Shop Notes 5/26 – New HF Amplifier, Trials and Tribulations


I received a high-power HF amplifier so I can properly perform some larger-scale Tesla and Cosmic Induction Generator in the near future (and at the upcoming conference). So of course I started tuning things up and seeing how it handles.

Being a fancy high-tech box, it has a lot of automatic interlocks that protect the tubes and electronics from damage. High SWR, overheating, overvoltage, undervoltage, etc.

Unfortunately, it also seems to have a feature ‘gain too low’, that cuts off power when the amplifier is not putting out enough juice. This is a problem for Tesla/HV work, because we’re working with a lot of variables and dynamic components that can easily throw the tuning off. So I am basically unable to tune to match a Tesla transformer due to this, or at least without some modifications……

Immediately, I opened the amp and had a look inside to see what kind of bypass or options I had to mitigate the low-gain cutoff..

under the hood

After looking at the board and the schematics, I settled on two possible plans-of-action.

1 – The interlock system seems to be driven by a single relay at the end-stage of the amplifier. If I force this relay on, I should be able to force it to transmit in the interlock’ed state and maintain power while tuning+testing.

2 – I looked at the control board and found that it uses a standard PIC32 processor (specifically pic32mx460f512l-80i/pt)

I also found an ICSP programming header, as well as a common serial port in the back of the unit to flash firmware. Using my knowledge of PIC’s, I think I might be able to download and hack the firmware to disable this single interlock while leaving the rest of the system functioning normally….

PIC chip near the bottom, and a plain-jane 6-pin ICSP programming header right below it. Doesn’t get much simpler than this 😄

Over the holiday weekend, the plan is to attempt to dump the current running .bin files for the amplifier, decompile into hex, and attempt to reverse-engineer the section that handles the various interlocks in the unit.

I also grabbed a couple versions of firmware updates to play around with as well, in case the ICSP option falls flat.

I’ll probably also try a quick bypass of the interlock manually to confirm that the system will run OK in this mode.

I was definitely not planning on hacking the firmware of this thing within 24hrs of buying it, but such is life😂

Will report back once I have more data. If *nothing* works then I suppose I can try using a ferrite coupler to help stabilize the input to the Tesla Coils.