Trainwreck Pages bias spec.

[leadfootdriver]

On the TWP on page 28, Ken specs -30vdc on pin 5 on the power tubes. This is way below 70% max dissipation.

Anyone have some insight on this?

[Phil_S]

According to the 113 schematic in circulation, bias voltage is -33 and there is a 25K pot in the bias circuit. I didn't do the math, but I'd guess range is about 5V +/- whatever the center value. Looking at EL34 specs, I'd guess bias voltage might be set closer to -40, but then again, it is all a matter of how a particular pair of tubes might actually bias and your personal taste. IMHO, -30 is a bit on the hot side, but I'm no expert, and certainly not in Ken's league. I don't imagine the listed voltage is a hard and fast number given the variability in tubes. It may suggest that Ken liked to bias them hot.

[Structo]

Not to butt in but that -30v figure is meaningless if one doesn't know the plate voltage or screen current.

Many things are not cut and dry with amps.

As long as you stay within the safe operation of the power tubes, then you can be safe with great tone.

In other words, one man's recommendations for bias might not fit an amp even with similar parts.

[martin manning]

This is one time KF screwed up... :^o

[Reeltarded]

Throw in JJ 6ca7 bias it up and then swap in GE 6ca7. YIKES!!! RFTs can take the punch too.

Modern tubes are barely tubes. They just look like tubes.

[Tillydog]

This is one time KF screwed up... :^o

You want to know what I think?

(Well probably not, but this is the internet, so it's coming anyway....)

The distortion on an Express kicks in when the power tubes are driven into grid conduction. Once this happens, because of the way the negative feedback works, the PI output increases until it is driven to compression. Where things end up here determines how heavily distorted the amp is once it has been tipped into the 'mean' mode. More distortion then comes from driving the (already compressing) PI harder.

That initial distortion and the way it comes in and cleans up is extremely sensitive to the headroom of the PI relative to the onset of grid conduction in the power tubes - i.e. to a first approximation: the PI B+ (which determines the available headroom) and the power tube bias voltage (and, of course, reliable, predictable behaviour of the PI tube itself).

SO.... *I* think that KF specified a *voltage* for the bias setting in order to preserve the 'clean to mean' transition as he intended it.

YMMV, FWIW, IANAL, etc... etc...

IDGAFA!

[martin manning]

I'm always interested to hear what you and others think, T-dog!

I'm not ready to buy the notion that power tube grid conduction is required, though. In fact that could result in nasty blocking distortion. OTOH, a collapse of the NFB can be brought on by overdriving the PI into clipping, and the circuit is tailor-made to do that.

[leadfootdriver]

So then, is the greater the negative voltage at the grids, the colder the bias, or the other way around.

I have 1 ohm resistors on the cathodes, and have been biasing at 43.7ma's using the cathode current method which puts my bias at 70%. But setting the grids at -30vdc (or 28.7 because that's as far as my pot goes back with this set of tubes) puts the cathode method at 16.4 ma's.

This IS a colder bias, is it not?

[Tillydog]

I'm always interested to hear what you and others think, T-dog!

Why thank you! Likewise

[martin manning]

So then, is the greater the negative voltage at the grids, the colder the bias, or the other way around.

I have 1 ohm resistors on the cathodes, and have been biasing at 43.7ma's using the cathode current method which puts my bias at 70%. But setting the grids at -30vdc (or 28.7 because that's as far as my pot goes back with this set of tubes) puts the cathode method at 16.4 ma's.

This IS a colder bias, is it not?

Yes, lower current = colder bias. 16mA and 400V (?) would be 6.4W... ~25% Pa max with screen current included. Burrrh!

[katopan]

That initial distortion and the way it comes in and cleans up is extremely sensitive to the headroom of the PI relative to the onset of grid conduction in the power tubes - i.e. to a first approximation: the PI B+ (which determines the available headroom) and the power tube bias voltage (and, of course, reliable, predictable behaviour of the PI tube itself).

I like that description.

But setting the grids at -30vdc (or 28.7 because that's as far as my pot goes back with this set of tubes) puts the cathode method at 16.4 ma's.

I can't understand how it's so cold with -28.7V, but that must be the valves you're using. I had to adjust my bias range and have mine set from memory around -35V to get the current around 40mA.