HV psu design and assorted questions

[NwcNlgUfh2]

Hi
At the moment I am trying to design a variable HV psu for some projects (some of which haven't even been though up yet, but I want to ability to run everything for hv off of the one psu).

The idea is to use a transformer(which would double as an isolation transformer) with a 600:2500 ratio to convert the 240V ac into 1000v (E(secondary)/E(primary)=N(secondary)/N(primary)) then feeding this into an 8 stage cw multiplier for an output of roughly 10,000V dc (8*sqrt(2)*1000v=11313, or 1313V worth of 'v drop'). This would then have some 'stuff' happen to it to vary the voltage (refered to as 'stuff' here as it will be detailed later) and would then end up in a switching system for either a dc output, a pulsating dc output (using a vacuum tube triode hooked up to a variable frequency ocilator) and ac (by applying the pulsating dc into a 1000:1000 transformer).

Now for the 'stuff' part. I was originally thinking of using 2 rotary switches (value+multiplier) to control which output from a massive (2,000,000 ohm total) potential divider (v=IR therefore V/I=R, 10,000/0.005=2,000,000) in order to maintain the output current at roughly 5miliamps (which would hurt, but probably not do much else). Unfortunately, this creates an issue wherein the power (V*I) for the divider to dissipate is soemthing in the vicinity of 50W, way more than any of the 1/2W film resistors I have available can take. Any ideas as to how this could be accomplished? Since I have to get some mineral oil for the multiplier anyway, could I improve the ability for the resistors to cope with this heat by placing them in a tube of mineral oil and cooling it using a pump and 120mm radiator? Is there a way this can be accomplished using some funky circuitry and a vacuum tube triode instead?

Other than the issue of changing the volatge output, how do I 'complete' such a circuit in a way to prevent Earth potential rise? I know that for the ac out, it would be through both outputs of the transformer secondary, but for the dc and pulsating dc, where do I loop it back to (using the neutral or ground on the power cord would probably cause an Earth potential rise and/or break something.

Safety is a big deal for this and so the ideas were that there would be a) indicator LEDs for when it was powered, for when it is turned on and which output is connected b) a rocker switch on the active line in, a button to check if the case is closed and having the actual ouput ports inside a bit of PVC tube, with the cable passed through a cover and another button to check if the cover is in place (can't remove the cord if powered on) & c) have the actual power switch with a 'missile lock' so it can't be knocked on accidentaly. Anything else that should be done?

 

[berkeman]

Hi
At the moment I am trying to design a variable HV psu for some projects (some of which haven't even been though up yet, but I want to ability to run everything for hv off of the one psu).

The idea is to use a transformer(which would double as an isolation transformer) with a 600:2500 ratio to convert the 240V ac into 1000v (E(secondary)/E(primary)=N(secondary)/N(primary)) then feeding this into an 8 stage cw multiplier for an output of roughly 10,000V dc (8*sqrt(2)*1000v=11313, or 1313V worth of 'v drop'). This would then have some 'stuff' happen to it to vary the voltage (refered to as 'stuff' here as it will be detailed later) and would then end up in a switching system for either a dc output, a pulsating dc output (using a vacuum tube triode hooked up to a variable frequency ocilator) and ac (by applying the pulsating dc into a 1000:1000 transformer).

Now for the 'stuff' part. I was originally thinking of using 2 rotary switches (value+multiplier) to control which output from a massive (2,000,000 ohm total) potential divider (v=IR therefore V/I=R, 10,000/0.005=2,000,000) in order to maintain the output current at roughly 5miliamps (which would hurt, but probably not do much else). Unfortunately, this creates an issue wherein the power (V*I) for the divider to dissipate is soemthing in the vicinity of 50W, way more than any of the 1/2W film resistors I have available can take. Any ideas as to how this could be accomplished? Since I have to get some mineral oil for the multiplier anyway, could I improve the ability for the resistors to cope with this heat by placing them in a tube of mineral oil and cooling it using a pump and 120mm radiator? Is there a way this can be accomplished using some funky circuitry and a vacuum tube triode instead?

Other than the issue of changing the volatge output, how do I 'complete' such a circuit in a way to prevent Earth potential rise? I know that for the ac out, it would be through both outputs of the transformer secondary, but for the dc and pulsating dc, where do I loop it back to (using the neutral or ground on the power cord would probably cause an Earth potential rise and/or break something.

Safety is a big deal for this and so the ideas were that there would be a) indicator LEDs for when it was powered, for when it is turned on and which output is connected b) a rocker switch on the active line in, a button to check if the case is closed and having the actual ouput ports inside a bit of PVC tube, with the cable passed through a cover and another button to check if the cover is in place (can't remove the cord if powered on) & c) have the actual power switch with a 'missile lock' so it can't be knocked on accidentaly. Anything else that should be done?

Welcome to the PF. :-)

I see you are in grade 11 in Australia and are studying for your advanced amateur radio license. Good for you. And good that you are thinking about safety issues in this project. What low-voltage projects have you done so far? It's important to get some experience working with SELV (safety extra low voltage, basically <60V) before venturing into AC Mains or HV projects...

 

[NwcNlgUfh2]

Welcome to the PF. :)

I see you are in grade 11 in Australia and are studying for your advanced amateur radio license. Good for you. And good that you are thinking about safety issues in this project. What low-voltage projects have you done so far? It's important to get some experience working with SELV (safety extra low voltage, basically <60V) before venturing into AC Mains or HV projects...

Low voltage I have done some messing around with an arduino for led cubes etc & I have an 'almost' finished a scanning tunneling electron microscope which I have been building on and off fir a while now. Currently nearing completion is a nixie tube clock, but that's at about 300V (powered from 9V battery and inductor).

 

[berkeman]

Low voltage I have done some messing around with an arduino for led cubes etc & I have an 'almost' finished a scanning tunneling electron microscope which I have been building on and off fir a while now. Currently nearing completion is a nixie tube clock, but that's at about 300V (powered from 9V battery and inductor).

Good stuff! What are you using for the vacuum system on your STEM?

http://mrsec.umd.edu/images/SEF/SEM-lg.jpg

 

[NwcNlgUfh2]

Not

Good stuff! What are you using for the vacuum system on your STEM?

http://mrsec.umd.edu/images/SEF/SEM-lg.jpg

We aren't using a vacuum system on it, the design is based off of one that appeared on geo-cities a short while ago.
We have a couple if op-amps set up to control a peizo disk (bending it where it needs to go) and create outputs for an oscilloscope's x and y inputs. We then have the 'z' axis done using an opamp to ensure its voltage stays constant going into the tip, and another few set up to amplify the outputs and turn it into a useable z input for the oscilloscope.

Other than that we etched a set of tungsten tips and have them stored in parafin oil, but we are probably only going to be able to get one good output from it anyway before 'something' breaks-a vacuum system would be ott for it since the tips are already probably oxidized a little and/or bent,

I say 'we' in this because this was done alongside my slightly crazy high school physicist teacher, who has guided me through most of it.

Also, any ideas on making a 2Mohm potentiometer that can dissipate 50W? I was thinking using something like a dc motor as a 'dump load' but I don't know if anything could really take the 10Kv that would be going in.

 

[berkeman]

Can you post the schematic for your 9V to 300V boost circuit? That sounds like a good background for the PSU you are wanting to build now.

You generally wouldn't use a potentiometer to give you a variable output for a PSU. Instead, you build a switching boost power supply that can vary its output voltage. And you can incorporate some of the HV boost ladders that it sounds like you have been looking at.

I'm glad you have a Mentor in all of this (albeit a :"slightly crazy" one). That's an important part of learning electrical safety, as well as building fun projects together.

I probably won't have Internet access for the rest of the weekend, but will check back in on your thread on Monday. :)

 

[NwcNlgUfh2]

Can you post the schematic for your 9V to 300V boost circuit? That sounds like a good background for the PSU you are wanting to build now.

You generally wouldn't use a potentiometer to give you a variable output for a PSU. Instead, you build a switching boost power supply that can vary its output voltage. And you can incorporate some of the HV boost ladders that it sounds like you have been looking at.

I'm glad you have a Mentor in all of this (albeit a :"slightly crazy" one). That's an important part of learning electrical safety, as well as building fun projects together.

I probably won't have Internet access for the rest of the weekend, but will check back in on your thread on Monday. :)

Hi, thanks for the reply, the reason I was thinking of using a potentiometer/potential divider circuit was so that I could also drop the output amperage to about 5mA, which would be painful, but not too dangerous (relatively speaking) to be shocked with. Since the value of whatever resistor I used to do this current drop was going to vary with the voltage, I though I may as well use it for that purpose at the same time. Is there a way I can just limit the maximum possible current that can pass through a section instead (probably would be placed on the ac in side, before being stepped up) or another way of achieving this same current limit?
I have included the psu design for the Nixie tubes (from instructables) however I don't think that it would scale up to well and I already have had great difficulty in finding a mosfet rated for for voltage this one would be dealing with.

 

[mfb]

The 2MOhm resistor would limit the current to 5mA if you use the full resistor as series resistor - but then you cannot use it for voltage regulation.
If you use it as potentiometer, it does not work as hard current limit.

 

[NwcNlgUfh2]

The 2MOhm resistor would limit the current to 5mA if you use the full resistor as series resistor - but then you cannot use it for voltage regulation.
If you use it as potentiometer, it does not work as hard current limit.

I feel like I'm missing something here, isn't it that for resistors in series the same current is across each if them and found through V=IΣR therefore, if there are two resistors, and they total to 2,000,000 ohms, that it would limit the current across them both when a 10Kv potential is applied to 5mA and then the voltage output at 5mA would be R(second resistor)/total resistance?
What am I missing and how else could I accomplish this?

 

[jim hardy]

A brute force 10KV with resistance current limiting would be dangerous, that high voltage is just waiting to break out through your resistor and ...
I'd think you might want a more modern design, some sort of flyback converter that'll be inherently current limited.

There's so much information out there that it's like tying to drink from a fire hose...

http://www.ti.com/lit/an/snva006b/snva006b.pdf
http://www.ti.com/lit/ml/slup261/slup261.pdf
http://www.smps.us/Unitrode.html
http://ww1.microchip.com/downloads/en/DeviceDoc/dsPICSMPS AC_DC Users Guide.pdf

search on the keywords in those links.

I'd look for an old plasma globe to study. Some folks modify CRT TV flyback transformers to make plasma globes.

You're experimenting at this level when still in high school? Marvelous ! In 12th grade i built a 15 watt tube hi-fi amplifier, the one in back of RCA tube manual.

Have fun, but always think "What If" and make your contraptions safe for those tiny little innocent and curious fingers .

old jim

 

[NwcNlgUfh2]

A brute force 10KV with resistance current limiting would be dangerous, that high voltage is just waiting to break out through your resistor and ...
I'd think you might want a more modern design, some sort of flyback converter that'll be inherently current limited.

There's so much information out there that it's like tying to drink from a fire hose...

http://www.ti.com/lit/an/snva006b/snva006b.pdf
http://www.ti.com/lit/ml/slup261/slup261.pdf
http://www.smps.us/Unitrode.html
http://ww1.microchip.com/downloads/en/DeviceDoc/dsPICSMPS AC_DC Users Guide.pdf

search on the keywords in those links.

I'd look for an old plasma globe to study. Some folks modify CRT TV flyback transformers to make plasma globes.

You're experimenting at this level when still in high school? Marvelous ! In 12th grade i built a 15 watt tube hi-fi amplifier, the one in back of RCA tube manual.

Have fun, but always think "What If" and make your contraptions safe for those tiny little innocent and curious fingers .

old jim

Thanks for the links, I'm at that unfortunate point where I know bits of what works how, but not enough of 'this can be done like this' examples. As for the fly back, I have heard of them (and where to find them) but I haven't ever dealt with one and I have conflicting information in that some sources say they emit ac, and some say dc (I'm after dc here to allow for the variable frequency pulsating dc and ac outs). Which is it? Also, how would I set one up for a variable output voltage?
Also, this would probably be mounted in something like Perspex or acrylic, with many of the parts being in plastic containers full of mineral oil for insulation/reduce corona loss -hence no hand should be getting in - and it will be designed to cut out if the case is opened.

I'm not quite sure what is meant by your 'Marvelous' comment, but I do have a mentor in this, unfortunately he is moving schools (family friend so will remain in contact) so anything that I can design/safely figure out for myself would be beneficial.

 

[zoki85]

Please, be careful with these things. Long ago I built a small full wave 6 stage Cockroft-Walton generator. I powered it with 6kV/30mA neon sign transformer. The generator could put out 2-3 cm long sparks and sparks seemed rather weak per see. I thought nothing serious could happen if I touched the output. It turned out to be quite a paintful jolt..

 

[NwcNlgUfh2]

Please, be careful with these things. Long ago I built a small full wave 6 stage Cockroft-Walton generator. I powered it with 6kV/30mA neon sign transformer. The generator could put out 2-3 cm long sparks and sparks seemed rather weak per see. I thought nothing serious could happen if I touched the output. It turned out to be quite a paintful jolt..

As long as it was only painful, not dangerous (didn't fibrillate/paralyze you etc).
Although, 30mA in, divided by 6*2^(1/2) is only about 3mA so I wouldn't expect much worse from it...

 

[jim hardy]

By "marvelous" i meant i think it's quite admirable that you have the interest and confidence to do hands-on experimentation..
In my high school electronics class every Friday was project day. Teacher required us to build something using real hardware and tools. That was in vacuum tube days when junk TV's could be scavenged for parts. (Probably very few here remember the "Zenith Hand-Wired Chassis" advertisements. ) Over my career i ran across several more of his students. We all agreed his pushing us toward the practical had been a big influence on our lives.Here's an introductory page.
http://www.kronjaeger.com/hv/hv/src/fly/

Of those links i posted, the first one is the best for getting started - it sort of defines terms and shows the concepts of how to set up the variable control.
Mind you I've built from scratch only one switching supply. I was an industrial repairman so got to fix them instead. I'm a believer in studying existing designs .Have you built simpler supplies? A variable dual tracking unit in the ~5 to 20 volt range is handy for an electronic hobbyist's workbench. The high voltage unit is more complex because of isolation and insulation needs and would be a good "next one".

Nixie tubes! I think i still have a Nixie tube frequency counter ! Hadn't thought of it in years, must look for it.

Whatever way you go, have fun !

old jim

 

[mfb]

"I feel like I'm missing something here, isn't it that for resistors in series the same current is across each if them and found through V=IΣR therefore, if there are two resistors, and they total to 2,000,000 ohms, that it would limit the current across them both when a 10Kv potential is applied to 5mA and then the voltage output at 5mA would be R(second resistor)/total resistance?"Let's take an example: you want 9 kV, so you divide it 200 kOhm <-> 1.8 MOhm, and connect your output in parallel to the 1.8 MOhm resistor. If your circuit outside gets a short, the equivalent resistance of [output plus 1.8 MOhm resistor] becomes zero, the whole 10kV are at the 200 kOhm part and you get 50 mA.

 

[zoki85]

It was definitely worse feeling than 3mA continuous.
In my circumstances, it was discharge of a small capacity charged to high voltage through my right arm several times in a row.. Peak currents measure in amps, but last for a mere a fraction of μs. Won't do much harm if the energy is small , but can be very painful (even at low average power)

 

[berkeman]

It was definitely worse feeling than 3mA continuous.
In my circumstances, it was discharge of a small capacity charged to high voltage through my right arm several times in a row.. Peak currents measure in amps, but last for a mere a fraction of μs. Won't do much harm if the energy is small , but can be very painful (even at low average power)

Let's all keep in mind that when getting shocked, the damage can depend on the path of the shock through the body. If you take a shock from your fingers to your elbow (like you are leaning on something with your elbow that is grounded, and get shocked on that same hand, then the damage may not be too great. But take that same shock arm-to-arm (like from one hand to the other), and that current passes through the heart. Very bad things can happen from that.

Stay smart and safe! :)

 

[zoki85]

Stay smart and safe! :)

Yeah, otherwise possibility of winning the Darwin award increases. When I was younger I deliberately exposed myself to electrical shocks of various sorts, in carefully prepared experimental set ups (still far from "safe"!). And shocks from capacitors were actually among worst by my judgement.

 

[davenn]

Nixie tubes! I think i still have a Nixie tube frequency counter ! Hadn't thought of it in years, must look for it.

LOL ... I still do ... The HP 5248L freq counter

 

[Mike_In_Plano]

When I was young, I build my variable DC monster using variac, a neon transformer, and some surplus high voltage rectifiers.
I've been involved in several supply designs that were switching in nature, but none perform as well when you attempt regulation electronically in comparison to when you simply regulate the voltage at the switching device with a variac.

Also, purely resonate typologies (i.e. sine waves) are a savior in this business. Impedance going through a transformer increases to the square of the turns ratio, so "little" capacitances become huge nightmares when you attempt to hard switch the primary.
The same "little" capacitances become frustrating when you attempt feedback - especially once you've added oil or potted the design. Then the leakage capacitances change. That is why many professional shops insist on Flourinert as a rapid means of test - It washes down clean for small changes.

Another heads up - If water gets in your oil, you can have "streamers" at high voltage that arc over and burn the oil.

Extra heads up - If you are impulsive, this is probably not the business to be in. Walking away before you become frustrated is important. Trusting your feelings regarding impending accidents is another. Finally, 1/4" Poly-carbonate is your friend. I would not take up work in a high energy field without Poly-carbonate shields. The help reduce the UV energy from arcs, keep expelled liquids from your face, hands from the work, localize fires,and in worst cases, keep flying debris from your body.

- Mike