Help with a voltage multiplier

In my junque box I have a High Voltage, Voltage multiplier. I pulled it out to help my son on a project and it doesn't work as I expected. Testing of the parts tells me they are all good. But the voltages don't seem right. I'm using a Bryman 235 DVM with a 10MΩ input impedance. So, I'm don't expect I'm loading the circuit. I took a picture and put voltage reads on the picture with some info about the diodes and caps. The diodes are about 1" long and have a 10v drop at 2ma. I would like to know the voltage ratings of the diodes and caps, but don't find any info on the net. Any Idea, why my voltages are way off? Mikek

formatting link

Reply to
Lamont Cranston
Loading thread data ...

What frequency is your AC source?

It would be easy to measure the caps. Then teach the kid about Spice.

Reply to
John Larkin

60 Hz, I did measure the caps, 3.1nF. Just need HVDC for a project, I thought this would be helpful, but it's not working as I expected. I have HV transformer, just wanted to use this to get DC. Mikek
Reply to
Lamont Cranston

At 60 Hz, each cap is about 800K ohms.

The mult was probably intended for high frequency drive.

Reply to
John Larkin

Oh, I hadn't thought about that. I thought it came with some ham radio junque and just assumed it was 60 hertz. Now I wonder what its original use was? Also why I can't find data on the diode. Thanks, Mikek

Reply to
Lamont Cranston

-----------------------------------------

** May well be 60Hz drive, if meant for use with an electrostatic tweeter. The component values and number of stages looks right for that app.
** Might be way obsolete types. Expect them to be rated at 10kV or more with very low leakage.

FYI: Such multipliers cannot be tested with a regular DMM alone. Needs a high voltage probe, with at least 1000 Mohms resistance.

...... Phil

Reply to
Phil Allison

-------------------------------------

** FYI:

Only one diode has EDI 7712 on it - the other 7 are all labelled EDI 7711. It's a DATE code, 11th or 12th week of 1977 !!.

The maker of the RHC8-5 is "American Microsemiconductor". See vid:

formatting link

....... Phil

Reply to
Phil Allison

Am 02.05.2023 um 09:04 schrieb Phil Allison:

Maybe EDI stands for this company:

formatting link
They have a RHC25-8
formatting link

Reply to
Matthias Czech

-------------------------------------- Phil Allison:

** OK - EDI make or supply the same number diode with their initials on it.
** Going on that example, the OP's diode is a fast, 5mA, 8kV type. All fits with the idea it supplies only a tiny (uA) current from a 60Hz supply.

An electrostatic tweeter, from the late '70s, is still a likely use for the multiplier. The molded ABS mounting plate is big clue too.

.... Phil

Reply to
Phil Allison

I hadn't seen you post for a while. For some reason I hope you're doing ok. The group would not be the same without you. (That's probably about as close you'll get to a compliment from me.)

Reply to
Chris Jones

Oh very good, thanks all, for the diode info. So, I have 60Hz HV transformers 8000V, If I use two diodes in series to build a half wave rectifier, should I put equalizing resistors across the diodes. What value? The largest value HV resistor I have is 2.5MΩ, is that to small?

Next, I want to limit the current to say 3ma, the transformers themselves are impedance limited to 15ma when short circuited, but I want to protect the diodes from over current. How would I accomplish that? Phil, it is good to hear from you, hope all is well. The speaker idea for the multiplier is interesting. I have had the multiplier board in the junque stash for so long, I don't know where I got it.

Thanks all, Mike

Reply to
Lamont Cranston

No. The HV diodes are already a lot of junctions in series.

A long time ago, diodes in series could fail because they had a negative-resistance breakdown mode, so people put a lot of equalizing passives. Modern diodes avalanche nicely.

What value?

Use the two diodes as a half-wave doubler and let the cap impedance limit current.

What do you want the HV for?

Reply to
John Larkin

If multiplication is ok in the first stage, at the frequency present, then subsequent stages should also work. The first stage passes the most current.

Check AC present on each stage - if it disappears, check for a short.

Keep in mind that a uA of load will cause a volt of droop in an 8nF cap at 60Hz - 1mA would cause 1KV of droop. So if this does not have a static load, it will only supply very low current.

RL

Reply to
legg

My son is doing some oil/water separation experiments at work, I don't really know what their vessel looks like or exactly what they are doing, but it looks like fun. I like the half-wave doubler idea to limit current. Mikek

Reply to
Lamont Cranston

He probably doesn't need much current.

Paper electrophoresis (sp?) is another fun thing that kids can do with high voltage. I still can't believe that it works, but it does.

I built a Kerr Cell as a high school experiment, using AC from a neon sign transformer. I used nitrobenzene, which turns out to be dangerous to deadly on skin contact, not to mention flammible. In a couple of Nero Wolfe mysteries, it was the murder weapon.

formatting link
Back when, any kid could go to a chemical supply house and buy anything, deadly poisons or obvious explosive ingredients. And we got used neon sign transformers free too.

Reply to
John Larkin

We think that's correct.

Yep, I got 6 or 7 when my bosses father cleaned out a build that previously held a neon sign company about 30 years ago. A few had an open on one side of the center tap.

So, experimenting, I put in 4Vac and get out 644Vdc of the doubler, that is a 160 to 1 ratio after the doubler. The transformer is about a 75 to 1 ratio. After peak voltage and a little load, seems right. Can I just measure the primary current and divide by 160 to get an idea of the DC output current? As it stands 20Vac in to the transformer gives me a nice arc, and 320ma of AC on the primary. If I can divide as I suggested, that is more current than I want thru the diodes, and I still have a 5 times multiple to get to 120Vac in and near 19kV out. i.e, I think I need more limiting. Mikek

Reply to
Lamont Cranston

I'm down one cap after experimenting. As I stated before the Dissipation of that cap went from 0.003 up to 0.320 and the Capacitance didn't change. Over night the capacitance did drop by about 10% and the dissipation dropped to 0.009. That's still 3 times the D of all the other caps. I'm calling it bad. So now I'm using it as a quadrupler, excess parts removed from circuit. With 150Vac input, I measured Voltage with 10MΩ meter at the output at 184Vdc, I have series connected Nine, 10MΩ resistors to make a 10 to 1 divider with my DVM. I now measure 684Vdc. Difference, 10MΩ load vs 100MΩ load. Using this data, I calculated a 43MΩ output impedance of the voltage quadrupler. I expect with no load it has an output of about 958V volts with 150Vac input. As John said at the start, my frequency is too low or said another way, my caps are to small.

Mikek

Reply to
Lamont Cranston

--------------------------------------

** This link might be useful to you:

formatting link

.... Phil

Reply to
Phil Allison

Thanks Phil. Mikek

Reply to
Lamont Cranston

On a somewhat different note: Looking through some experimental literature on oil/water separation using using electric fields. There is AC, sine and square, 50Hz to 50kHz, DC and pulsed DC, so I think the starting experiment is with AC, 60Hz. Experiments showed little affect of frequency on separation efficiency. I'm seeing glass vessels wrapped with a copper foil and an electrode in the center of the glass vessel in the solution. Applying the HV to this sets up the electric field, but I expect very little current. The AC out of the neon sign transformer isn't to bad to measure as the transformer output impedance is 295K, although 1000V open circuit measures 971V when loaded with a 10MΩ meter. ( I did build a 90MΩ prob e) I'm curious to know what current the Vessel with the electric field is drawing? So, I'm thinking of doing the usual, putting a resistor in series with the load and measuring the voltage drop across the resistor. In this case the current is so low, I'm thinking 1MΩ resistor in series will only drop 1V at 1ua, and I suspect current to be lower than that. Does that seem like the way to do the current measurement? Living vicariously through my son's experiments at his work, Mikek

Fun fact: I've seen the neon sign transformer say 30ma, impedance protected. The transformer we're using is 9000V at 30ma. I measured the output impedance at 295kΩ.

9000V / 295kΩ = 30.5ma, that could not have come out better! :-)
Reply to
Lamont Cranston

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.