Bias 6L6s in my Accomplice build?

[doveman]

Have at it! BTW, how are you measuring idle current?

Pin 3 to ground - both sockets are the same.

[martin manning]

Is "pin 3 to ground" voltage or current? I recall you said you have bias test ponts built into the amp- are there 1 ohm resistors to ground from those jacks? If so you are measuring cathode current, which includes screen current. Screen current doesn't count against plate dissipation.

[doveman]

Is "pin 3 to ground" voltage or current? I recall you said you have bias test ponts built into the amp- are there 1 ohm resistors to ground from those jacks? If so you are measuring cathode current, which includes screen current. Screen current doesn't count against plate dissipation.

voltage

Yes test bias points with 1 ohm resistors. They are from pin 8.

So how would that change my calculation? Where would I identify the difference?

[ampgeek]

Measure the voltage drop across your power tube screen grid resistors and divide by the value of your screen grid resistor.

This is the current flowing to the power tube screens and out through the cathode.

You are measuring the combined screen and plate current at the cathode. Subtract the the screen grid contribution calculated from above from your cathode current measurement to get the plate current contribution.

Use that value along with your plate voltage to get the plate-alone power dissipation.

Hope that helps!

Dave O.

[martin manning]

You can measure the voltage across the screen resistors and divide by their resistance to get screen current, which you would subtract from the cathode current. The difference will typically only amount to a few percent, and if you ignore it you are on the conservative side, so most people don't worry too much about it.

For example, from GE's 6L6GC data sheet at 400V Va = Vs, and plate current = 40mA, screen current is ~1.8mA. Using total current of 41.8mA (as measured from the cathode) and 400V you get 55.7%, and with plate current alone you get 53.3%, or 2.4 points difference.

[Firestorm]

For years, the basic rule on Russian 6L6GC's, both New Sensor/Reflektor and SED, was to treat them as 25-watt tubes even if the spec-sheet said otherwise. At one time the only real 30-watt tube New Sensor sold was the 5881WXT...and they didn't advertise it as such. Not sure about their current offerings.

Not to argue, but for future reference, this is not what I've heard.
The SED 6L6GC was a true and proper 30W tube. The Reflektor 5881WXT was a 25W tube. The Reflektor 6L6WXT+ was a 30W tube.

I could be wrong, but I won't admit it.

[wyatt]

For years, the basic rule on Russian 6L6GC's, both New Sensor/Reflektor and SED, was to treat them as 25-watt tubes even if the spec-sheet said otherwise. At one time the only real 30-watt tube New Sensor sold was the 5881WXT...and they didn't advertise it as such. Not sure about their current offerings.

Not to argue, but for future reference, this is not what I've heard.
The SED 6L6GC was a true and proper 30W tube. The Reflektor 5881WXT was a 25W tube. The Reflektor 6L6WXT+ was a 30W tube.

I could be wrong, but I won't admit it.

That is how they were marketed. That was not the performance that builders/designers got out of them.

Trying to think where to look up the info right now, but I was told always bias a SED, Sovtek 6L6WXT or 6L6EH is if it were a 25-watt tube, and NOS, Chinese or JJ as 30-watt.

But now that I'm looking for it, I can't seem to found any corroboration. I certainly can't find anything to back up my claims on the 5881WXT, since the Russian tube is really a 20-watt (not 23-watt or 25-watt) per it;s tube data.

[Firestorm]

Wyatt,

Frustrating when all the marketing departments do is republish old RCA or Mullard specs. But I'm pretty sure that St. Petersburg really made tubes to the specs they published. Not so much in Saratov, although now they might be more serious. When I bought tubes a case at a time, there were always outliers, so published advice might be based on the lower edge of QC.

Mullard, for example, was the only company to take rectifiers seriously. A GZ34 that was ridiculously good got branded Telefunken; a really good one said Mullard; a decent one (still within specs) could be sold to RCA.

Tube companies do not pay so much attention anymore.

[doveman]

I just had another idea. I have a pair of Tung-Sol 5881 reissues that I ran for a long time at 445v in my Brown Note. They are suppose to be able to take up to 500v according to some but I only have experience with 445v. They should be able to take 420v pretty easy and at 23v plate dissipation, I could run it up to maybe 38ma. I always liked those tubes as they were in between 6V6 and 6L6GC. Might be a good experiment.

I tried these today and sure enough the seem to sound better than both the 6V6s and 6L6s ... biased at about 38ma. But I still think the feel is wrong. Before going on the hunt, I asked David if he had a schematic with test voltages. I'll probably look at all the parts again too. I think I've missed something. It may just need to break in ...but it "feels" a bit wrong.

[martin manning]

What kind of speaker do you have in this thing? If it is not one that you are familiar with that could be a contributor.

[doveman]

What kind of speaker do you have in this thing? If it is not one that you are familiar with that could be a contributor.

I've tried it through:

• * scumback M75 8 ohm
  * celestion G1265 8 ohm
  * 16 ohm and 4 ohm combinations of the above two
  * EVM12L 8 ohm
  * Eminence 10516 2x10 at 8 ohms

I'm getting an Eminence RWB 8 ohm 12 with the cabinet but not here yet.

They all sound different but still seem a little choked. I'm going to send David some test voltages. I can say it seems closest to right with the 5881 power tubes and a 12at7 in PI slot.

[Roe]

usually 50-60% heat diss. sounds best with blackface type amps.

Just to be clear, is this based on true plate current or plate plus screen?

true plate current. but the screens do not draw much current at all in these amps. So it doesn't make that much of a difference really. A marshall, however, is something different

[martin manning]

usually 50-60% heat diss. sounds best with blackface type amps.

Just to be clear, is this based on true plate current or plate plus screen?

true plate current. but the screens do not draw much current at all in these amps. So it doesn't make that much of a difference really. A marshall, however, is something different

Yes, a similar analysis as above, but for an EL34, with Va = 425, Vs = 400V at 30mA Ia, or 51% Pa max has screen current at 4.4mA (Telefunken data sheet). Using total current (34.4mA) puts calculated dissipation at 58.5%.

Phillips data sheet numbers give a similar result at the same plate and screen voltages (425/400) and 70% Pa max: Ia = 41mA, Ig2 = 5mA. Using total current calculated dissipation is 78.2%

So, around 8 points difference using cathode current vs. anode alone.

[Firestorm]

Screens react to signal more visibly than plates. If you play through it and can see the screens lighting up in response, the fix is to change signal relative to what the screens see. Easiest way is to change the grid loads. If they're 220k make them 200k or lower.

[Roe]

Just to be clear, is this based on true plate current or plate plus screen?

true plate current. but the screens do not draw much current at all in these amps. So it doesn't make that much of a difference really. A marshall, however, is something different

Yes, a similar analysis as above, but for an EL34, with Va = 425, Vs = 400V at 30mA Ia, or 51% Pa max has screen current at 4.4mA (Telefunken data sheet). Using total current (34.4mA) puts calculated dissipation at 58.5%.

Phillips data sheet numbers give a similar result at the same plate and screen voltages (425/400) and 70% Pa max: Ia = 41mA, Ig2 = 5mA. Using total current calculated dissipation is 78.2%

So, around 8 points difference using cathode current vs. anode alone.

The RFTs (aka Siemens) draw even more screen current, often 6-7mA