High Voltage Pulser Expert

Hello Shyam, I wouldn't like to call myself an expert in this field, but 24 years ago I had to design a test instrument to show the effects of changing the frequency and amplitude charactoristics of an automotive ignition coil output voltage.

This was for use on racing car engine test beds to find the optimum frequency and amplitude to give the most efficient ignition at the spark plug.

The firing pulse was from a standard TTL signal and the frequency was given by the capacitance coupled to the primary side of the ignition coil, to give the necessary ringing as the two oscillated together. That is when the pulse was required the capacitor was connected to the primary of the ignition coil so the LC time constant gave the damped ringing output at well over 100kV.

The amplitude of the output pulse was adjusted by having a small resistor in series with the LC circuit, to give a more damped output but also a smaller starting amplitude....

I don't think I've described that too well?

It wasn't a precise output obviously, but it was sufficient to find the optimum spark duration and frequency to give the best fuel burning ignition for that particular F1 racing car engine.

Sorry I've read back and I'm sure that doesn't sound like it would be of interest to you...?!

John.

Dear John

In my application, I have a DC high voltage source of 60kV 200mA. Now, the idea is to make precise switch that can work at that high rate at which mechanical things will fail.

Ignition coil does not go for such a high current and even micro amp will do. My need is for 200mA 60kV and that means solid 12kW of energy to be switched. Normally relay electrodes will get welded at this power in single pulse.

Yes, this needs an expert to do and I can pay solid US$2000 for just one switch if some one dares to make one for me. I can make it, but I do not want to rediscover the wheel. That is why I am offering funds to any one who can do things for me. I also welcome details, links, and ideas that were tested.

Thanks for your comments. Whatever you have written is right. It falls short on power and pulse shape for me. I am already in high voltage pulser designs but never did 60kV with this high power. Perhaps will be doing it if there is no one else doing it within my budget.

I think vacuum switches can work but not very sure if they are made to this fast and for such a high power. I have solid state for 30kV only with 30A. I need 60 kV now. That beats me. I am sked to do when no one else does such things in India. I like to get involved in very tough designs. I also appreciate if some friend can do it and I pay for that.

Hello Shyam,

My business I started back in 1983 is to design instruments and equipment to suit the application.

I've certainly seen high voltage and power relays which would be adequately rated, but as you have said none that could give you the speed of operation.

Is there any possibility that switching could be at a lower voltage with higher current?

This would enable a heavy duty transistor arrangement to be used or maybe a GTOSCR to be used....

Although I would assume that with a dc output there are going to be output capacitors to be discharged and charge again....

Not an easy problem!!

I will keep checking to see what others have to say.

John.

Dear John,

In this application, the requirement is fixed by the user, who said he needs DC to 1kHz pulse switching for his 60kV 200mA source. I can't alter that. It is my starting point. My input is 60kV 200mA capable source and load will draw that power for sure. I now have to find something to do that. I may be using 60 MOSFETS of 1kV each in series to make them switch all at a time. This is not an easy task. I think it will become big instrument. I hate big instruments as I am miniature loving person. I will use DEI MOSFETS and they will cost me US$40 x 60 = 2400 US$ plus driver circuit. Hence I am raising my cost estimates to US$5000 now. It may look too much for others. I like it or not, but now I am in this tough business and have nothing to compromize.

So prize money is now US$5000 for it. Just for a switch. Let me see if some one has a courage to make one.

Wow!! If that's your starting point of having a supply of 60kV 0.2 Amps DC input and having to 'chop' it at various frequencies, 0 to 1 kHz......

Then I will have to say how about.... xyz..... in other words just suggest ways it might be possible....

thinking outside the box ... (as some people say) some ideas may sound crazy but as you well know....

"from little acorns mighty oaks can grow..."

So my first thought is what is this supply used for.... or does the user need an accurate 60kV or is it roughly 60kV?

Will he be needing to draw a constant 0.2 Amp or is it load determinate?

And most importantly does he need a fast rise time for the output or can he tolerate a sine (or similar) wave output??

I think just banging the output off and on at 1 kHz and expecting a pure square wave output is expecting a little too much, don't you?

John.

Dear John,

Most of the designs I do have 0.1% regulated and with good stability, low noise and not worst than 200ppm/oC for the 60kV. 200mA is continuous to the load and Peak currents often go as high as 60A.

Here I give you simple calculation for peak current, which comes from stored charge on a capacitor.

Assume that you have 1000pF load and then some resistive load. Taking 1us rise time.

60kV x 1000pF = 60 uC charge in 1us to be dumped or 60uC / 1us = 60A average current in that 1us period. Thereafter it is resistive DC current if you ignore discharge leakage.

Hence 200mA is only what happens after 1us period. I often design with this parameter and limit the charging current to 60A peak value. It comes through cable and capacitance of the load and there may not be any real capacitor in the load. If I ignore that, then my switch will go under fumes. Another serious problem in this design is to suck back the charge on the capacitor to Ground to stop the voltage to the output such that we get perfact shape with about 1us rise time and decay time, both at DC to 1kHz.

I have many circuits already designed that have 25ns rise time for 15kV and hence, I am used to such needs.

Perhaps this must be a requirement for some continuous x-ray generator with rotating target to get a continuous x-ray beam as well as pulsed x-ray beam with stable output to within 0.1% x-ray beam current. However, they may never disclose to me the application but I can smell it because of my experience in specilities.

I manufacture a high specility delay timer that can have pulse rate from DC to 1kHz and pulse dlay 0-2000,000,000,000ns in steps of 100ns and pulse width also in steps of 100ns 100ns-2000,000,000,000ns or DC. This equipment I call ST2006.

www.sensorstechnology.com/st2006.pdf

My research is on pulsed Ion Beams, Charge imaging devices, X-ray generator, ps charge amplifier for position sensitve micro-channel plates (MCP), pulse shape signal processing in ps to ns time domain and time of flight experiments.

Hi - just a thought. Several manufacturers make photomos relays, each with 600V ratings, max current when on being 900 mA. They are about $1 each, in quantity, and you would need at least 100 of them, totalling about $100. You would be driving LED's on the logic side, probably in some series/parallel configuration.

There are at least 2 "gotcha's" I see in this scheme. 1) they have an OFF impedance that is very high, but not predictable. All in series, the 60KV may not distribute itself evenly across each, and some would probably have too high a voltage. Maybe you can use 200 of them, or more, so that the variation never exceeds the 600V rating for each. 2) The isolation rating on them is 5KV each. I wonder if there might be a way to "float" the logic (if the 60KV supply and load are referenced to ground) so that the 5KV rating on each part is never really a problem.

I think I'd operate the controls on this with a long stick ! Good luck !

Hi,

Yes, there are ways to make the control float and be with isolation up to 100kV. I can use laser fired photo cell drivers and can start the cell using light. I need fast photo cell driver. You are right. LED can also be a light source. This part is to be built in vacuum as no one will manufacturer such parts as they are specility.

Thanks for communication. It is good thinking.

Hi,

Yes, there are ways to make the control float and be with isolation up to 100kV. I can use laser fired photo cell drivers and can start the cell using light. I need fast photo cell driver. You are right. LED can also be a light source. This part is to be built in vacuum as no one will manufacturer such parts as they are specility.

Thanks for communication. It is good thinking.

This may be off the mark, but Switch Gear in any Factory have an extremely Fast High Voltage and High Current trip. So they would be rugged enough to handle your needs. It may be possible to devise a module to set the Trip frequency. The units are adjustable. for rate of rise, and settable for limiting voltage trip. Hope this helps, Fris.

Dear Fris

No electromechanical switch comes near this speed of 1us. They are in 10ms up.

I have no expertise in this field at all, but, my gut feel is that putting lots of switches in series (60 MOSFETs as above or Mike's suggestion) will be disastrous, even if you could simultaneously activate all the gates at EXACTLY the same time the different propagation times through the devices would result in one turning on last: with most of the 60 KV and a substantial amount of stray charge to switch.

I agree with you Randell.

Distriduting the voltage equally or even moderatly equally would be practically impossible....

All this is bringing me back to an old fashioned way of handling high voltages.... Using a vacuum tube.... Why not? high power transmitters used some very special valves in the output stages easily giving outputs of many 100's of kilowatts......

What about a modified mercury arc rectifier? they operate well into the 100 kVolt region at many amperes.....

Maybe something used 35 years ago is worth a backward thought??

John.

"I have solid state for 30kV only with 30A. I need 60 kV now. That beats me. I am asked to do when no one else does such things."

Why not put 2 of these in series?

Just a thought..

Sounds like a good start. Many broadcast stations still use good old vacuum tubes for their "final". Now all you need is a couple driver stages to translate the TTL level signals to the voltage required by the grid the tube. Maybe a radar tube??? With this voltage, power range and pulse rate almost sounds like a radar application anyway.

Dear John,

You are right. The valves are good and they work fine for 10kV or even more but getting Peak 10A current from them is a serious problem as they give current from filament and at 10A it will evaporate in ms. I am very happy to see that you used your brain quite a lot and are very near to good design. If current was not so high I might have gone for it.

In ESCA and SIMS single atom detection system many such high voltage valves were used and they worked fine for several years. kW energy from valves is possible. I need DC 12kW and pulsed 3600kW peak power for to shape 1us rise time pulse at 1kHz with 200mA DC out 60000V. It is falling short and will need 6 valves in series and that is not easy. MOSFETs and IGBTs are much smaller.

I think some one can make those vacuum relays that can fire lightning pulse at 60kV. I am again afraid as the current may not be possible with these to Peack 60A. I will try this as they are used for such purpose as high voltage discharge and safety.

Dear Randall,

Actually this design needs 60+60 = 120 MOSFETS of 1000V each. That may sound crazy to work on. However, that is the way it is to be done. I have switch scheme that can fire in less than 100ps skew to all MOSFETs. I manufacture such instruments. I agree it is not easy but not impossible. Engineering means near impossible work so I do that and that is what others also want me to do. Cost is further going up as cost of MOSFETs alone will be US$4800 and then other circuits. Hence IGBT may be a much better choice as I can get 3kV each IGBT and hence 20 of them required.I will welcome design ideas and not scared ideas. If you have done something like this then it is fun of life. This is something I need to build in just 90 days from now. No going back. 60kV 200mA DC and 60A Peak current fired at 12kW DC.

We are not talking something that is impossible. However, there are only few HV experts so I am hunting the right person and right experience. I know how to do it.

Shyam,

The more I think about this the more I think of valves..... They are more than capable of handling excessive voltage loads and are available in very powerful models.....

You should be able to do this with just one high power device....

John.

Dear John,

Just look at their Peak current. Can they reach 60A Peak current? Voltage yes, they can handle any. Is there such a single valve that can go for 60kV and 60A peak and 200mA DC. X-ray generators use such power sources and perhaps an expert in x-ray power supply can tell. Deep x-rays are up to 400kV and current may be only few mA and think they can not generate or pulse 200mA DC or at 1kHz with 1us rise time. Looks that they never need such high specifications. I just need that to implement.

Dear Shyam,

The last time I designed anything using valves was in the late 60s maybe early 70s when semiconductors were fragile little things that just needed a spike of over voltage to kill them!

I'm not acquainted with the specs of valves made today, I merely know that the output valves in the low frequency transmitters were giant, water cooled beasts that could handle continuous power well into the 100 kwatt range.

I am sure the surge you mention must have led to a design of cathode, either indirectly or directly heated, that would resist stripping at these power pulse levels?

I keep thinking of the mercury arc rectifiers with their continuously changing cathode of liquid mercury.

In this application the power surge is significant but it is not an excessive average power level, afterall the device you would need only needs to switch it doesn't need to be a linear device?

I'm just thinking out loud now.... let us know if and how you do solve this intriguing problem.... Please!

John.

Dear John,

Your injected this Mercury Idea is a new thought. Good. Unfortunately Mercury is now a banned item. It is still used for bounce free switching relays. I am not sure if any of those relayes were ever designed for such a high voltage. I remember that one of my friend was actually using a RF furnace using a valve with high voltage and must be using it for high current as well. There are 100kW RF furnaces. The filament was red hot and was used to heat gas to generate electron stripped plasma, hitting at deuterium and tritium targets nucleus to produce Neutrons beam of 14.2MeV.

I will sure like to look more into valve now. Perhaps those used for RG furnaces in industries to melt steel may work. Yes valve are fast switching devices. Thanks. Good thinking.

I also found one manufacturer source for solid state relay for 65kV 100A 200ns switching ON time and that will cost me 10000 Euro. It is because it uses 130 MOSFETS as I thought I should design with large array of those powerful tiny things. It is already done by some one. I may have to buy this to see how it works but is up to my specifications. Only togh people can make such things. I think I can now finish the project in 90 days as planned. I will but few numbers of those things immediately.

Dear Shyam,

It seems as if we are living in an age that is seeing most things that are useful being banned..... Cadmium, Mercury, Lead ...etc..

What will be next? and what will we have left to use?

John.

Wood.

Dear John,

Banning does not mean they will disappear from earth. Some people use these things to stear the business. This ban must be a lobby funded by XRF manufacturers, who otherwise lost their business. They are now selling these as hot cake. XRF instruments of US$100000 are now used to see if there is 1ng Lead on US$1 PCB or in 0.1mm thin SOIC leads (pins). Insn't that funny banned ROHS directive. And then you have 1000kg lead in the power inverter batteries at the same place in COOOOOOL way. It is a joke and it will kill Europian Union Business. We don't care about it. One friend from Germany asked my help to get a smal USB board developed ROHS and all my friends who otherwise used to help me raised their hand. I asked then to put lots of Gold and Silver in place of lead and they did it as I asked them to do but they are still confused, about that word ROHS. They are asking me how it is now ROHS. I asked them not worry else their sleep will be lost. In India smoking is banned, drinks are banned, looking at girls is banned. You know that it all goes not here and every where in the world. Law is one thing and your needs is another.

I think we should reduce these items from getting into environment. As we are getting dirty to consume all shorts of things and through them anywhere, this is going to be a problem. Hence, for mass production the ban looks OK. For small use, you can play with anything and if it is private then nothing to worry about. one should avoid creating public mesh. There were seriius diststers in Japan and Brazil due to Mercury mining and Mercury entering the fish and then men. Arsenic was used for poisoning when bullets were not there.

You may like to read this which was gift of Union Carbide Corporation to India.

"Hiroshima of the Chemical Industry"

http://www.hu.mtu.edu/hu_dept/tc@mtu/papers/bhopal.htm

Dear John

We can perhaps have a chat in London City. How far are you from that place?

Dear Shyam

Are you visiting then? I'm not far at all from London, it would be nice to meet for a chat....

John.

Hello,

I know that you said no electromechanical switch can come close to 1us. For a somewhat cheaper version that is not as complicated and probably very quick to build would be to get a rotary switch (a mechanical device) that is able to spin repeatedly in either direction. The rotary switch can be hooked to a motor and the motor can be PWM so that you achieve the correct rate. The hooking mechanism can be like a simple two gear transmission or any type of belt. The idea is that you can find a mechanical switch with such a high voltage rating for a cheaper price that is capable of rotating for example... google "high voltage rotary switch" and the first one that pops up will show you such a device. So now you got two modes, ON and OFF. I hope that helps you out in some way.

Oh, I forgot to mention that if you really wanted to get crazy, how long the signal stays high... there are two options to go about this. Ask the guys you get the the switches from if you can specify the contact width of the switch. The con is that it will always be a specified pulse width, but extremely precise. Another idea, use a microcontroller to control the motor along with the PWM and in this manner you can actually change the pulse width to a longer time or a shorter time. For example, have the microcontroller tell the PWM to speed the motor up when the switch is not in contact and then slow down when in contact. You would get a longer pulse width. The same idea can be used to get a shorter pulse width. So, in reality you could even transmit data with this system with high voltage. This would be extremely useful for a power company, I guess. I would say that a PIC might be able to handle this. Even nowadays they make a PIC16C that is USB capable and contains a PWM within its harvard architecture. On the plus side, it is a SMALL programmable chip. That may be more than what you want to do and this is only because you have to understand the mechanical, electrical, and computer science engineering to make this device.

Hi,

The switch works from DC to 1kHz and hence continuous ON to perhaps 500us ON and 500us OFF time with signal rise time of about 1us without any contact bounce.

while rotating contact or CAM device is a very good idea if contact frequency is known and just rotate the shaft in perhaps mercury pool to get zero bounce, hell of a current. If this is done using Teflon insulated CAM then 60kV will be just nothing.

I will say that it is a very good thinking. Only problem will be an start up speed and need of an extra switch to keep the CAM OFF until full speed is gained.

I think this all can be built within US$1000. I will give high credit to this thinking. Excellent job. This can reduce expenditure heavily and I will try it for some applications. However, this can not be exported to or from many countries as it will require mercury pool for contact debouncing.

Some of you may be interested in looking at mercury contact based rotatory connectors.

http://www.kuroda-precision.co.jp/e-top/Products/uni/uni21w.htm

Nice thinking. Good idea.

I figured you would like the idea and that it may help you out. I have double B.S. in electrical and mechanical engineering from the University of California, Davis. I just graduated UCD and work for a robotics company. I wanted to see if what I learned could help other engineers out as well. I saw that you wanted to buy a very exspensive piece of equipment and I said I bet I can make it for cheaper and I thought up that idea in just under 1 hour after reading through all the posts. I was like, you guys dont like to use the ol' basics of mechanical stuff even when the voltages are extremely hard to find for the xtors. Albeit, the mechanical idea is great and all, however, technically it wont be as perfect of a signal as the other exspensive item. This idea is a great way to make a test product or prototype, though.

You are joking aren't you?

I threw that idea out right away.... a mechanical switch operating with a connect and disconnect current of 60 A at 60kVolts!!

The contacts, mercury or other, wouldn't last long.... even if you found / designed a switch that had a lifetime of 1million operations at that current and voltage level, at 1000 switchings a second the contacts will last 1000 seconds!!

Sorry but I wouldn't even contemplate the idea unless I knew I could design a switch to operate for well over 10^8 cycles with that power level, and even then I would supply it to the customer with several sets of replaceable contacts!!

John.

exactly why I said it is a prototype... meaning it is not meant for consumers only for testing purposes.

Technically, where are you getting those numbers from... for example a statics equation? I know that every mechanical engineer wants 10^8 cycles and this is far too little. What is the function related to in terms of inputs for the number of cycles ouput. In reality, those numbers should be based off the geometry of the contacts as well as the material(s) and quite possibly the amount of power flowing through them at any given minute. Plus, you would also have to know the speed at which the switch is turning or know the time the switch is in contact. And, even more technically, you can reduce the duty cycle if you have a crummy switch to make it where the switch is actually on is less. For example, change the pulse width. You can still have a frequency but just at a different duty cycle. OR, you can go a step further and actually use the exremely brutal form of math where statistics is not involved and model the the forces applied along with the friction coefficients, using dynamic and static equ..etc etc etc.... just to get the number of cycles its capable of producing repeatedly. So, I would have to say that you would not know how to get the number of cycles of a switch built by a separate company unless you asked them specifically. They should know how long the switch will last within the confines of your problem. OR, you can build a prototype and see if any modifications need to be made to the design!

actually look at a DC motor, they have a switching mechanism in there that uses copper contacts with an iron filling bushing type of contact. How long do they last? So, I think it's probably very doable to get long life out of a rotary switch.

Another idea, if you upgrade the type of motor to a servo type you can have almost full control over what the waveform or signal generated looks like. This means it can be possible to make a square wave, rectangular wave, or just about almost any shape you could imagine within the confines of the servo. Plus you could plot the waveform easily on a laptop and do all kinds of crazy stuff. The downside is that it costs a lot of money and the lead times are very long for servos. Just another idea for those people who want to make a high voltage signal generator.

Hello to all,

I have seen this small 60000 RPM brushless DC motor that can go up to 1kHz switching. Mechanical contact motion will not stand even in prototype. Only wet mercury contacts can stand. If high pressure chamber is made say up to 25 atmospheric pressure then the corona ark distance can be reduced to about few mm for 60kV operation.

One can find motors up to 100000 RPM here

http://www.faulhaber-group.com/

As I have stated earlier, my clients need this for research and they will set the switching frequency, start delay, pulse width etc DC to 1kHz using my specialized delay timer ST1006 or ST2006 models. These timers can generate anything from DC to 1kHz and pulse delay and pulse width in steps of 100ns with 1ns accuracy 0-2000,000,000,000ns level. These are computer programmable instrument you can fill 13 digit numbers to set the pulse with rate etc and pulse generated is accurate as instruments are of highest quality in the world used for controlling high voltage pulser switches. See these links if you want to know more

www.sensorstechnology.com/st2006.pdf

www.sensorstechnology.com/st1006.pdf

With mechanical CAM devices, one can not set the timing accurately, but sure will be of great use in many applications. I greatly appreciate the thinking of many of you now pointing to electro-mechanical directions.

These are abolute research requirements and I can not alter specifications. Hence, solid state device is of greater preference. Hence, using 120 MOSFETs is a much preffered idea and that can easily give me switching frequency of DC to 1kHz and will start working immediately with first trigger pulse.

We synchronize high voltage pulses with only 100ps skew error among signals. It is possible using MOSFETs as they have that kind of skew in about 2ns ON time.

In rush currents are very high due to cable capacitance. If there is 1000pF cable capacitance, then for 1us rise time at 60kV requires greater than 60 Amps of inrush current capability for short period. This is difficult for normal devices and can weld mechanical contact in such a short time.

Marx Impulse Generators can generate 100 Amps in Million Volts easily in very short time of even 10ns. That will be my next program to build one for generating lightning pulse in the lab.

http://www.amazing1.com/marx_generators.htm

http://home.earthlink.net/~jimlux/hv/marx.htm

Often magneto type distributors are used in race cars to switch 40-60 KV at high currents.

where do you buy these? Any specifications?

http://www.msdignition.com/1distributors.htm Here is the site. It may take some digging to find the specs. These are mechanical rotating ignition parts, but they also have solid-state programmable versions. I don't know if these are adaptable for your application, but it was a thought..

Ooowww that sounds a good idea......

But having worked for an automotive design company these distributors only select the spark plug for the next ignition pulse, they shouldn't be used for switching ignition voltages and even less the currents that Shyam needs...

But for a prototype /trial version, why not?

John.

Dear Steve

I located dealer in India for MSD.

http://www.n1engineering.com/products.htm

This link for MSD was good information. If you come across any one else, then kindly give me link. I am considering this discussion very valuable.

Thanks.

I have checked out on the web site.

Yes they not only have those high voltage high current ignition coild but very accurate contact switch with rotating assembly used for firing within 4 degrees. That sounds great.

Some of the ignition coils give 200mA plus at about 48000V and perhaps that can easily make higher voltage version. These are only auto-transformers or inductors with lots of high voltage coil, with three points on same coil. not isolated version. Their pulse is for about 200us.

I think I can get some of these for testing. Very good idea. We do not have such cars in India so there will be no supplier for formula one car accessories. We have only one good car racer from India.

Dear John and Dear Steve,

This something I will love to have.

http://www.msdignition.com/dist_83_8394.htm

Dear Shyam,

Glad to help. Here is a photo of a nitro-methane fuel coupe, which uses this type of ignition. They produce 8000 horsepower and travel from 0 to 523 KPH in 4.5 seconds!

Very fun to watch. Cover your ears, though!

Dear Shyam,

I don't want to dampen your spirit, but please remember that these distributor switches are, as I said, only to select the spark plug and NOT to switch the spark at 60kVolt at any current.

The ignition coil is fired when the contact is made and the distributor does not break or make the contact at the full voltage and current. if you tried to use one to do so I suspect you would end up with a continuous arc round the distributor and no switching.

Its worth a try as they are so cheap and if it does manage to switch the full voltage then the contacts would very quickly vapourise, but could easily be replaced.

I'm sceptical of using one in your application.

An ignition coil is a handy way to generate a few hundred kiloVolts though, but only in short pulses.

John.

Yes, this is not a switch, it is a rotating arc-gap of probably less than 1 mm. The whole system would have to be utilized (processor controlled firing at the exact right time) and I'm not sure what the output waveform would look like. I am not an expert on ignition systems, I was just outputting some general ideas think-tank style to be bounced around. Switching 60KV is not an easy task and these ignition systems are built to do that. Some re-designing of the system I'm sure will be necessary. The racing type rotor caps do, however provide superior resistance against arcing at high voltages. The magnetos, I believe, are self contained high voltage pulse generators, which may or may not be useful.

You All have a nice day, Its almost time to go home here

Cheers

Also, its worth remembering that the output from an ignition coil is ac and is therefore muchless likely to maintain an arc than dc.

In case anyone wants to know why the output is ac from an ignition coil, its to stop the spark plug electrodes from eroding and depositing as for dc. The ignition coil is only a 100:1 transformer that gets its energy from breaking the primary current, which creates a large back emf in the primary of a 1000 volts and so generates over 100kVolts on the secondary, but only at a small duty cycle....

John.

John,

Good point but, I believe what you are referring to is a stock automotive setup. The systems in which I'm referring are specialized fully programmable microprocessor controlled systems which you hook to your laptop PC They deliver selectable multi-pulse discharges in which you can control the pulse width and duration. They also deliver well over stock voltages for these applications. In modern day racing, the crew chiefs and tuners have to be able to control all aspects of the engine in order to be competitive; especially the ignition system. These are not stock systems, they are very specialized and flexible.

As far as the distrbutor cap goes, yes, like you said, it just selects the circuit (spark plug) to be fired. It may be able to be written out of the equation; utilizing the pulse firing part of the system, I don't know. This is all up to Shyam. Thoughts, Shyam?

Steve

Hi steve-o,

It is some years since I worked on ignition systems for racing cars, in fact about 20 years....

It sounds like things have become much more sophisticated in that time!

20 years ago I was involved in the first designs for electronic engine control management using mapping of engines etc...

So it sounds as if you are better acquainted with these new systems than I...

I will read and keep learning hopefully

John.

John,

I am definately no expert on these systems, I just know some info from being a drag racing fan for some time. I'll try to get data for Shyam, It'll require a bit of research.

I like your logo--the Panda rolling over and blinking. That's fun!

Steve

PS--I think that these are amazing times we live in when we can intantaneously communicate from the UK to India and the US in our conversations that we are having here.. It's almost unbelieveable!

In place of electrical contacts, people also use hall effect sensor for position sensing and then electric firing. This technology is better and also ignition pulse is monitored. If there was no firing then perhaps better firing mechanism can be selected for next firing phase.

Good morning Shyam! Oops, I guess it's about 7:00 PM there? Good evening, then.

We use Hall sensors in some of the servo motors that we build here for feedback devices.

I don't quite understand your application, though.

Hi John,

Yes I have noticed that they switch only battery voltage, which drives the HV choke to generate 40kV to 48kV 200mA pulse for about 200us. Yes it will be funny ark dischaarge with sharp peak and decay.

I have worked on spark pulse generators for some experiments for up to 100kV. It was self triggerred at that time at about 1000Hz to 100000Hz and I used to count the spark pulses to measure the spark generated UV effect on LiF:Mg,Tl crystals to store energy from spark UV (high energy vacuum UV). My instrument was a thermoluminescence reader, which was heating the LiF crystal at the rate of about 40C/s from 25C to 425C in about 10 seconds. This generates light pulse from crystal called thermoluminescence phosphor, and that is what you see as glow from meteroids also in the sky (from heavenly objects).

This research was started by Marie Curie and she proposed that material can store radiation energy and it can come out on heating in the form of light. I actually have met her (Marie Curie's) grand doughter in India in one conference on radiation safety and perhaps at that time she (Marie Curie's grand daughter) was 75 and that was some 20 years ago. She was seen in nice French flowery frock dress. She is very shoft spoken person and I was sure very happy to see her. I did research on what was stated by her grand mother Marie Curie.

This makes my association with high voltage spark study of nearly 30 years.

Do a search for rotating spark gap. Alot of folks build them out of scrap materials in their garage for Tesla coil projects. Electrode vaporization and deposition is a problem but can be mitigated using multiple pairs of tungsten electrodes. Alternating each electrodes polarity per 1/2 revolution. As this is not for long term continous use, you can simply file down to fresh tungsten and reset the gap between runs. But basically you are operating on the same principle as a car distributor. Another method is to simply use a large motor commutator with brushes made of carbon graphite or precious metal.

slo