Biasing by Percentage of Maximum Plate Dissipation

[martin manning]

Seems like this has settled down and I appreciate all the inputs, which have definitely improved this guide.

I like the suggestion to move this thread to the FAQ section, it will be easier to find there. Can one of the moderators do that? Thanks!

[pdf64]

... The '62 RCA and '63 GE 6V6GT(A) datasheets specifify PDmax as 14W (Desing-Max-Rating).

So this might suggest a crude conversion from design max to design center of 12/14, or -14%. This is of course for GE's assumptions and production quality standards for this particular tube type.

The 1st generation 6L6 (inc G, GA, GB) had a 19W anode dissipation limit under the design centre system, 22W design max.
https://frank.pocnet.net/sheets/137/6/6L6G.pdf
https://frank.pocnet.net/sheets/049/6/6L6.pdf
https://frank.pocnet.net/sheets/049/6/6L6GB.pdf
https://frank.pocnet.net/sheets/127/6/6L6GB.pdf

That further supports the design centre rating being 0.86 x design max.
I think that relationship has to have been industry wide, as I suspect it would have been unacceptable to the EIA etc for a valve type from some manufacturers not meeting the type's minimum requirement.
Also I suspect it would have been unacceptable for one manufacturer to publish ratings for their version of eg 6L6GB higher than the type standard, even if they could demonstrate it. Hence GE stating in the Ham Radio publication I linked previously that their 6L6GB had a 24W limit.

Obviously there's that EI 6L6GC info throwing a spanner in the above hypothesis. I think it's an inexplicable outlier, perhaps by the time it was published, the EIA or whoever had given up attempting to regulate the vacuum tube market.

The relationship between design centre and design max seems pretty well defined, providing an allowance for component and supply variations, and so seems unrelated to the valves themselves.

RCA AN174 introduces the design maximum system, explains its rationale and how to use it https://www.one-electron.com/Archives/R ... 0Tubes.pdf

I'd stick to the design-max system as it seems to be most widely used since the 60s.
Derating to design-center is likely to cause confusion.

I think that there should be a warning not to use absolute maximum limits, as those seem to have no fixed relationship to the other systems. My impression is that they depend on the manufacturer's acceptance criteria.
The GEC 'KT' valve info notes a pretty wide difference between the design max and absolute max limits, eg
https://frank.pocnet.net/sheets/086/k/KT77.pdf

[Helmholtz]

I think that there should be a warning not to use absolute maximum limits, as those seem to have no fixed relationship to the other systems. My impression is that they depend on the manufacturer's acceptance criteria.
The GEC 'KT' valve info notes a pretty wide difference between the design max and absolute max limits, eg
https://frank.pocnet.net/sheets/086/k/KT77.pdf

Agree, but I didn't speak of absolute max ratings.
I would actually be surprized by a fixed absolute-max to design-max ratio for all tubes.
Absolute limits need to be found from long term statistical evaluation at the manufacturer.

Another remark:
Within the 6L6 family (including specialised types like the 6L6WGB or the 5881) only the 6L6GC has a 30W design-max PD as well as highest screen and plate voltage limits.
So, if a circuit was designed around the 6L6GC, none of the other 6L6 types should be used.
Leo Fender specified the 6L6GC when required by the circuit.

As indicated earlier I think the PDmax for a 6V6GT should be 14W to be in line with the design-max values for the other tubes.

[jabguit]

Another remark:
Within the 6L6 family (including specialised types like the 6L6WGB or the 5881) only the 6L6GC has a 30W design-max PD as well as highest screen and plate voltage limits.
So, if a circuit was designed around the 6L6GC, none of the other 6L6 types should be used.
Leo Fender specified the 6L6GC when required by the circuit.

7581A @ 35W plate dissipation would be acceptable substitute fro 6L6GC.


cheers,

[martin manning]

More examples of the difference between the Absolute Max and Design Max Pa rating:

Genelex KT77 (linked by PDF above)
Pa 32W Abs Max vs. 25W Des Max (22% derate)
Pa+Pg2 Same

GEC KT88 https://frank.pocnet.net/sheets/084/k/KT88_GEC.pdf
Pa 42W Abs Max vs. 35W Des Max (16.7% derate)
Pa+Pg2 46W Abs Max vs. 40W Des Max (13.0% derate)

GEC KT66 https://frank.pocnet.net/sheets/086/k/KT66.pdf
Pa 30W Abs Max vs. 25W Des Max (16.7% derate)
Pa+Pg2 32W Abs Max vs. 27W Des Max (15.6% derate)

A note on the KT66 data sheet says the Pa+Pg2 rating is for triode or ultra linear operation, which does not appear on the KT88 data sheet.

[pdf64]

...
Within the 6L6 family (including specialised types like the 6L6WGB or the 5881) only the 6L6GC has a 30W design-max PD as well as highest screen and plate voltage limits.
So, if a circuit was designed around the 6L6GC, none of the other 6L6 types should be used.
Leo Fender specified the 6L6GC when required by the circuit.

I think that most 60s and 70s Fenders using 6L6GC (and the modern HRDeluxe) had pretty much the same / no harsher operating conditions than their preceding later tweed / 'brown panel' models specified for 5881.
Certainly though it might be argued that many of the 5881 models seemed to exceed their limits, and they may be better off using GCs
eg the 6G6 Bassman https://el34world.com/charts/Schematics ... _schem.pdf

... I think the PDmax for a 6V6GT should be 14W to be in line with the design-max values for the other tubes.

Of the valve types mentioned in the document, 5881, EL34 and EL84 are all rated under the design centre system.

[Helmholtz]

I guess the 6L6GC was not yet available when Leo designed his early amps?
I'm sure he learned from tube failure experience.

Can we be sure that European and US rating systems used the same criteria?

[martin manning]

Fender used the industrial/military 5881 for the large amps produced from the mid-fifties. The 5E8-A layout shows 6L6G, 5F8-A and 6G8-A show 5881. The 5F8 is the later model in Fender's nomenclature, and 6G8 marks the new decade. [edit: probably due to the 5881's more rugged construction]. 6L6GC was not available until the end of the 1950's, and it shows up in the brown panel amps that were produced in the very early 1960's.

[pdf64]

I guess the 6L6GC was not yet available when Leo designed his early amps?
I'm sure he learned from tube failure experience.

I suspect that Fender swapped to GCs from 5881 when they became cheaper and easily available, the H-FA August 1961 revision of the 6G9A Tremolox being the earliest model I've noticed with GCs. 5881, being a military grade valve type may have been quite expensive.
https://el34world.com/charts/Schematics ... a_h_fa.pdf

Can we be sure that European and US rating systems used the same criteria?

I've looked but can't find it documented, so I can't be 100% certain. But my understanding is there was a lot of cross Atlantic communication etc, so it would be unusual for the same terminology to be used for such a key thing, but for it to have differing definitions.

[pdf64]

Here's an interesting Tung Sol publication, introducing the 5881 https://www.qsl.net/dl7avf/roehren/e5881.html

[jabguit]

Here's an interesting Tung Sol publication, introducing the 5881 https://www.qsl.net/dl7avf/roehren/e5881.html

gold plated grid wire...


cheers,

[Metanoisus]

Thanks for the doc Martin. You may be able to help with somewhat related questions. How do I calculate my actual output power in RMS for both tubes in a class A/B? Or more specifically do I need to bias my tubes lower so I don't exceed my OT rating in RMS.

My actual situation is that I am running my Trainwreck Express 6V6 at about 9.5 watts per tube (approximately 180/190 Ohm, ~420Va, ~4.5Vpr). My output transformer is a Hammond 6,600 C.T. 20 watt RMS rated (1760H). How do I calculate that I am biasing within range of my OT capability?

Second question is the % power for my 6V6. I am using TunSol 6V6 GC and have seen 12 watts and 14 watts as 100%. The spec sheets I see online are not rated that high. 9.5 of 14 is around 70%, but 9.5 of 12 is around 80%. Should I be conservative and run these down at 8 watts? This amp is loud as a MF running at 9.5 watts and crossover distortion is fine down at 8-8.5 watts.

proof of build

[martin manning]

How do I calculate my actual output power in RMS for both tubes in a class A/B? Or more specifically do I need to bias my tubes lower so I don't exceed my OT rating in RMS.

Output power into a resistive load is calculated using Watt's law, I^2*R or V^2/R, where I and V are RMS values and R is the dummy load resistance. For an unclipped sine wave the RMS voltage is Vpeak/sqrt(2), so you would measure peak voltage on your scope just at clipping, and calculate output power as (Vpeak/sqrt(2))^2/R = Vpeak^2/2R.
Zero-signal dissipation doesn't have much to do with the max power case. The danger to your OT is in driving the output tubes to square wave, where the RMS value is equal to the peak value. That will leave you with just Vpeak^2/R, or ~200% of the measured power at clipping.

Second question is the % power for my 6V6. I am using TunSol 6V6 GC and have seen 12 watts and 14 watts as 100%. The spec sheets I see online are not rated that high. 9.5 of 14 is around 70%, but 9.5 of 12 is around 80%. Should I be conservative and run these down at 8 watts? This amp is loud as a MF running at 9.5 watts and crossover distortion is fine down at 8-8.5 watts.

If you want to be conservative use 12W. GE's 6V6-GT data sheet from 1955 gives max plate dissipation at 12W, design center. The 1967 GE data sheet for 6V6-GTA gives design max at 14W. JJ's current production 6V6S data sheet says 14W, no information on whether that is design center or design max, but it looks more robust than the traditional 6V6GT.