6L6 cathode bias

[rzenc]

Hi guys!

I would like suggestions on what O.T. you are using for a pair of 6L6** running cathode biased. I wanna build a standard AB763 - no vib no verb - preamp + P.I. and P.A. but running cathode bias.
RCA chart calls for 9Kohms plate to plate with 360V @ Plates and 270V @ G2....

Thanks in Advance
Best Regards

Rzenc

[jjasilli]

6L6** vs 6L6GC:  Are you sure your chart is for the GC??  9000K_p-p seems wrong. 

[rzenc]

6L6** vs 6L6GC:  Are you sure your chart is for the GC??  9000K_p-p seems wrong. 

Yeah, it's from RCA 6L6GC. File is too large to post.

[phsyconoodler]

Look at some other 6L6GC charts.They all say 4.2k or 6.6k or very close to that.
  Does this not mean both tubes add to get 9000 ohms? So 4.5k + 4.5k.  Makes sense to me.

[jjasilli]

Ditto.  And, here's some more info:  http://www.thegearpage.net/board/archive/index.php/t-678299.html

But those early Bassmans & Twins were not yet using GC's.  They probably get the large voltage drop from the plate supply to the screen node by using the rather large 10K dropping resistor.  I still suspect that  9000Kp-p spec, along with the max power output.

Output Wattage = Plate voltage² / OT primary Imp.  Assuming the OT prim imp. matches the plate-to-plate imp. shown on your chart, then:

per your chart Fixed bias @ 360V should yield only about 19W; Cathode bias only abnout 14W.  This does not match your chart.  Something seems to be wrong somewhere.

[HotBluePlates]

What is your proposed supply voltage? The OT primary impedance is not selected in isolation.

Also, cathode bias does not lend itself to deep class AB, high supply voltage or high power output for the tubes. The large current variations of deep class AB mess up the operating point when you sit the tubes on top of a cathode resistor, which ruins the possible output power.

Also, if you run class AB that's not very deep into AB, you tend to use a lower supply voltage, and a smaller current swing. That means you need to use a higher primary impedance to have a big enough voltage swing to get close to maximum output power for that configuration.

No-calculation examples: A higher voltage (450v) deep class AB amp might use a 4k primary (possibly smaller), while a lower voltage (350v) hotter class AB amp might use a 6.6k primary.

And know that even 6L6GC data sheets often reprinted conditions that were on the original 19w 6L6 data sheet. They may not show a condition that takes advantage of all that the 6L6GC is capable of.

Output Wattage = Plate voltage² / OT primary Imp.  Assuming the OT prim imp. matches the plate-to-plate imp. shown on your chart, then:

per your chart Fixed bias @ 360V should yield only about 19W; Cathode bias only abnout 14W.  This does not match your chart.  Something seems to be wrong somewhere.

Your equation is handy for quick sanity check, but also guesses output power well above real circuits. For your equation to work, the tubes need to have perfectly horizontal curves, equally spaced (perfect linear and no distortion), and the tube must be capable of pulling the plate all the way down to 0v.

It also ignores the fact that watts are calculated using RMS voltages and currents; normally peak plate voltage swing is found from the curves and multiplied by 0.7071 to get RMS/effective voltage swing for power calculation.

But the errors won't be so bad as to confuse 10w output with 1w output, and is close enough for figuring a design on a napkin. It won't help you make sense of data sheet numbers.

[PRR]

> cathode bias does not lend itself to ...high supply voltage

High voltage is fine; except it leads to awkward high load impedance and suggests G2 voltage awkwardly lower than plate.

> or high power output

Correct. Cathode-bias is upset if current changes from idle to full output, which essentially means Class A, much less output power than possible in Class AB.

(The sheet-snip above says "Class AB", but the cathode current rises only 25%, which is a lot closer to A than to B.)

> RCA chart calls for 9Kohms

Look at the page before that. Class A push-pull cathode bias. 250-270V 5K and 14W-18W.

The 270V condition has a nice side-effect: the ~~150mA total cathode current is just about what a 12AX7's heater needs. Free DC for preamp heater.
_______________________________

For marketing reasons, as HBP says, 6L6GC suggestions are mostly old 6L6 suggestions, not taking advantage of the GC's higher G2 and Pd ratings.

Look at 7027. Different pinout, same guts. Middle of page 2 shows P-P self-bias operation.

http://www.mif.pg.gda.pl/homepages/frank/sheets/049/7/7027A.pdf

[rzenc]

Cool guys!!! Many thanks for all the information!!!
It's a first time with 6L6** cathode biasing....I have built tweed types - 5E3; 5E9-A - with 6V6 and really liked the sounded. Don't know how much of it's sound comes from output being self vs. fixed. Have you ever have the opportunity to do such test?
I remember some years ago, El Guapo Loco (former member) built a TWIN 5D8 and I had the opportunity to hear him playing harmonica thru it...I must say it was one really fine sounding amp indeed!!!!
I have a 6550 p-p amp (my fav big bad tube) with ability to switch from fixed to cathode bias and although they behave differently, in fixed bias mode there seems to be a kinda punch which is not present on cathode bias mode. On the other hand, cathode bias shows a more smooth amp - biased on the hot side ~ 100mA's per tube @ 430V on plates - with none of the agressive character of  fixed mode.




The 270V condition has a nice side-effect: the ~~150mA total cathode current is just about what a 12AX7's heater needs. Free DC for preamp heater.

I think I did not understand what you meant...


Correct. Cathode-bias is upset if current changes from idle to full output, which essentially means Class A, much less output power than possible in Class AB.
(The sheet-snip above says "Class AB", but the cathode current rises only 25%, which is a lot closer to A than to B.)

RCA 7027A data sheets shows variations around 30% of cathode current from no-signal to max-signal conditions. Max power varies ~ 37% from condition 1 to condition 3. Effective Load resistance varies a lot. The higher the load, the lesser amount of power output in watts it's seems capable of generating.
Heat generated between specs - heater power was not included - :

condition 1> MAX-SIGNAL POWER OUTPUT: 32W.
no-signal: plate @  44,8W D.C.; G2 @ 2,1W D.C.>>>>TOTAL DCWATTS: 46,9W
max-signal: plate @ 51,2W D.C.; G2 @ 4,8W D.C.>>>TOTAL DCWATTS:  56W

condition 2> MAX-SIGNAL POWER OUTPUT: 36W.
no-signal: plate @   52,44W D.C.; G2 @ 2,128W D.C.>>>TOTAL DCWATTS:  54,56W
max-signal: plate @ 64,6W D.C.; G2 @ 7,6W D.C.>>>TOTAL DCWATTS:  72,2W 

condition 3> MAX-SIGNAL POWER OUTPUT: 44W.
no-signal: plate @ 63,75W D.C.; G2 @ 3,4W D.C.>>>TOTAL DCWATTS:  67,15W
max-signal: plate @ 83,3W D.C.; G2 @ 8,5W D.C.>>>TOTAL DCWATTS:  91,8W   

Except for condition 1, it seems that audio power output is roughly 50% of generated heat (I'm not considering necessary power from driver).

According to the data sheet PRR provided for RCA 7027A, the most common part I can find down here is 425V @ plates and G2 with 3K8 load, so I will give it a try.
I would like to hear your opinions on how big P.T. needs to be..

Secondary: H.T.> 330 - 0 - 330 @ 350mA's; Fil.1> power tubes 4A's; Fil.2> preamp 2A's. Choke: 20H/50ohms @ 150mA's D.C.

Many thanks in Advance!! You guys rule!!
Best Regards

Rzenc

[jbefumo]

Ahh -- I KNEW if I searched long enough I'd find the answer I WANTED .....


I'm building an ultra compact 45-Watt amp using a pair of 6L6s. Would like it to come in around 45-55W.  My initial assumption was that I'd have to go with fixed bias, but I generally prefer cathode biasing.  Just glancing at the 7027A datasheet, and adding a bit of hand-waving, it seems like I should be able to get away with 450V plate I was already planning, coupled with a 4.2k OT load.  Is this reasonable, and more importantly, will it be safe in terms of not destroying power tubes?


Thanks!


Joe

> cathode bias does not lend itself to ...high supply voltage

High voltage is fine; except it leads to awkward high load impedance and suggests G2 voltage awkwardly lower than plate.

> or high power output

Correct. Cathode-bias is upset if current changes from idle to full output, which essentially means Class A, much less output power than possible in Class AB.

(The sheet-snip above says "Class AB", but the cathode current rises only 25%, which is a lot closer to A than to B.)

> RCA chart calls for 9Kohms

Look at the page before that. Class A push-pull cathode bias. 250-270V 5K and 14W-18W.

The 270V condition has a nice side-effect: the ~~150mA total cathode current is just about what a 12AX7's heater needs. Free DC for preamp heater.
_______________________________

For marketing reasons, as HBP says, 6L6GC suggestions are mostly old 6L6 suggestions, not taking advantage of the GC's higher G2 and Pd ratings.

Look at 7027. Different pinout, same guts. Middle of page 2 shows P-P self-bias operation.

http://www.mif.pg.gda.pl/homepages/frank/sheets/049/7/7027A.pdf

[jbefumo]

Oh, neglected to mention that I'm thinking in terms of ~45W.

Ahh -- I KNEW if I searched long enough I'd find the answer I WANTED .....


I'm building an ultra compact 45-Watt amp using a pair of 6L6s. Would like it to come in around 45-55W.  My initial assumption was that I'd have to go with fixed bias, but I generally prefer cathode biasing.  Just glancing at the 7027A datasheet, and adding a bit of hand-waving, it seems like I should be able to get away with 450V plate I was already planning, coupled with a 4.2k OT load.  Is this reasonable, and more importantly, will it be safe in terms of not destroying power tubes?


Thanks!


Joe

[drgonzonm]

Hi guys!

I would like suggestions on what O.T. you are using for a pair of 6L6** running cathode biased. I wanna build a standard AB763 - no vib no verb - preamp + P.I. and P.A. but running cathode bias.
RCA chart calls for 9Kohms plate to plate with 360V @ Plates and 270V @ G2....

Thanks in Advance
Best Regards

Rzenc

I have the same datasheet in my collection.  dated 8-60

[tubenit]

Yrs ago, I have used a 20w (1620)  Hammond OT in a couple of 6L6 builds.  I compared volume once between my 23w 5881 cathode biased amp with a Marshall 50w plexi.  The amp heads were side by side played thru the same speaker cab and same guitar.  Not much difference in useable volume, IMO  and in the opinion of the guy who owned the Marshall.  He agreed the 23 watt amp I had would have enough headroom to play in the local clubs in our area.  Just a thought to consider.  Doubling the wattage only gets you a few more decibels. 

With respect, Tubenit

[shooter]

He agreed the 23 watt amp

I got about the same comment comparing my 23wattish against a 45watt fender, although speaker cabs were different, otherwise same guitar, same player, He even argued my amp was NOT 23watts

[HotBluePlates]

Oh, neglected to mention that I'm thinking in terms of ~45W.

Then why not copy the 44w cathode biased condition on the 7027 data sheet?  425v plate & screen, 200Ω cathode resistor, 3.8kΩ OT.  Plate idles at 75mA each tube (150mA total) and 4mA screen (8mA total).

200Ω * 0.158A = 31.6v
425v - 31.6v = ~393v

393v * 0.075A = ~29.5w

Perfect for maxing out class A in a pair of (true) 6L6GC's.

[jbefumo]

Pretty much what I'm going to do -- only that 4,2k OTs are easier to find -- I suspect that the difference doesn't amount to much, and figured a slightly higher plate would compensate?

Oh, neglected to mention that I'm thinking in terms of ~45W.

Then why not copy the 44w cathode biased condition on the 7027 data sheet?  425v plate & screen, 200Ω cathode resistor, 3.8kΩ OT.  Plate idles at 75mA each tube (150mA total) and 4mA screen (8mA total).

200Ω * 0.158A = 31.6v
425v - 31.6v = ~393v

393v * 0.075A = ~29.5w

Perfect for maxing out class A in a pair of (true) 6L6GC's.

[HotBluePlates]

Pretty much what I'm going to do -- only that 4,2k OTs are easier to find -- I suspect that the difference doesn't amount to much, and figured a slightly higher plate would compensate?

It's already very close to 100% dissipation, so you would not go higher on plate current.

Rather, you'd just accept very slightly less clean power output by keeping everything else the same except the OT primary impedance.

[jbefumo]

Yeah, been cypherin' and simulatin and came to the conclusion that was a bit too close for comfort.  I am, however, going to use Tung-Sol 7581As, which will give me a bet more margin of safety. Else I may just go with fixed bias. 

Pretty much what I'm going to do -- only that 4,2k OTs are easier to find -- I suspect that the difference doesn't amount to much, and figured a slightly higher plate would compensate?

It's already very close to 100% dissipation, so you would not go higher on plate current.

Rather, you'd just accept very slightly less clean power output by keeping everything else the same except the OT primary impedance.

[drgonzonm]

.......  Just a thought to consider.  Doubling the wattage only gets you a few more decibels. 

With respect, Tubenit

Sound like pH is measured in logarithms.

[jbefumo]

Yeah -- all marketing ...

.......  Just a thought to consider.  Doubling the wattage only gets you a few more decibels. 

With respect, Tubenit

Sound like pH is measured in logarithms.