6BQ5 Cathode Bias

[bobloblaws]

Have you checked that the screen and control grid voltages (and screen resistor values) are they the same for both tubes? If the imbalance follows the tubes then it's most likely the tubes. They will balance better if individual 270R cathode resistors (and capacitors) are used. That 8mA difference will come down to around 2mA if I remember correctly and it's cathode biassed so you could run the tubes at more than 70%

Edit: I see nick has already explained the 70% "rule"

Dave, do you mean to say that IF it turns out to be the tubes that are unbalanced I should disconnect the common cathode connection and bias each cathode separately with its own resistor and bypass cap? Or are you suggesting it would be advantageous to do that in any case? I think you mean the former but I wanted to clarify.

[Dave H]

Dave, do you mean to say that IF it turns out to be the tubes that are unbalanced I should disconnect the common cathode connection and bias each cathode separately with its own resistor and bypass cap? Or are you suggesting it would be advantageous to do that in any case? I think you mean the former but I wanted to clarify.

Both

I was thinking it would reduce the dissipation of the hotter tube. The heaters connect to one power tube cathode.

[Chuck H]

re: idle current, the 70% rule is usually applied to fixed bias amps. Cathode bias, and especially cathode biased el84's, are typically between 80% and 90% for AB and 100% to 120% for A. Don't forget to base dissipation figures on the plate to cathode voltage and not plate to chassis. And don't forget to account for screen current since it makes a larger proportional difference in the figures for lower wattage tubes.

I thought we determined that the higher current stays with the socket and doesn't follow the tubes? As mentioned, check screen grid resistors and control grid bias feed resistors for balance. Also possible there is a coupling cap leaking a little and skewing the bias on the hot socket.

[Bloomfield]

Hi Bob, if by any chance it is actually an A-160 rather than A-60, I posted a schematic on here a while back. I seem to recall that one side of the heater string was originally grounded, and this may have led to the confusion about the cathode resistor.

Andy

[Chuck H]

I just went and grabbed it.

[bobloblaws]

Both

I was thinking it would reduce the dissipation of the hotter tube. The heaters connect to one power tube cathode.

OK. So the bypass cap as it stands is 20uF, would I use a similar value cap for each cathode? Or closer to 10uF?

[bobloblaws]

I just went and grabbed it.

Chuck and Andy,
Thanks, the amp I have here definitely says A-60, but it is the same as this schematic for all intents and purposes. Much appreciated.My next task is to get the tremolo going so this is a great help.


B.

[bobloblaws]

Cathode bias, and especially cathode biased el84's, are typically between 80% and 90% for AB and 100% to 120% for A.

Thanks Chuck, good info to go by.

I'm going to look into what is going on with the imbalance I saw.

[pdf64]

This is just Internet lore. It's only a guideline for amps that are operating in class AB...

I would re-phrase that as 'the set bias to idle at up to 70% guideline is only applicable for amps operating in fixed bias class AB'.

Also my perception is that the guideline causes misunderstanding of amp operating class, so I suggest that the note 'operating class is not defined by the plate dissipation at idle' should accompany it.

re: idle current, the 70% rule is usually applied to fixed bias amps. Cathode bias, and especially cathode biased el84's, are typically between 80% and 90% for AB and 100% to 120% for A..

Ha, Chuck spotted it earlier.
FYI, Vox AC30 are class AB but typically idle over 100%; the late 60s/70s versions, with solid state rectification and a high HT, tended to idle well above even that.

OK. So the bypass cap as it stands is 20uF, would I use a similar value cap for each cathode? Or closer to 10uF?

Whilst the signal level is such that both p-p sides are conducting (eg 'class A' operating area), the bypass cap of a shared resistor doesn't do much, because the (hopefully) balanced signals present at the control grids then mostly cancel out at the cathodes; the bypass cap only really comes into effect when a tube gets pushed into cut off.
When that happens, the absence of its plate/cathode current prevents the signal from its control grid appearing at its cathode, and so the cancellation stops and the bypass cap then has a job to do.
Whereas with individual cathode resistor and bypass caps for each cathode, the bypass cap is needed at all signal levels. 20uF is a pretty small value anyway, and as explained above, it's only ever really bypassed one cathode at any one time.
So I wouldn't reduced its value much, if at all, as even though the cathode resistor value has increased, the total effective cathode resistance (internal cathode resistance // cathode resistor) probably hasn't increased that much.

[Dave H]

OK. So the bypass cap as it stands is 20uF, would I use a similar value cap for each cathode? Or closer to 10uF?

I think you need bigger caps when the cathodes are separated or the bass will be rolled off early. LTspice tells me 10u caps and 270R resistors give a -3dB point at 200Hz. I used 270R resistors and 220u caps because that's what I had. It sounded the same as the single 150R and 220u I was using before.

[bobloblaws]

I think you need bigger caps when the cathodes are separated or the bass will be rolled off early. LTspice tells me 10u caps and 270R resistors give a -3dB point at 200Hz. I used 270R resistors and 220u caps because that's what I had. It sounded the same as the single 150R and 220u I was using before.

Hey thanks, Dave, as a matter of fact I replaced the 20uF with a 50uF per the schematic that was provided. I also decided to leave well enough alone regarding the common cathode resistor. When I swapped positions on the two power tubes I ended up with plate currents of 36mA and 38mA so I figured that was close enough for rock and roll. Plate to cathode voltage is 378 so it's running near 120% dissipation which apparently is fine for cathode bias EL84's. The schematic shows that it is in fact push-pull with a cathodyne phase inverter. My understanding is that push-pull is usually class AB but in some cases can be class A. I don't know how to determine which class this amp is, I imagine it can be determined with a scope. I have a scope but I'm still green when it comes to most operations. So unless i'm told different, I'm assuming the bias is good and is as intended by design.

[pdf64]

...I don't know how to determine which class this amp is...So unless i'm told different, I'm assuming the bias is good and is as intended by design.

See post #12 for a means of assessing amp operating class.
'drive the output to clipping' means that a dummy load be connected in place of the speaker, the amp's controls should be turned up high, and a small sine wave be connected to the amp input, with the level of the signal being adjusted to push the amp's output signal (ie scoped) to the onset of clipping.
Your amp may just tend to round the sine peaks off, rather than create hard clipping, so (without a distortion meter) deciding which point is the max clean output may be a judgement call.
Whatever, whilst the amp is putting out its max clean signal, measure the Vdc across the cathode resistor and compare it to the Vdc at idle; if it's more than ~10% higher, class AB is indicated.

When making any voltage measurements on an amp, the first step is to measure the mains and heater Vac; the heater voltage is our best indication of whether the mains voltage is around the level that the amp was intended for. If the mains is higher than the PT designer was expecting then all secondary voltages will follow suit.

[bobloblaws]

See post #12 for a means of assessing amp operating class.
'drive the output to clipping' means that a dummy load be connected in place of the speaker, the amp's controls should be turned up high, and a small sine wave be connected to the amp input, with the level of the signal being adjusted to push the amp's output signal (ie scoped) to the onset of clipping.
Your amp may just tend to round the sine peaks off, rather than create hard clipping, so (without a distortion meter) deciding which point is the max clean output may be a judgement call.
Whatever, whilst the amp is putting out its max clean signal, measure the Vdc across the cathode resistor and compare it to the Vdc at idle; if it's more than ~10% higher, class AB is indicated.

When making any voltage measurements on an amp, the first step is to measure the mains and heater Vac; the heater voltage is our best indication of whether the mains voltage is around the level that the amp was intended for. If the mains is higher than the PT designer was expecting then all secondary voltages will follow suit.

Ahh, yes. I will give that a go, thanks!

Edit: Cathode voltage at idle is 10.66V and at clipping it is 11.61V. That works out to approximately 9% difference so I'm gonna call it Class A.

[bobloblaws]

I just went and grabbed it.

Edit: please disregard, I found this thread discussing the same trem circuit, I'll start there.

http://music-electronics-forum.com/t45620/

Sorry to go off topic, can you suggest any troubleshooting for the Tremolo? I tried swapping out the caps in the trem circuit as well as the 6AT6 but still nothing. The circuit looks very similar to the blackface Princeton style but there is no negative bias voltage line that comes into play. Is this one designed to modulate the B+?

[bobloblaws]

I returned the amp chassis to its rightful owner and got a look at the cabinet it came out of. To my mild consternation I discovered the schematic glued to the inside bottom. Attaching for reference.