High voltage diode

[Carl Smith]

I recently acquired a high voltage diode that has a peak voltage of 80 kv at 200 mA. I am using a 1:4 ratio and 1:120 ratio potential transformer to feed in 57.6 kv RMS. Are high voltage diodes rated at RMS or peak voltage since the peak voltage of my power supply is 81.5 kv. I am also using a 1440 watt at 120 volt water heater element to limit the current in my system.

[Chris Bradley]

Carl Smith wrote:
> I recently acquired a high voltage diode that has a peak voltage of 80 kv at 200 mA. I am using a 1:4 ratio and 1:120 ratio potential transformer to feed in 57.6 kv RMS.
Are we to presume you are attempting to use two transformers in series to take 120V mains to 57kV? What sort of transformers are these? Are they really capable of 80kV?


>Are high voltage diodes rated at RMS or peak voltage since the peak voltage of my power supply is 81.5 kv.
How do you know the peak volts is 81kV?


>I am also using a 1440 watt at 120 volt water heater element to limit the current in my system.
Not sure what you think a 10 Ohm resistance will achieve. 10 ohms to a high voltage mA current is nothing more than a straight wire!

[Carl Smith]

I am using two transformers. The first has a 120 primary and 480 secondary. The second is a 120 primary and a 14400 volt secondary. They are both potential transformers however I am feeding the secondary of the first transformer into the primary of the first transformer. The 10 ohm resistor is in series with the primary of the first transformer which should limit the current provided to both transformers. I did some research on the type of potential transformer that I have. Like its name plate suggest it should deliver 14.k kv output with 120 volts input. However I am unable to confirm its output because I am currently unable to measure the output voltage. I am in the process of getting the supplies so I can build a wiring diagram similar to Richard Hulls NST wiring diagram however I will acquire a beefier resistor. However I calculated that if my peak RMS voltage is 57.6 kv then my peak voltage should be 81 kv because 57.6*(2^.5) = 81kv.

[Tyler Christensen]

You can to d 120:480 -> 120:14400. You'll saturate the core. A transformer core has a maximum volts/turn, and you can be sure it's not going to be 4x under-rated. If lucky, you might find one that's 1.25-1.5x under-rated primary. So unfortunately, your plan is flawed and you can't get more than 14.4kv with this transformer.

Keep in mind a 10 ohm resistor in series with the parallel will drop V=IR, so you'll have a linear secondary voltage drop as power increases.

[Carl Smith]

Thanks

[Chris Bradley]

Carl, notwithstanding the basic electrical issues (as raised), if you think it through, consider whether your 14400V transformer is going to be designed to withstand 81000V? Why would designers not simply string transformers together to achieve the desired output, if this was an OK way to do it.

I think you would get some performance out of this arrangement, if you had a 480V transformer of a sufficient power to drive the HV transformer primary into saturation and suffer the power losses - I guess there would be some margin of both isolation and saturation tolerance - but you'd probably be burning it up at twice the voltage. But you might as well run a voltage doubling circuit if that was all you were after.

[John Taylor]

Carl,
I have a potential transformer that I use in my power supply, and it has a nameplate specifically stating what the breakdown voltage is (mine is rated at 500kv). These things are connected to lines that will have substantial voltage spikes at times and (at least all I have seen) are specifically overdesigned as far as insulation goes. I control it by use use of a variac feeding a transformer that was intended to step 240v to 480v. Into this I feed 0-280v this gives me 0-560 volts into the 1:200 terminals of the potential transformer for a maximum voltage (theoretically) of 112kv output. Of course, these are RMS ratings and the peak voltage is what will matter as far as breakdown goes.

I currently have a rectifier rated at 80kv Peak Reverse Voltage and so have only run this setup to around 40kv RMS, but have run it up to the maximum setting on the variac when it had no load on it (nothing connected to the hi-voltage connectors) with no ill effects. Of course at these voltages, a breakdown would probably be pretty spectacular!

Just my 2 cents...

John Taylor

[Chris Bradley]

John Taylor wrote:
> Into this I feed 0-280v this gives me 0-560 volts into the 1:200 terminals of the potential transformer for a maximum voltage (theoretically) of 112kv output.


As you raise the volts/turn in a winding, you drive higher magnetic fields in the core. This is lossy (unless it's an RF air core), and is very non-linear for most materials. A typical mass-produced commercial transformer will be designed for a given input voltage to squeeze every last efficiency point out of the available core (because less core, less cost). There will be a little latitude, but I bet sometime before twice design input voltage you will be mostly just pumping heat into the core. You may well get some weird inductive behaviours and high unloaded volts appearing, but try to draw just a mA or two of such a setup with commonly available transformers and I suspect you will find you've achieved very little.

Of course, you may have some atypical transformer that works nominally at very low core flux densities so you've got that latitude to do this. Have you driven an actual load at 40kV? I'd be interested to hear what the load was, how you measured it, transformer spec, etc..

[Rich Feldman]

Carl and John,
Here is a small-scale teaching experiment about core saturation.
It may save you some time and expense, before you try to double-volt your potential transformer (which is designed to tolerate saturation by overvoltage spikes, not steady-state operation). Aside from secondary insulation issues, 2x or 4x voltage on primary has consequences similar to DC voltage on primary.

Take a small 120V-input transformer, such as one made for low voltage lighting / lawn sprinkler control / doorbell etc. Know its rated maximum primary current. Then with its secondary unloaded, measure and chart the primary current as you progressively increase the primary voltage beyond the nameplate rating.
If you are in the US and have split 240V electricity, but no 240V variac,
you could cover the range of interest with a 120V variac on one side of a 240V circuit.

Here's another safe way to explore the hockey-stick-shaped curve of no-load current, with its "knee" at a voltage typically not far above the nameplate value. Take a 120V variac, set near full voltage. Apply 120V to the "output". Now measure the current as you gradually turn the control "down", increasing the volts per turn. You can monitor the voltage on the fixed-tap "input" circuit. Be sure you have an appropriate fuse to protect the brush and short winding.

A transformer rated for 50 Hz will tolerate 20% extra voltage (w/r/t saturation) when run at 60 Hz.

[Richard Hull]

Rich is right, I operate my 220 volt primary X-ray transformer via variac and can push it to about 250 volts input before the current, due to saturation, piles in like gang busters. I get about an extra 5kv out of the transformer. Most transformers can easily handle this 5-10% push over rated nameplate for the short operating periods of a fusor.

Once you fully saturate all you get added on for more volts input is a bigger electrical bill and a possible burned out transformer. However, most all 60hz transformers can be pushed safely into saturation yielding extra volts out, provided you monitor the primary current.

Richard Hull

[AFW]

Don't forget that, if the PIV of your diode is 81kV, It is only guaranteed for a rectified output of up to 40.5 kV, as the PIV it experiences will be twice the output voltage.
Tony Webb