Anode current metering in an HV supply. Is my pal trying to kill me?

[Chris Wilson]

I have a friend who says I have done the plate amp meter install
wrong. I attach a back of fag packet diagram showing both his method
and my method of measuring plate current, Method 1 and Method 2. Which
is correct please?  Using Method 2 if the negative return line was disconnected
from the meter in some way I would be seeing 5kV this thing produces on this line, right? Not good!
He says method 2 is correct. I think he's wrong and maybe trying to kill me! Cheers.
All my reading says my method (Method 1) is correct and the 33 Ohm 100 watt resistor
and the 2 6A10 diodes should hold everything on the negative rail close to ground
potential including the  meter and possibly its case if the insulation failed, and the meter wiring.

[ElectroIrradiator]

Method 2 looks most correct, though neither is likely to kill you. What will actually work, will depend on what the rest of the amp+PSU schematic looks like. The trick is to ensure the anode current meter only measures just that, and for instance not the sum of anode plus screen current.

Your sketch do have some nasty errors, though. The 6A10 diodes should be used in anti-parallel pairs, not in a series configuration. The idea is that they act as bidirectional 0.7V zener diodes with a hu-u-u-uge surge current capability, in case the tube flashes over, or whatever. The 33 ohm resistor is there to ensure that even if the diodes dies horribly, the negative HT rail doesn't suddenly float (which can be quite deadly, if the PSU is in a chassis separated from the amp proper).

So anti-parallel 6A10 diode pairs directly across the meter and the 33 ohm resistor. Location of the diodes protecting the meter should be across the connecting terminals. IE. they are part of the PSU circuit, and not the range setting resistors. The total voltage across the anode meter circuit ought to be below 0.7V, so this shouldn't be an issue.

[Chris Wilson]

OK, thanks, I will reverse parallel the diodes across the safety 33 Ohm resistor. The meter itself also has 2 off 6A10's as a reversed pair across the terminals, as well as a shunt resistor of well under 1 Ohm, and a cap of 0.1uF across the terminals. A trimmer pot is in one of the meter leads to set the calibration. Or do I not need 2 pairs of reversed parallel diodes?

[ElectroIrradiator]

You should put an anti-parallel diode pair across any component or sub-circuit in the negative supply lead, regardless of number. The only noticeable resistance from filter cap battery to the tube should be the surge limiting resistor in the positive supply line, which you call the glitch array.

So if you need the meter at location 2, you 'cut the wire', place an anti-parallel diode pair across the gap, then connect metering sub-circuit with shunts, series R and whatnot, in parallel to the diode pair. Then add another pair across the 33 ohm resistor. Frequently one of the diodes in the pair will never see any surge current, regardless of type of fault condition. Yet given the cost of a few extra diodes, then it is frequently easier to go 'belt-and-suspenders', compared to trying to decipher exactly what fault currents are potentially likely to occur.

Note that I would not recommend using any series/parallel combination of smaller resistors for the glitch resistor. The whole idea is that this resistor will be able to survive having the full filter cap battery connected across it, when - not if - the tube flashes over. As the glitch resistor has a resistance much higher than the spark inside the tube, almost the full anode potential develops across the glitch resistor, extinguishing the spark.

This trick relies heavily on the glitch resistor not flashing over, once it sees 3+ KV. I'd use a single, ridiculous glass/ceramic 100W resistor, as they can take the abuse. The actual power rating is less important, but you want to make sure the resistance wire and everything else doesn't give, once the manure hits the ventilation device. So choosing a mechanically rugged resistor is important.

The same is true for the 33 ohm 'ground lift' resistor. That one is your last defense against potentially being wired in series with the HT rail, with only the tube as a current limiter. So you want to make very certain it never goes open circuit, no matter what.

[Chris Wilson]

OK, I have decided to go the full Monte and have bought 2 stud rectifiers, cathode and anode reversed, 1.6kV and 40 amp with very high surge, for across the 33 Ohm 100 Watt safety resistor. I will replace the glitch array with a single 100 watter which I have. Thanks again!

[Paul Price]

If the plate current is, for example 100mA, then the drop across the 33 ohm resistor will be 3.3V without the diode protection, and with the single or double diode protection the plate current reported by the meter will be quite low yet quite large in error.

A better solution would be quan. 6 of 1N4002 diodes in series. Would do the protection job well and able to withstand 100's of amp surges and recover room enough to hide your lunch inside the equipment's enclosure that the 100-W 33 ohm resistor was taking up.

[Chris Wilson]

Hmmm, I see. Plate current will be circa 1.5 amps. Thanks!

[ElectroIrradiator]

If any plate current flows through the 33 ohm ground lift resistor, then you are doing it wrong. The anode current should flow through the meter at position 2, hopefully.

Also, it is quite possibly suicidally stupid to leave the 33 ohm resistor out, even though it actually shouldn't do anything during normal operation. As I already mentioned, then its purpose is to be your last line of defense against killing yourself, should the diodes in parallel with it (usually no current here either) somehow cease to work (for any reason).

I think you need to post the complete schematic of PSU+amp if you want me to explain further here. I am sort of guessing what the circuit looks like at the amp end, and that part is also important, safety wise.

If the metering and various voltages sources for you tetrode amp is wired the 'usual way', then it is possible to develop the full anode potential between PSU and amp chassis, with only the tube wired in series as a current limiter. This usually requires something like a broken wire (or component) to actually happen, but is obviously very, very bad if it does. So you want to avoid this possibility at all cost. Literally.

Happy 1N4002 soldering...

[Chris Wilson]

OK, will try and get a full schematic together.  Thanks. It started life as a Henry 2kW RF generator for plasma work, with a 3 phase HV supply in the one cabinet as the exciter, amp and control boards. I built a single phase HV supply in its own cabinet, the amp itself and control board are in the original cabinet. Both share a seriously hefty ground strap. The HV supply voltage and current are measured at both the new HV supply cabinet, and the remote mounted, on its original Henry umbilical, Henry control box, again, a separate cabinet in the shack.  So effectively I have divorced the HV supply from the two original cabinets, and home brewed a single phase HV supply. I'll see what I can do about a schematic. It won't be pretty, sorry. It's a triode though, not a tetrode, it's a 3CX3000A7. Thanks again.

[geraldjhg]

method 1 is the correct but the 33 ohns resistor and diodes has to be of good make and
LOCATED in the ground circuit near the power negative- no accidental disconection
the meter then can go anywhere

[Chris Wilson]

Thanks Gerald, not been able to up a schematic as my PC suffered a total hard drive failure the other day. Luckily all mail and important stuff was backed up EXCEPT for amateur radio related items Been recreating links blah blah. I have now sorted the meter wiring as you and others have described. No one has spotted my brain fade in drawing the fag packet schematic though! I have drawn the filter cap bank in PARALLEL. Of course, they are really in series Thanks again.