Princeton Reverb bias question

[Randall]

A beautifully done blackface Princeton Reverb came to me, built recently by a local longtime builder who has since passed away. It looks to be AA1164, at least the bias circuit is. However, it has 5881's in it, and there is a 10k pot where a bias pot would be, and it is wired in series to ground with the 22K bias resistor. Other than that, the bias circuit follows the schematic.

Problem is, the bias current is way hot, 44 - 46mA @ 413v. That's 19 watts by my calculations. This is with the pot dialed full open, putting 10K between the 22k and ground, the grids are seeing -39v. As I started to close the pot and make the bias voltage less negative, the current goes much higher. I dialed it quickly back when I glanced over and saw 75mA!

So... why? Is it the 5881's require more bias voltage than this circuit can deliver? Am I looking at this the wrong way, or am I incorrect in my thinking that Tung Sol 5881 has a max plate diss. of 23 watts? Is the data sheet I am looking at not relevant to the Russian made ones in the amp?

http://www.thetubestore.com/lib/thet...-Schematic.pdf

http://www.tubeampdoctor.com/images/...20Tung-Sol.pdf

[lowell]

Is the bias CURRENT equal on both tubes? Is the negative grid bias voltage equal on both tubes? Finally, how does the amp sound?

Could be you just need more negative voltage. Could be the tubes. Finally, could be the OT. Assuming the answers to the first two questions are - yes, and the third question answer is - like shit.

[g1]

What is the negative DC voltage going to the grids, and is there any AC ripple on it?
How is the resistor feeding the bias diode, 100K ?

Also, if the coupling cap to the trem pot is leaky, you will lose some bias voltage, you may want to try disconnecting that cap.

[Dave H]

So... why? Is it the 5881's require more bias voltage than this circuit can deliver? Am I looking at this the wrong way, or am I incorrect in my thinking that Tung Sol 5881 has a max plate diss. of 23 watts? Is the data sheet I am looking at not relevant to the Russian made ones in the amp?

You could reduce the value of the 100k in series with the diode but if it sounds good and those Russian tubes are Sovtek 5881 then I wouldn't worry about it. They are good for 30W.

[Randall]

The current to the tubes is within a couple mA. The grids are getting -39v, which is more than the schematic shows, so I don't think the circuit is failing. No appreciable ripple. I don't want to mess with any soldering on this guy if I don't have to, it's a new and very well done amp, and the owner just bought it. I have fixed the complaint, a dirty PI socket connection. I just was curious about the 5881's so I measured them, and said, "whoa!".

The amp works and sounds great, and is loud, as it also has a Celestion G12 H in it. (barely fits).

[Tom Phillips]

It is not at all surprising that a circuit that was designed for 6V6s, that now has 5881s installed (For practical purposes a member of the 6L6 family), would need adjustments to the bias supply circuit to allow a reasonable adjustment range. The tube types may be interchangeable in some respects but they are not equivalent. A look at some old 2x 6L6 fender circuits such as the Bandmaster will show you that, in general, the 6L6 requires a more negative bias supply than a 6V6.
Edit: Also useful and interesting to review the spec sheets for the two tubes.

[Justin Thomas]

Hey Dave H,

Randall said they're TS. If they're reissue (likely), then they certainly are not 30W. My Prosonic redplated a pair in under 60 seconds. Maybe REAL TS 5881s could take the punishment, but yeah, those Sovteks, though... never killed one yet.

Justin

[g1]

Ooops, I didn't even pay attention to this being a 6V6 amp, thanks Tom. Yes 5881 or 6L6 type will have different bias requirements than a 6V6 and it should be adjusted. The 100K feeding the diode that Dave H mentioned would be a good spot (decrease value).
The Tung-sol 5881 re-issue are supposed to be a 23W tube with special considerations noted here:
www.thetubestore.com - Tung-Sol 5881 Audio Tubes

[Dave H]

Hey Dave H,

Randall said they're TS.

Thanks for pointing that out. Tung sol didn't register. I only saw 'Russian' and thought they would be the indestructible Sovtek 5881WXT

[Randall]

I replaced the 100k bias resistor with 82K 2 watt, and all is well. With the 10k pot I mentioned that is in series with the 22K to ground, dialed to about half I get 35mA and 37mA at 418v, which puts me at 60 - 65% of 23 watts. That makes me a lot happier. Something tells me this amp was still a work in progress of a dying man who didn't get to finish it. I also found an unsoldered connection on the bias adjustment pot, and one wrong value and one missing resistor in the non-functioning tremolo circuit. But it all works now, and it sounds great!

Thanks guys!