VVR issue - voltage drops under load

[Zozobra]

So I decided to add a VVR circuit to my Shiva/Hot Cat build and it behaves as expected under no load conditions but then drops 100V with the pot at 100% rotation when I put the power tubes in.
I have also used Zeners in the tail of the bridge rectifier to drop the supply to the general desired level. Could this cause a problem?

Here is a schematic of the power supply with the voltages under load and the pot at 100% rotation.





What have I done wrong? I think something isn't right as it doesn't seem like I'm getting the Zener voltage drop as I remember the plate node being around 360-370VDC before I added the VVR but this may be a red herring as I didn't note the wall supply...

[Dave H]

So I decided to add a VVR circuit to my Shiva/Hot Cat build and it behaves as expected under no load conditions but then drops 100V with the pot at 100% rotation when I put the power tubes in.

The B+ will drop when the power tubes are put in but you also have 68V across the 100k resistor so there is 0.68mA flowing either into the FET gate or leaking through the 12V zener. I suspect the zener. Check that it's not reversed.

[Zozobra]

Thanks. I'll swap out the Zener to see if that is the culprit. I'd expect the B+ to drop down to 360-380 with the tubes in as that was the ballpark they were operating in before. If the leaky zener is dropping another 68V then that tallies with the plate voltages I'm seeing.

[Zozobra]

I swapped out the zener which didn't help. I'm pretty sure I've got it the right way around so maybe the MOSFET is borked. I have another 14nk60 so I'll try that. I also have a few irf840s if they're suitable.

[J M Fahey]

1) Please label *all* parts (R1 - R2 - ... C1 - C2 - ... Q1 - .... etc.) and repost.

2) to start checking/thinking: how come you have 396V on D1 cathode but can only reach 368V on Vr1 cursor?

If so, you have a bad or very poor quality pot.

[Zozobra]

I've updated the schematic with part numbers for clarity as suggested.

Before I set out for work this morning I had time to quickly snip out the D5 Zener and measure the output voltage. This fixed the issue, however, the Zener was in the orientation shown in the schematic so I'm not sure what the problem is. Could the inrush to C4 be damaging it?

As for the pot; it is an Omeg pot rated for 500V so the quality should be okay but it may be bad. With D5 removed the drop from D1 cathode to VR1 wiper is reduced to around 10V or so.

[R.G.]

Before I set out for work this morning I had time to quickly snip out the D5 Zener and measure the output voltage. This fixed the issue, however, the Zener was in the orientation shown in the schematic so I'm not sure what the problem is. Could the inrush to C4 be damaging it?

Given that, the issue is that the current through R7 is causing a voltage drop that increases until the zener conducts. The zener is there to protect the gate of the MOSFET, and its anode should connect to the source of the MOSFET, not to the far end of R7. When you reinsert a zener, connect it with cathode to the MOSFET gate, and anode to the MOSFET source.

For RF stability, insert a 100R to 1K resistor in series with the MOSFET gate, as close to the MOSFET gate pin as you can possibly get it.

As for the pot; it is an Omeg pot rated for 500V so the quality should be okay but it may be bad. With D5 removed the drop from D1 cathode to VR1 wiper is reduced to around 10V or so.

That's because the zener is no longer letting current through.

The point of the circuit is to let the MOSFET meter the current through, not let it do an end-run through the zener.

By the way, that is not a voltage variable resistor circuit - it's a source follower fed a DC reference voltage from the pot wiper; a simple voltage regulator.

[Wombaticus]

Given that, the issue is that the current through R7 is causing a voltage drop that increases until the zener conducts. The zener is there to protect the gate of the MOSFET, and its anode should connect to the source of the MOSFET, not to the far end of R7. When you reinsert a zener, connect it with cathode to the MOSFET gate, and anode to the MOSFET source.

No. The Zener here is *not* meant to protect the gate (14NK60 series has built-in gate protection). In this circuit, D5 and R7 are meant to be a crude current limiting circuit. Implementing current limiting this way is slightly iffy because of the somewhat variable Vgs drop of MOSFETs, but typically this will be about 4 or 4.5V. In this case the Zener will conduct and current limiting will kick in when the voltage across the R7 sense resistor exceeds about 12-4.5 = 7.5V. So with a 20R resistor one would expect current limiting, roughly, to start at about 7.5V / 20R = 375mA. Give me sec, let me check for other issues here.


EDIT: One other thing to consider here is that this form of current limiting is really only suitable for protection against transient shorts to ground. If the circuit goes into persistent current limiting, the pass device will just sit there and fry.

[Zozobra]

I'll add a resistor to the gate when I rebuild the circuit.

Yup, it's there for current limiting as described in Merlins power supply book. IIRC the MOSFET Vgs was 4.1V so it should limit at around 395mA. The amp shouldn't pull anywhere near that outside of a fault condition and certainly not with the 6V6s that are currently installed.

[R.G.]

No. The Zener here is *not* meant to protect the gate (14NK60 series has built-in gate protection). In this circuit, D5 and R7 are meant to be a crude current limiting circuit. Implementing current limiting this way is slightly iffy because of the somewhat variable Vgs drop of MOSFETs, but typically this will be about 4 or 4.5V. ...
EDIT: One other thing to consider here is that this form of current limiting is really only suitable for protection against transient shorts to ground. If the circuit goes into persistent current limiting, the pass device will just sit there and fry.

Ah. I just misconsidered it. That is indeed a very crude way to do current limiting. There are much better ways that don't destroy things or rely on the imponderables of the zener and MOSFET.

A more predictable way to do this is to use a current sampling resistor there where the 20R resistor is, and to use an ordinary NPN transistor with its base to the MOSFET source and its emitter to the outside end of the sampling resistor. The collector connects to the gate circuit of the MOSFET to steal gate voltage. The NPN turns on somewhere between 0.45 and 0.6V reliably, and it eats all the gate drive by shunting the current on the MOSFET gate to the output. With any given NPN, the threshold is predictable and there's a lot less wobble in guessing the MOSFET turn-on voltage. This makes the 20 ohm resistor value change to something like 1-2 ohms, and the threshold, once adjusted for the exact NPN, is quite reliable. I used this for current limiting to solve overcurrent issues when switching out of standby.

I didn't originate this, I adapted it from a circuit in the ARRL handbook from about 197x? and it was old then.

This addition just makes it predictable, it doesn't keep the output device from frying. But there are other ways to do that.

Simplicity is good - however, as attributed to Albert Einstein, everything should be as simple as possible - but no simpler.

[R.G.]

The problem there is that you can't predict the Vth of a power MOSFET that closely. It may or may not be 4.1V - could be from 3.5? to 5.? for the same part numbered device.

You might think about using the NPN + 1.x ohm resistor to replace the zener. It's a much sharper and more predictable cutoff.

A variation is to use a 500R trimmer across a ???2.2R??? resistor for sampling current and tie the base of the NPN to the wiper. You can then twiddle the trimmer and adjust to any threshold you like.

[J M Fahey]

The unexplained/unjustified ~30V drop from wiper to top end of Vr1 means you have a poor quality track there with a residual resistance of ~45k, not unheard of but very high , specially here.
Might pass unnoticed if used as a volume or tone pot.

So please, with everything unplugged (preferrably with VR1 unmounted, just sitting alone on the table), set wiper to 10, measure residual resistance frpm wiper to hot lug; then set pot to 0, measure wiper to ground terminal, post results.

0 o0r 10 do not mean just a number on a scale but shaft rotated until full stop either end.

I have actually made pots at a factory and suspect what's going on, worst case get a couple different brands.

[Zozobra]

I'll whip the pot out later and measure it to see if it is bad

[R.G.]

This may not be your problem, but then it might be, too.

Many pots are less tolerant of current through the wiper than current through the length of the pot. The wiper works by physically touching the element, and there is additional resistance just from the contact. Some pots actually specify the wiper-to-element resistance. A lot of current can burn one spot on the element where the wiper sits, or burn the wiper itself, or both.

Since your pot has had (relatively) lots of current for its wiper, it may be that the wiper has a problem now. Maybe not, but it's worth checking.

[Wombaticus]

Since your pot has had (relatively) lots of current for its wiper, it may be that the wiper has a problem now. Maybe not, but it's worth checking.

Under normal operating conditions, there shouldn't be more than maybe 500 microamps through the whole pot, and maybe 200uA through the wiper, tops. But while it's all just slightly fuzzy owing to the variable Vth issue R.G. and I have pointed out, it's worth noting that if the output goes dead short (or even just enough for current limiting) the circuit as drawn will pull significantly more through the wiper, maybe up to nearly 4mA. Depending on the pot position this could easily be enough to blow through the 1/4W limit of most pots. Of course, with persistent current limiting you'll have other problems as well owing to the pass device dissipation. Nevertheless, you can make this aspect of the circuit more bulletproof by changing R6 from 100K to something more like 270K, which should keep a 1/4W pot pretty much unconditionally safe. If you're determined to stick with a Zener, you should then choose one with a reasonably low Zener current..But R.G.'s suggestion of an NPN transistor there is time-tested. Note that it doesn't even have to be an especially high voltage device - a 2N3904 (Vcbo = 60V) is plenty adequate.