

Sedit - 13-4-2009 at 20:30

Quote:
These seem too short for such high voltages. If so, what are some ways to prevent arching across the diode?


Im sure there may be better ways But I found lots and lots of epoxy to work for isolation of the diode and prevent arc over. This should be put after the RF choke because the High frequencys can have whats called skin effect and travel alot larger distances the lower frequencys.

The comment about keeping your left hand in your pocket cant be stressed enough. Its to simple a safety precaution not to follow. I have been bitten by 10kv 23 MA OBITS so many times that I would grab them streight out if you paid me but arm to arm is a whole another story. That happened once and its like you just been hit by a car... and thats if your lucky enough to recover. It was the primary 10kv feed into saltwater capacitors until it was roughly 50-75 kv and unknown current. I am damn lucky to be here and if I kept my hand in my pocket I would have never reached out to grab the object that was falling over.

Mr. Wizard - 13-4-2009 at 20:57

Maybe you could build a Marx Generator?
http://en.wikipedia.org/wiki/Marx_generator
As far as I know, the individual components don't have to be as high as the target voltage, because they are all basically in parallel with each other at a lower voltage, until the spark cascade starts, which switches them all to series. Maybe it's for you?

Twospoons - 14-4-2009 at 14:21

Check out http://www.4hv.org Its like Sciencemadness, but for high voltage freaks :D