Bias Failsafe AB763

[joesatch]

How would i modify this bias circuit for failsafe in case the pot malfunctions my tubes wouldn't get cooked?

[sluckey]

Put a 2.2M resistor between the pot left and center lug.

[pdf64]

A bias failsafe seems a good idea. Merlin’s guideline is for its value to be about 10x to pot value, so as per sluckey’s instruction, but around 220k.

http://www.valvewizard.co.uk/bias.html

[sluckey]

A bias failsafe seems a good idea. Merlin’s guideline is for its value to be about 10x to pot value, so as per sluckey’s instruction, but around 220k.

Yes, that's another OPINION.

[pdf64]

Yes, that's another OPINION.

Since Professor Ohm snatched claim to the only law around these parts, opinions are all we’ve got 

[sluckey]

When the pot wiper fails, the pot value becomes irrelevant. You just need a resistor that's big enough to not affect the bias adjustment while the pot is working properly, yet still pass full bias voltage to the tube grid. 2.2M definitely gets the job done.

[pdf64]

When the pot wiper fails, the pot value becomes irrelevant. You just need a resistor that's big enough to not affect the bias adjustment while the pot is working properly ...

Hence I think the suggestion for the safety back up resistor value to be about 10x the pot's track.

yet still pass full bias voltage to the tube grid. 2.2M definitely gets the job done.

That would seem to assume 0 grid current? From my (limited) observations, at (AB) idle, it tends to be in the range 0.1 - 1uA, so with a pair of valves, 2M2 should only lose a couple of volts from the available bias.
So yes, at idle it should be fine, even though it's over an order of magnitude above the typical grid circuit resistance limit of beam pentodes.
However, my understanding is that output valve control grid current may increase as the overall valve dissipation increases (hence the grid circuit resistance limit).
So if the amp is used with such a high grid circuit resistance, there seems a potential for a process of an ever increasing spiral upwards of anode current to initiate

[sluckey]

Same idea applies to the LarMar MV which commonly uses a 2.2M safety resistor. Yes, I see the 10:1 ratio. But, as said, once the wiper fails, the pot is no longer connected to the grid, so bias voltage travels through the 2.2M safety resistor to the grid. Seems the same as a 2.2M across a failed 10K bias pot to me.

[pdf64]

I suppose a difference is that if the master vol wiper goes open, the sound will go rather wonky, especially at high vol settings, so an issue should hopefully be apparent to the user.
Whereas if the bias goes open, nothing much may change, and the amp carries on being used.

[PRR]

There's a real difference between Ig at happy -45V bias and Ig at wiper-off -65V bias.

I think you two are hitting both ends of a tolerable range. If I found 680k on the floor I would use them.

[sluckey]

When the bias pot wiper fails, the grid bias goes to max negative. Surely anyone interested in this thread could tell that something was not right.

[Latole]

This

[sluckey]

This

pdf64 and Merlin love you.   

[sluckey]

Put a 2.2M resistor between the pot left and center lug.

You could also achieve the same protection by putting a jumper between the pot left and center lug.

[pdf64]

Put a 2.2M resistor between the pot left and center lug.

You could also achieve the same protection by putting a jumper between the pot left and center lug.

In a Princeton Reverb etc with a big AC dropper, that would be fine, but with a bias tap and 470ohm dropper shown in post #1, I think that would lead to a poor range of bias adjustment. There’s been a thread where this was an issue, see https://el34world.com/Forum/index.php?topic=27767

[joesatch]

Put a 2.2M resistor between the pot left and center lug.

Like this?

[sluckey]

In a Princeton Reverb etc with a big AC dropper, that would be fine, but with a bias tap and 470ohm dropper shown in post #1, I think that would lead to a poor range of bias adjustment.

This slightly different circuit has plenty of adjustment range...

     http://sluckeyamps.com/misc/bandmaster.pdf

[pdf64]

In a Princeton Reverb etc with a big AC dropper, that would be fine, but with a bias tap and 470ohm dropper shown in post #1, I think that would lead to a poor range of bias adjustment.

This slightly different circuit has plenty of adjustment range...

     http://sluckeyamps.com/misc/bandmaster.pdf

I regard that as being more than slightly different.
From the perspective of the trimmer’s variable resistance, it’s drawing current via a >15k supply impedance.
Hence it’s closer to a Princeton Reverb than the arrangement shown in post #1.

Attempting to vary the output of a comparatively lower impedance supply by loading it down seems inherently the wrong approach to me, as the ripple will tend to increase and, taken too far, things are going to become stressed.

[sluckey]

I concede. Wrestling with an armchair engineer is always futile!

[shooter]

 
the lab engineers would come out to the field every now and again,  they had fits when they found our hardware mods n software patches!!
the funny thing, most of those temp fixes eventually made it into "official upgrades"