The Bias King Pro modification that it desperately needs

[SoulFetish]

I got this as surplus from a store I used to work at for a fraction of it's retail price - (but still not cheap, by any means). I bought it with the hopes that it would make the bias procedure much more efficient and convenient. In particular, it had the ability to quickly toggle between two tubes see what kind of offset there was. So, it comes with two molded, flexible cable assemblies and 8-pin socket adapters, and is powered directly of the 6.3V heater supply. Great, right?
Almost.... ?
To do all that and leave out the ability to measure the plate to cathode voltage makes no sense to me. Particularly for the retail price they go for. In order to make any adjustment to the bias any tube, plate voltage is one of the 3 criteria that you have to know. So, the fact that a separate step is needed to take a manual reading of the plate voltage, almost renders the convenience moot and the device loose most of it's appeal.
I decided to take it home and see if there is a practical way to modify it to read both plate voltage and cathode current. I took some shots of the device here:









A couple of quick observations reveal how the circuit works simple enough, you can see how it switches between tubes reading the voltage drop across the current sense 1Ω resistors internal to this enclosure mounted one the smaller PCB.
But the device itself is ultimately a voltmeter, which is great because there's no reason that i shouldn't read a ¹/₁₀₀₀ or even ¹/_10,000 of the plate voltage.
But, here's the complication. They designed the device with the 1Ω resistor internal to the enclosure, rather that internal to the sockets assembly, and use an SPDT to select between the two tubes. This means that here is to easy or convenient solution for switching between the plate and cathode readings by simply adding another DPDT switch.
It can totally be done by adding a single switch to the enclosure, but it would require rebuilding or replacing the socket adapters and installing the plate voltage divider and 1Ω sense resistor internal to sockets. Then I could keep the shared ground connections by running 3 wires (ignoring the 6.3V power supply wires for the moment) from the sockets to the device. That way I could keep the shared ground, use a new DPDT switch installed to switch betweenb plates and cathodes, and rewire the existing switch to switch between the + voltages for the main board reading and display.
Oh, there is one more thing. The existing cable only has 4 conductors. That means ditching the existing sockets AND retractable molded cables and running 5 conductor cable instead. Or using some other way to power the device. But, I've spent the better part of the night trying to see if can affordably source that coiled cable with 5 conductors. Because, holy shit, there is no way I'm paying what they go for on the retail market. check out the prices for a 5 'footer and up on Mouser or Newark.
Luckily, I found a couple of sources where I can get 'em for around $12-$15
What do you guys think?

[Helmholtz]

Don't forget that connecting meter wires to power tube plates often causes HF oscillation and thus wrong DC readings. There are tried and tested countermeasures, but one needs to be aware.

Also make sure you have sufficient voltage rating. With signal or oscillation, peak plate voltages often considerably exceed 1000V. There's a reason why quality meter leads have heavy insulation.

[g1]

You're probably not looking for quick & dirty, but I'd be inclined to drill a hole so I could get a meter probe on pin 3 of the socket adapter.

[SoulFetish]

Don't forget that connecting meter wires to power tube plates often causes HF oscillation and thus wrong DC readings. There are tried and tested countermeasures, but one needs to be aware.

Right, there were a couple of things I was thinking about to minimize this. First, I was thinking about installing a MOV across the plate to cathode pins to protect against overvoltage (figure somewhere between 800-1000V rating). Given the MOVs internal capacitance, I had hypothesized that the shunt capacitance would be an advantage in this regard. I have doubts about being able to fit a MOV, as well as 2 resistors inside the socket, so my second thought was an small mounted to the plate pin (between the plate and R1 of the voltage divider).
What do you think?

Also make sure you have sufficient voltage rating. With signal or oscillation, peak plate voltages often considerably exceed 1000V. There's a reason why quality meter leads have heavy insulation.

Absolutely! In fact, because R1 in the divider will basically drop the entire voltage across it, I wanted to make sure to sufficiently limit the current for measurements up to 1000VDC so not to exceed the rated dissipation of the resistor as well. In reality, I kept it under half the rating.
The Vishay HVR37 series looks like a good choice here: 500mW/3.7kV rating
hvr2537.pdf

Newark has overstock pricing on 4.7M & 3.3M values
4.7MΩ Metal Film
3.3MΩ Metal Film

[SoulFetish]

I also realized that I don't need to replace the existing cables either. If you look at the orange and yellow wires which supply the 6.3VAC power for the supply, they only use one of the tubes for that (the one which has the white wrap near the socket). I really only need one of the tubes to get the plate voltage since they will only differ by a small % in a working amplifier. So, I can use the available red or yellow wires to take a voltage reading at the junction of the divider. But I do need to build another test socket adapter.
I mean....I have to now anyway, because I completely destroyed it trying to disassemble it.

this is a simplified schematic/diagram of the device. The only change will be the addition of SPDT 2.

[nickb]

You can lose two of the wires to the probe by using a battery instead of the heater supply. I see there is a transformer in the control box to provide isolation and possible a step up to power the meter. Maybe a little voltage regulator too. Fire it up, measure the voltage and current to the meter. I'd expect it to be in the 5-12V range with current in the few mA. So, given intermittent use and an on/off switch, the battery/-ies will last for ages.

[SoulFetish]

You can lose two of the wires to the probe by using a battery instead of the heater supply. I see there is a transformer in the control box to provide isolation and possible a step up to power the meter. Maybe a little voltage regulator too. Fire it up, measure the voltage and current to the meter. I'd expect it to be in the 5-12V range with current in the few mA. So, given intermittent use and an on/off switch, the battery/-ies will last for ages.

Yeah, I thought about maybe using batteries to power it up. Its is convenient not having to worry about turning it on and off, or batteries though.
One thing that concerns me about powering it from the heater is the potential of a tube shorting and damaging the unit. The other idea I was tossing around was maybe adding a USB port and a switching regulator for the DC supply. It's rare to enter a room at this point without at least one USB charger within reach

[Helmholtz]

Given the MOVs internal capacitance, I had hypothesized that the shunt capacitance would be an advantage in this regard.

I wouldn't rely on the MOV's capacitance. This will just add to the self-capacitance of the OT primary of several hundred pF. As a result the amp might just oscillate at a somewhat lower frequency.

[SoulFetish]

I finished up with the mod to read plate voltage as well as cathode voltage, but your suspicious (and my fears) were confirmed about oscillations.

I went through 2 socket adapters trying to secure the cable with miniature Heyco strain reliefs. This would have been more secure and ideal, but ultimately there was a problem of fitting the cap on, and breaking both attempts lead to me using a grommet and zip tie. Sometimes ideal doesn't end up being practical, and "good enough" wins the day.





However, it's gotta get opened up again to suppress the oscillations. I'm thinking ferrite beads on the plate, grid and screen pins (hopefully I can fit them). I'm guessing long runs of unshielded wire are probably the issue here.
1. - Do you guys think this is a good solution?
2. - What value inductance would you suggest?

[nosaj]

I finished up with the mod to read plate voltage as well as cathode voltage, but your suspicious (and my fears) were confirmed about oscillations.

I went through 2 socket adapters trying to secure the cable with miniature Heyco strain reliefs. This would have been more secure and ideal, but ultimately there was a problem of fitting the cap on, and breaking both attempts lead to me using a grommet and zip tie. Sometimes ideal doesn't end up being practical, and "good enough" wins the day.

[ATTACH=CONFIG]57427[/ATTACH] [ATTACH=CONFIG]57428[/ATTACH]

[ATTACH=CONFIG]57429[/ATTACH]

However, it's gotta get opened up again to suppress the oscillations. I'm thinking ferrite beads on the plate, grid and screen pins (hopefully I can fit them). I'm guessing long runs of unshielded wire are probably the issue here.
1. - Do you guys think this is a good solution?
2. - What value inductance would you suggest?

Is there a problem with using shielded wire? just ground one end and not the other like on a signal cable? Granted you loose the coily cable.

nosaj

[The Dude]

Hard to say EXACTLY what's causing the oscillation, but you might try a small value cap from plate to screen inside the sockets.

[SoulFetish]

Is there a problem with using shielded wire? just ground one end and not the other like on a signal cable? Granted you loose the coily cable.

nosaj

Right, I was just going to say that the original cable is not shielded. I would like to keep the original cable. There is the reason of vanity, of course, but it is nice and flexible as well.

[SoulFetish]

Hard to say EXACTLY what's causing the oscillation, but you might try a small value cap from plate to screen inside the sockets.

yeah, plus, that's something I could test by tack-soldering onto an amp socket to test out before I open it

[nosaj]

Right, I was just ghttps://www.amazon.com/MC-100-Replacement-Coiled-Microphone-Cord/dp/B07G9P4CB4 oing to say that the original cable is not shielded. I would like to keep the original cable. There is the reason of vanity, of course, but it is nice and flexible as well.

Do they make coily microphone cable?
they do
https://www.amazon.com/MC-100-Replac.../dp/B07G9P4CB4

nosaj

[Helmholtz]

However, it's gotta get opened up again to suppress the oscillations. I'm thinking ferrite beads on the plate, grid and screen pins (hopefully I can fit them). I'm guessing long runs of unshielded wire are probably the issue here.
1. - Do you guys think this is a good solution?
2. - What value inductance would you suggest?

Without knowing the frequencies involved and understanding the positive feedback path, it's hardly possible to recommend specific measures without testing.
I tend to think that using shielded cable (withstand voltage!) is more promising than ferrite beads, as these mainly damp HF currents.

But it might not work with all amps. I generally recommend to pull the PI tube before taking plate measurements.