Load testing HV power supplies, and HV transformers (safely....)?

[Chris Wilson]

I am in the midst of making a couple of HV power supplies for RF linear amplifiers..
I would like to know if there's a safe means of testing these under load,
without using the RF deck and an expensive tube to put a load on?

One supply is hopefully going to give about 3800V DC at about .8 of an
Amp, after full wave rectification and capacitive smoothing,
the other about 5kV at about 1.8 Amps. I would also like to be
able to load a transformer's AC output itself up to see how it reacts
under load. Has anyone "done this at home" and lived to share the
story? Thanks.

[Rerouter]

my first thought would be a HV bulb, but 10KW of light is not somethign you want to share a room with,

as that is a $#!%-load of power its hard to reccomend much more than serious passives as most conventional things cannot be reworked for those kind of voltages,

also i pray your wearing proper HV stuff, for 2A the smoothing capacitance would easily fry most people.

[SeanB]

You probably will be best by using a water dummy load, a large plastic bath ( well insulated) and 2 electrodes at each end, with some salt added to increase conductivity.  You will need to have 2 baths, one of around 4k7 resistance and one of 2k7. Both will get hot, and fast. Place the dummy load far away when using it, and use a lot of care.

[ejeffrey]

Electrolysis of salt water will evolve large amounts of chlorine gas, hydrogen gas, oxygen gas, along with heat, while leaving behind sodium hydroxide.  Poisonous, explosive, and corrosive!  That is assuming the electrodes don't immediately disintegrate.  The conductivity of the solution will depend on the bath temperature, so it will change a fair bit as you heat up the bath.

Done carefully and outside, this could work.  Still, if at all possible, I would suggest a different approach.  10 kW is not too much for a normal resistive heater.  Off the shelf heating elements won't be designed to operate at 5kV but you could connect several lower powered 230 volt heaters in series, or wire one yourself from a whole bunch of nichrome wire.

[Chris Wilson]

Not too sure on the bath, rather cumberesome, and operating this (heavy() gear outside in a typical UK Autumn is going to be problematic. Someone sent me this link on another forum:

http://gs35b.com/hvload/index.html

I like that idea, but higher wattage resistors, less in number, would appeal more than that modern art / bird cage like structure

Thanks for the ideas so far.

[ptricks]

 That is seriously high power levels and learning off an internet forum how to handle it is a bad idea. I have seen people hurt by HV because they thought latex rubber gloves, the type used for cleaning would protect them, others thought vinyl electrical tape would insulate them . It is extremely hazardous and little things like forgetting to take off your watch while near a wire can kill without ever touching.  Things like using the wrong load can destroy the equipment due to the surge of current that occurs when the load is disconnected, so you have to be sure to drain off the load and any surges and not just cut power if anything wire wound is involved like high power resistors.

I would contact your local electric utility and ask to speak to some of the lineman. They can point you to what they use to test their systems in the field and also any precautions you can take.

[Monkeh]

others thought vinyl electrical tape would insulate them

It will. Up to about 6kV per layer. But it has to be applied correctly, not sloppily.

[vk6zgo]

I am in the midst of making a couple of HV power supplies for RF linear amplifiers..
I would like to know if there's a safe means of testing these under load,
without using the RF deck and an expensive tube to put a load on?

One supply is hopefully going to give about 3800V DC at about .8 of an
Amp, after full wave rectification and capacitive smoothing,
the other about 5kV at about 1.8 Amps. I would also like to be
able to load a transformer's AC output itself up to see how it reacts
under load. Has anyone "done this at home" and lived to share the
story? Thanks.

I've spent most of my life working on high powered transmitters,& my suggestion is to trust the transformer specs if you know them.

We never found the need for a test load for these type of supplies.
The main things to look for are high voltage arc-overs,etc,rather than current capacity.

Trying to make a load for these sort of power & voltage levels is a nightmare,and to do it properly would cost a lot more than the "expensive tube".
After all,the thing is unlikely to produce voltages in excess of the tube ratings,& if it falls short under load,that won't  hurt the tube.

If you don't trust,or don't know the transformer specs,you can probably get them tested by a transformer rewinding service,but it will probably be costly.

[NiHaoMike]

Maybe some microwave ovens modified to accept high voltage DC?

[Chris Wilson]

Have decided it's probably a lot safer to assemble the PSU and see how it loads up against the valve in the RF generator. Anything else I might be able to kludge up at home sounds a bit risky Thanks for the ideas though everyone!

[SteveyG]

10 kW is not too much for a normal resistive heater.  Off the shelf heating elements won't be designed to operate at 5kV but you could connect several lower powered 230 volt heaters in series, or wire one yourself from a whole bunch of nichrome wire.

We've just done a project at work for a 3.3kV 15kW motor driving a vibration plate, but used resistive elements for early testing of the SMPSU Lots of heat...

[Alana]

When i was doing my job experience we had some electrofilter assembly to be tested. 100kV, several kilowatts of output power.

Guys had some problems with making artificial load for this device. Whenever they assembled something HV arced across and burned plastic boards that held power resistor network together. Eventually they ended  up with something like this [pic]
Power resistors looked quite like those , i believe they were 1kohm

[Chris Wilson]

It gives the correct voltage OK, so half way there! Thanks for the further info and diagrarms

[SeanB]

To test the transformer under load connect the secondary to the mains and load the primary till you get the rated input current flowing in it. This will load the copper to the right power point ( but at a low voltage) and allow you to test the core and windings safely.

[Chris Wilson]

Hi Sean, thanks for replying. It's not so much that I doubt the windings, it's more that I want to know how much current I can safely pull from it at say 50% duty cycle, with or without any forced cooling. I am hoping with a full wave bridge rectifies, as I have built, with a capacative input filter, I might be able to get 1.6 / 1.8 amps at a 50% duty rate. I would really like to test tha, but it seems pretty scary given my very limited resources