HV power supply help

[kffern]

I am trying to build a HV PS for my electrostats and am having a few problems.
I have bought a fluoro invertor and I am having a problem building a voltage multiplier. With 12v AC in I measure 18V at the first cap, 34v at the second cap and 26v on the next and the same till the end.
I built the same using 2 screw terminal blocks and new components just to see if one of the components was bad but it measures the same.
What am I doing wrong? This is supposed to be childs play!

An externally hosted image should be here but it was not working when we last tested it.

With 12v in to the inverter the output starts at around 200v AC and slowly climbs up to around 250V. Only my Fluke autoranging multimeter will read it and it has a lowZ meassage on the display.

An externally hosted image should be here but it was not working when we last tested it.

 

[kffern]

This will be fed by a variable DC using a LM317 with a trimmer giving 0 to 15v DC.
kffern

 

[geraldfryjr]

You won't be able to reliably measure the voltage with your average multimeter.
The input imepedance of a meter is to low and will load down the multiplier as it won't produce enough current to drive the meter.

I have a whole discussion on this in one of the threads.
I use a very high resistance 1000:1 Voltage divider with a unity gain opamp buffer on the bottom resistor to drive the meter.

This is the most accurate way to measure such voltages.
See this thread for more info and especially post 50 for the complete schematic and details,

http://www.diyaudio.com/forums/plan...tor-insulation-mylar-coating.html#post2767031

and this thread,

http://www.diyaudio.com/forums/plan...estion-esl-bias-power-supply.html#post2597135

My second version uses a 300Megohm resistor stack and I use it to directly measure the output of the step-up transformers directly into my sound card and/or O-Scope!!

jer

P.S. One more thread that may help you,

http://www.diyaudio.com/forums/plan...easure-bias-voltage-acoustat.html#post3791524

 

[geraldfryjr]

Here is the post that shows the schematic of my simpler Transformer Test Jig and the power supply for reference.

They both use the same methods just one has a higher resistance voltage divider (Much Better!!) than the other.

http://www.diyaudio.com/forums/plan...me-fix-these-poor-ml-sequels.html#post3775769

jer

P.S. My First version in the HV power supply used the 300meg stack and the Test jig used a 20meg stack Vice-Versa of what I had stated earlier.

 

[bolserst]

...Only my Fluke autoranging multimeter will read it and it has a lowZ meassage on the display.

Which Fluke model do you have?
When operated in lowZ mode the meter input impedance is about 3Kohm.
There should be another AC voltage measurement setting that gives you a 10Mohm input impedance that will not load down the circuit you are measuring so much.

 

[kffern]

Thanks for all those links Jer. I searched HV power supply not bias.

I am using one of these http://www.altronics.com.au/p/k2559-high-voltage-eht-probe-dmm-adaptor-kit/http:// but read nothing with the 12V AC supply - to low I guess. It is as you say but with a 50K pot to calibrate. I set it to 25K which is meant to be close enough if you don't have a reliable reference.
I measured a few of the Kingsound supplies which had voltages scribbled on the bag and they were within 0.1Kv. I can't vary these supplies as I mentioned in my Kingsound thread.
http://www.diyaudio.com/forums/planars-exotics/282553-kingsound-king-electrostatic.html

I will hook it to the inverter and see what happens. I haven't found a thread where someone has used the variable DC - inverter - C-W multiplier option.

I had the multiplier mounted on a small prototyping PC board but took it apart thinking I had it wrong.

Regards,
kffern

 

[kffern]

Which Fluke model do you have?
When operated in lowZ mode the meter input impedance is about 3Kohm.
There should be another AC voltage measurement setting that gives you a 10Mohm input impedance that will not load down the circuit you are measuring so much.

Hi bolsert,
Its a Fluke 7-600
Fluke 7-600

I've had this for ages but rarely use it because I thought it was giving trouble but I may have been wrong. Just downloaded the manual.
Thanks,
kffern

 

[bentoronto]

I use a very high resistance 1000:1 Voltage divider

Nobody could be more supportive of creative people doing audio projects. More power to you.

On the other hand, the goof you are making should be a wake-up call that you may be short on the electrical background needed to proceed safely.

Everybody working with ESLs needs a high resistance high voltage meter. Like Gerald, I use 1000:1, 100 meg to 100k. Any digital voltmeter can read voltages across 100kOhm resistor (their input resistances are a bunch of megs).

You can buy high value, low wattage, high voltage resistors (which are special creatures) to form a 100meg chain (plus one 100k ordinary resistor) on eBay from the Ukraine. (I hope you will buy some so I can find out if they are legitimate.)

Ben

 

[kffern]

Nobody could be more supportive of creative people doing audio projects. More power to you.

On the other hand, the goof you are making should be a wake-up call that you may be short on the electrical background needed to proceed safely.

Everybody working with ESLs needs a high resistance high voltage meter. Like Gerald, I use 1000:1, 100 meg to 100k. Any digital voltmeter can read voltages across 100kOhm resistor (their input resistances are a bunch of megs).

You can buy high value, low wattage, high voltage resistors (which are special creatures) to form a 100meg chain (plus one 100k ordinary resistor) on eBay from the Ukraine. (I hope you will buy some so I can find out if they are legitimate.)

Ben

I don't know how many times to say this but I am using one of these:

http://www.altronics.com.au/p/k2559-high-voltage-eht-probe-dmm-adaptor-kit/
(Silicon Chip Magazine April ‘10) Measuring high voltages on a standard DMM is usually limited to 750AC-1000VDC. This probe allows measurements up to 23-25kV DC as found in CRT based scopes, computer monitors and TVs. High voltages are also commonly found in laser printers, photocopiers and microwave overs.


Thanks Ben, but yet again you type before you read.

kffern

 

[bentoronto]

Sure. Me real bad boy.
Ben

 

[alexberg]

RNX0381G00FNEE Vishay Dale | Resistors | DigiKey
Use two or three of them in series. Do not forget to clean with alcohol before use to get correct reading.

 

[bolserst]

Its a Fluke 7-600

Huh, not familiar at all with that model. According to the manual it always uses a lowZ setting for measuring AC voltage and auto adjusts the input impedance to something between 2K and 300K depending on the input voltage. It also isn't clear how it decides whether to show AC or DC voltage for a signal containing both a large DC component and large ripple. It appears there is no way to manual set the input impedance to the more common 10Mohm used for most DMM. So, overall a DMM not suited to measuring low current HV multiplier circuits.

Is this the DMM you are using with the HV probe kit you mentioned?
According the the instructions for the kit, it should be used with DMM having 10Mohm input resistance to get proper readings.

You mentioned using this DMM for measuring the inverter, but now that I reread your original post it isn't clear if it was the same DMM you used to measure the multiplier with 12VAC at the input.

One last question, you mentioned that only the Fluke 7-600 could measure any voltage from the inverter...what other methods did you try to measure it?

 

[kffern]

Is this the DMM you are using with the HV probe kit you mentioned?
According the the instructions for the kit, it should be used with DMM having 10Mohm input resistance to get proper readings.

No. I used a now expired old "Micron" - manual select marketed by our local electronics diy shop. It was a pretty good one but now discontinued. I do have the manual somwhere.
Late night, day before yesterday I decided to work on it. I had connected the multiplier to the inverter with variable DC input. I forgot that the HV probe wasn't connected as it was a bit of a jumbled mess on the bench and picked up the wrong probes. I had it for around 8 years and it measured correctly when I used it on some spare Kings HVPS that had fixed voltages written on them. Bought a new meter yesterday.
http://www.altronics.com.au/p/q1070-20-range-true-rms-digital-multimeter/
Tried it on one PS and it seemed Ok but will try again when I get back home tonight.

You mentioned using this DMM for measuring the inverter, but now that I reread your original post it isn't clear if it was the same DMM you used to measure the multiplier with 12VAC at the input.

When measuring the output of the inverter I used the Fluke which started around 200V and slowly built up to 250-260V AC. My other (now expired) meter would not read anything. When I used the HV probe with the Fluke (on a working HVPS) it measured much lower than the now dead meter. I mainly used the old Micron with the HV probe.

One last question, you mentioned that only the Fluke 7-600 could measure any voltage from the inverter...what other methods did you try to measure it?

Just the Micron (now expired)

Hope I have been a bit more clear this time. Thanks for all the help.
Actually the reason I am doing this is that after rediaphragming the old ER audio panels I realised that one of the speakers would buzz intermittently. This is what I think caused the diaphragms to be full of small holes before. Measuring the HV over time showed it was jumping around a bit upto 1KV or more. I changed the 5K pot and LM317 on the DC PS, but it is still the same. Its quite simple, using a 2N2219A transistor which I'm waiting on a replacement. So in the meantime I thought I'd try and diy a new one.

Thanks,
kffern

 

[geraldfryjr]

HV resistors are nice, However they are costly and have to be special order in some cases.
I used just plain ole' 1/4 watt resistors from Radio shack in my divider.

Using 20x1Meg resistors I was able to have a disipation of only 5 watts at 10Kv, this desien was for my transformer test jig.

Typically those types of resistors have a voltage cofficient of about 250V-350V and a maxium for some brands at 500v each.

That particular divider never sees more than 5Kv max on my bench the a dissipation of only 1.25 watts and a maximum of 250V across each resistor in the string.

When I built it, I had measured each individual resistor as they don't have a very low percentage of tolerance, this allow me to select the bottom resistors range to get exactly the division ratio that I needed.

I also used a multi-turn trimpot in the lower resistance to be able to adjust it to where it exactly needs to be.

I am sure that my accuracy is atleast .001% to .0001% and/or better.
It is hard to be sure since my cheapy average DMM has only 4digits to work with.

I used my scope and sine waves to make the finest adjustment by comparing the input signal with the output signal since my attenuation range on each cahannel 1:1000, or 5v to 5mv plus 5X gain.

Here is one more thread (post 6-7) that I was looking for that explains them in a little more detailed with pictures of the units in action.

http://www.diyaudio.com/forums/plan...as-voltage-sound-output-data.html#post4070779

For the "low power" supply that you are working with I would consider something closer to the 100megohm range as previously suggested.

You can't go wrong with this method as it is cheap to build and can be very accurate depending on how meticulous your are in your final calibration and the type of opamp you chose to use, I used a TL082 and similar with great results, but when I do finally make it permanent in a case I will be using some precision opamps form LT.

Cheers !!!

jer

 

[WrineX]

i just buy hv inverters. add the cockroft walton cascade on the output. now you can use normal 1-12 volt dc to control the ouput, and the switching and thus riplle will be at a way way higher frequency as well. also you only need a few multipliers so less expensive caps. and pretty safely fed with a adapter.

might need a resistor on input and on output. just for the safety of it the shocks are real


inverter 5 dollar adapter 5 dollar voltage control completely assembled 2 dollar(can be set with a pot) or something. cascade few bucks


damn ME !! you use a inverter already and a LM same setup hehe. i usually measure the AC out the inverter since it is still within the limits of the meter, then just multiply by the stages you got.

excuse me for not reading the whole thread

 

[geraldfryjr]

Also one more note nobody had seemed to mention,

What type of Diodes are you using?

Those little inverters run at a much higher frequency than what the line is at the line is at.
You should be using diodes that have a Fast Recovery or else you will have great losses and the output voltage won't be as high as it should be when you are trying to draw some current from the supply.

Using a higher capacitance value ( If you have them) helps too, This just increases the current capability for the most part.
I actually used a mixtured of values as that was all I had on hand, I used some .1uf caps on the very First few stages and then .01uf for the rest of the stages in my supply.

Here are a couple more excellent links to help you,

http://www.voltagemultipliers.com/html/multdesign.html

http://www.extremeelectronics.co.uk/calculators/cw-voltage-calculator/

By using the above calulator you can determine almost exactly the performance of your supply excluding the Diode losses.

Those inverters run at higher frequency's, usually in the 10's of kilohertz range and most DMM's wil not give you a proper measurement readings at those frequency's.

jer

 

[WrineX]

Also one more note nobody had seemed to mention,

What type of Diodes are you using?

Those little inverters run at a much higher frequency than what the line is at the line is at.
You should be using diodes that have a Fast Recovery or else you will have great losses and the output voltage won't be as high as it should be when you are trying to draw some current from the supply.

Using a higher capacitance value ( If you have them) helps too, This just increases the current capability for the most part.
I actually used a mixtured of values as that was all I had on hand, I used some .1uf caps on the very First few stages and then .01uf for the rest of the stages in my supply.

Here are a couple more excellent links to help you,

High Voltage Multiplier Design Guidelines

CW Voltage calculator | Extreme Electronics

By using the above calulator you can determine almost exactly the performance of your supply excluding the Diode losses.

Those inverters run at higher frequency's, usually in the 10's of kilohertz range and most DMM's wil not give you a proper measurement readings at those frequency's.

jer

that calculator is nice Thx

 

[kffern]

Also one more note nobody had seemed to mention,

What type of Diodes are you using?

jer

I used IN4007 diodes and 0.01uF/3KV caps.
I haven't had the time to do anything more as yet.
Thanks for the links. Will read up.
Regards,
kffern

 

[kffern]

damn ME !! you use a inverter already and a LM same setup hehe. i usually measure the AC out the inverter since it is still within the limits of the meter, then just multiply by the stages you got.

excuse me for not reading the whole thread

The inverter link is at the start of the thread - A$8.8 plus postage, which was the smallest wattage and cheapest i could find on the web.

The DC reg is a tiny board kit from Altronics
http://www.altronics.com.au/p/k3220-universal-regulator-kit/

I substitute a 5K trimmer in place of the resistors/jumper. It is very sensitive to adjust and I might fit an offboard multiturn pot if I can.
kffern

 

[geraldfryjr]

My plan is to get one of mini variable CFL modules and do another build of my regulated version using one of those.

Most the them use the exact same circuitry that I came up with, only I don't have a recipe for the transformers I made, I used some torroid cores that I had torn out of an old dead PC supply and I had to do a lot of experimenting to get them to work correctly.

I scope is essintial to do that kind of design work.

jer