Help Me With My Cockroft Walton Multiplier

[nshin31302]

Hello. I have used 17 capacitor and 16 diodes with the designs from here: http://www.diyphysics.com/2012/02/09/d-i-y-250-kv-high-voltage-dc-power-supply-with-neat-trick-for-switching-polarity/. The diodes were 12 kV 350 mA microwave oven diodes from http://www.amazon.com/gp/product/B00JFOR95Q?psc=1&redirect=true&ref_=oh_aui_detailpage_o01_s00 while the capacitors were 20 kV 1nF ceramic capacitors from http://www.ebay.com/itm/10pcs-20KV-1nF-0-001uF-1000pF-102-High-Voltage-Ceramic-Capacitor-/300678663614?hash=item4601d84dbe. After soldering together the Cockroft Walton Multiplier, I submerged it in USP grade mineral oil to prevent corona discharge. Then I hooked it up to mains A.C. with a 1 amp fuse and a resistor to limit the current. But when I used a multimeter, it said 120 V. What might have gone wrong? Did I use the wrong components? Thanks in advance for replying.

[c4757p]

uh... where exactly did you use a multimeter?

[Fungus]

The diodes were 12 kV 350 mA
the capacitors were 20 kV 1nF ceramic capacitors from. After soldering together the Cockroft Walton Multiplier, I submerged it in USP grade mineral oil to prevent corona discharge. Then I hooked it up to mains A.C. with a 1 amp fuse and a resistor to limit the current. But when I used a multimeter, it said 120 V. What might have gone wrong? Did I use the wrong components? Thanks in advance for replying.

a) You don't need such high voltage components or mineral oil if you're only going to connect it to the mains.

b) It's a good idea to test it before you make it all oily.

c) 120V sounds suspiciously like mains voltage. Where/how did you measure that?

[tron9000]

  oh god! PLEASE don't measure the output of that circuit directly with a DMM, unless you know that DMM can take it!

[Fungus]

oh god! PLEASE don't measure the output of that circuit directly with a DMM, unless you know that DMM can take it!

Maybe that's the reason the DMM now only reads 120V 

[station240]

There is no way known his multimeter is rated to 5,440V, lets hope he's using a HV probe.

output voltage of multiplier = input voltage (peak) x 2 x number of stages
                                            = 170V x 2 x 16
                                           = 5,440V
reverse the equation, if you want a far safer 50v output, you need to input 1.5V peak (1V AC <- RMS).
Then you can test the thing outside it's oil bath.

[Alex Nikitin]

But when I used a multimeter, it said 120 V. What might have gone wrong? Did I use the wrong components? Thanks in advance for replying.

You need to think about either changing the hobby (before you kill yourself) or (at least) learning how to use, say, LTSpice simulation software. It should take ten minutes to simulate the circuit, would save you some time, money and wasted oil, and, perhaps, may give you a bit more insight on what the multimeter measures.

Cheers

Alex

[Alex Nikitin]

output voltage of multiplier = input voltage (peak) x 2 x number of stages
                                            = 170V x 2 x 16
                                           = 5,440V

Unfortunately (or fortunately for the TS) this equation is always incorrect in practice.

Cheers

Alex

[Fungus]

output voltage of multiplier = input voltage (peak) x 2 x number of stages
                                            = 170V x 2 x 16
                                           = 5,440V

Unfortunately (or fortunately for the TS) this equation is always incorrect in practice.

Yep. Cockroft Walton Multipliers look good on paper but in practice they're so feeble that maybe having a multimeter connected is preventing it from charging.

OP: If your plan is to make big exciting sparks then forget it. This won't work.

[bktemp]

output voltage of multiplier = input voltage (peak) x 2 x number of stages
                                            = 170V x 2 x 16
                                           = 5,440V

Unfortunately (or fortunately for the TS) this equation is always incorrect in practice.

The equation is correct, but only for no load. But in this case there is a mistake: He used 16 capacacitors, not 16 stages, so the output voltage is only half of that.
High voltage diodes also have a high forward voltage, probably in the range of 10V. The expected output voltage with no load should be around 2.5kV.
Since 1nF is way to small for mains frequency, the output impedance is very high and even connecting a DMM will reduce the output voltage considerably.
If the DMM was connected before switching the multiplier on, it may have survived.

[PA0PBZ]

I doubt that the 1nF capacitors will be very effective at 50Hz, normally you run these things at 10-20Khz.

[Alex Nikitin]

output voltage of multiplier = input voltage (peak) x 2 x number of stages
                                            = 170V x 2 x 16
                                           = 5,440V

Unfortunately (or fortunately for the TS) this equation is always incorrect in practice.

The equation is correct, but only for no load.

No, and you've just explained why:

High voltage diodes also have a high forward voltage

A small correction: all diodes have a forward drop voltage (and some leakage too, which in this particular case may be a problem as well  ).

Cheers

Alex

[c4757p]

A much safer idea, and more enjoyable for at least one of the involved parties.

[Kalvin]

There is also Dickinson charge pump ...

[station240]

Just in case he does manage to get this thing working.

This is what you are supposed to use to measure thousands of volts.
http://www.amazon.com/Fluke-80K-40-High-Voltage-Probe/dp/B000LDQ672
Rated to 40KV, 1:1000 ratio so 10KV reads as 10V on the meter.

Yes I was aware the multiplier needs a higher frequency that 60hz mains, but there are obvious safety issues with anything that does operate at a higher frequency.

[nshin31302]

So is the following what all the replies imply:

1. Either don't build the Cockroft Walton Multiplier
2. Or use a circuit simulator
3.  And/or use a high voltage probe

But you still have not answered how I could fix the Cockroft Walton Multiplier. Are the components the wrong types? Is the problem the multimeter? Or is it just human error? Should I discontinue the project?(If so I still want to know the electronics theory of why the Cockroft Walton Multiplier did not work.) Is it possible to make a high voltage probe? (Most high voltage probes cost a few hundred dollars.) I am not too interested in making huge sparks; however, I want to use the Cockroft Walton Multiplier as a constant high voltage D.C. power supply. Any helpful replies will be appreciated.

[PA0PBZ]

The answer is right above your post.

[dan3460]

I work on antique radios, were voltages are in the 200-400V, we are always extremity careful when working inside the radios and extra careful when the radio is connected. A shock from this voltage could kill you, but certainly will hurt you. With the voltages you are talking about, it will definitively kill you.
For the questions you are asking it doesn't seem that you have a lot of experience handling high voltages. Whatever you are trying to do I would suggest that you look for guidance from someone with experience that can see what you are doing.
You are asking questions here but you have not answered one crucial question, how and where are you using your meter to get 120V?
On top of that, you said that you want to make this thing as a constant high voltage supply, for what? If you are thinking on tubes, be aware that at those voltage many tubes are capable of producing X rays, that could be hazardous for your health.

[nshin31302]

So I will take your advice and discontinue this project. Instead I will build a Van de Graf.

[Howardlong]

On main page, this thread shows like this:

Re: Help Me With My Cock

There is also Dickinson charge pump ...

**Cough** 

[Alexei.Polkhanov]

There is no need for HV probe if it is for one time - just use a 1:100 or 1:1000 resistor divider like 1M:10K. I work with 1200-2000V circuits often and I always use divider built with two HV resistors that I got from broken Bertan HV supply. Input impedance of DMM should be like 10M so result should be close or adjusted accordingly.

[tron9000]

So I will take your advice and discontinue this project. Instead I will build a Van de Graf.

Would be "safer" - I've not heard of anyone being killed by a Van De Graf (except maybe an unfortunate dutch politician killed by someone called Van Der Graff), but at least the end at which you want to make the sparks will be somewhat isolated from the mains.

The problem I have with the Cockroft circuit your designing is the failure mode: what if one or all of the diodes or caps failed and you'd plugged it into the wall directly and you had one end in your hand, waving it round like a wand? You'd be pretty crispy! Seems an unlikely scenario, but its still more likely than impossible.

Also it would be a less dangerous introduction to electronics, I think all you'd need to build is a DC motor controller?

[Fungus]

So I will take your advice and discontinue this project. Instead I will build a Van de Graf.

It always helps if you tell people what you want to achieve, not just what you're doing.

If you want big sparks, a Cockroft Walton Multiplier doesn't work anyway.