Fender 5f6a build low output & distortion

[steves3972]

here are my voltage readings, pin 1 & pin 4 on both 6l6s seems low to me, most of everything else seems a little high, not sure about pins 2 and 7 on the v3, looks like they're half of what they're meant to be as well.
v1
1.180
2.0
3.2
4.fil
5.fil
6.178
7.0
8.2
9.fil

v2
1.188
2.0
3.1.3
4.fil
5.fil
6.371
7.188
8.198
9.fil

v3
1.280
2.15
3.53
4.fil
5.fil
6.261
7.17
8.53

6l6
1.6.1mv
2.fil
3.494
4.330
5. -40
6.491
7.fil
8.6.1mv

6l6
1.6.3mv
2.fil
3.491
4.328
5. -40
6.489
7.fil
8.6.3mv

[Firestorm]

Pin 1 of a 6L6 has no connection so it doesn't matter what it reads. It's usually grounded along with pin 8 so what you're reading there is just noise or the margin of error of your DMM. Pin 4 does look low. What value is your screen resistor?

Is this pretty much a stock build? Your supply voltages are obviously higher than a stock 5F6A, but that's your transformer.

[steves3972]

thanks Firestorm, using the ceriatone layout for this build with classictone transformers.
The screen resistors are 470ohm 2w (just double checked them)

Should I try a different value?

[Firestorm]

Should I try a different value?

No, but your voltages imply that the screens are pulling 342mA, which is about 10 times what the plates should pull at idle, so something ain't right. Double-check all your output tube wiring. For example, pin 6 is used as a tie point for the screen resistors and both pin 6s are connected to the same point and should have exactly the same supply voltage on them. Yours don't. Have you cross wired something?

[Firestorm]

Also, considering your B+ voltage, -40 negative bias voltage seems low. An original 5F6A had it's plates at about 430VDC and had a bias voltage of around -48. I'm thinking yours might want to be more like -50 to -55.

Are you seeing any redplating?

Do you have a 1-ohm current sense resistor on the cathode? If so, then the 6.3mV reading on pin 8 is meaningful: you're way overbiased. Doesn't explain the extreme behavior of the screens, though.

[steves3972]

checked the wiring to both 6l6s and pin 6 on both actually read exactly the same, I must have measured it wrong originally. I have a bias pot that was turned right down so I turned it to almost half way and the bias now reads -50 but all voltages have gone up now too. There is a 1ohm resistor on the cathodes as well. I'm not sure I'm using my voltmeter correctly for the readings on pin 1 & 8, could it actually be 33mv not 3.3mv

both 6l6s read
1. 3.3mv
2. fil
3.518
4.353
5.-50
6.519
7.fil
8. 3.3mv

I don't think I should run the bias too hot for now though if you think I am cooking the tubes, no red plating though either. I am using some mesa 6l6gc tubes, not sure if this is worth mentioning or not.
So if it's all wired correctly, what else could it possibly be?

[surfsup]

No, but your voltages imply that the screens are pulling 342mA, which is about 10 times what the plates should pull at idle

...that 2W resistor would be toast with 342mA through it...

[Firestorm]

checked the wiring to both 6l6s and pin 6 on both actually read exactly the same, I must have measured it wrong originally. I have a bias pot that was turned right down so I turned it to almost half way and the bias now reads -50 but all voltages have gone up now too. There is a 1ohm resistor on the cathodes as well. I'm not sure I'm using my voltmeter correctly for the readings on pin 1 & 8, could it actually be 33mv not 3.3mv

both 6l6s read
1. 3.3mv
2. fil
3.518
4.353
5.-50
6.519
7.fil
8. 3.3mv

I don't think I should run the bias too hot for now though if you think I am cooking the tubes, no red plating though either. I am using some mesa 6l6gc tubes, not sure if this is worth mentioning or not.
So if it's all wired correctly, what else could it possibly be?

I wasn't sure about the 1-ohm resistor. That changes the importance of the cathode reading. If it was 6.3mV originally, that means the tube is passing only 6.3mA of current, so the tube plate is idling somewhere around close to 3W -- almost off. When you increased the negative voltage to -50, the idle dissipation became 1.7W. At that point the tube really should be completely off (or nearly so). Sometimes, when a power supply is marginal, you will see large changes in the plate voltage as you adjust bias for more or less current. Yet here, you've gone from plate volts of 491 to 519 with only a very small change in current (just 3mA).
And, with the tube biased essentially to cutoff, your screens should not be drawing current either. (And as surfsup observes, if you really are passing 342mA through a 2W 470R resistor, you should be able to have a weenie roast in the chassis.)

Let's try a few things.

Make sure you have fresh batteries in your meter. Remove the 6L6GCs and (with the amp off) confirm that the screen resistor reads 470 ohms (rather than 4,700 or 47K). Confirm that the cathode resistor reads 1-ohm (rather than 10).

Then with the output tubes still out, power on the amp and take readings with respest to ground at the output tube sockets. Since no current is flowing (no tube), pin 3 should show the full B+ (same as at the center tap), pins 4 and 6 should both show the full screen supply voltage (same as from the choke output) and pin 8 should be 0 volts.

Then put the tubes back in and take readings again (amp on) and we'll compare. Make sure you know how to read the small voltage on the cathode; meter should be at its lowest DC volts setting (unless it's auto-ranging); DC not AC, millivolts not milliamps.

Let us know what you find.

[The New Steve H]

Make sure the OT output wires (#3 pins on the 6L6s) go to the correct tubes. This is something I screw up all the time. You will get distortion and low output if you have them connected wrong.

If I were you, I would take a couple of minutes and switch them just to make sure.

[steves3972]

ok so I checked the screen resistors again and they actually were the wrong value, thanks so much for that. The cathode resistors appear to be reading closer to 2ohms with the voltmeter on the lowest setting, I am getting no reading on pins 1 & 8 of the 6l6s. I now have 470ohm resistors on the screens and all other voltages are reading fine now. I plugged in my guitar last night and the amp worked quite well but this morning after playing for a little while the amp started to cut out, and now it's sounds as if a tube is bad or something (low output farty sounds, worse than before). Have replaced each of the preamp tubes one at a time and there was no change. So as I said, all voltages look good now but the bias reads -57 and it does not change when turning the bias pot. This happens with the tubes in or out, why is that now the case?
Here are the voltages anyway

v1
1. 148
2.0
3. 2
4.
5
6.143
7.0
8.2

v2
1.153
2.0
3.1
4.
5
6.296
7.153
8.161

v3
1.218
2.12
3.42
4.
5
6.217
7.14
8.42

6l6
1.0
2
3.399
4.398
5.-57
6.398
7.
8.0

6l6
1.0
2
3.400
4.398
5.-56
6.397
7
8.0

[Firestorm]

With the negative voltage that high, the tubes are likely in cutoff, so there wouldn't be any reading on pin 8. Did you happen to check what the grid voltage was when the amp was working?

Hard to say what happened, but the only thing to do is check whether the pot is still functioning and especially whether there is continuity to ground from the righthand tab (looking at the back of the pot). It sounds like you have the raw bias voltage being applied to the tubes.

[Jana]

Unless you have a really good multimeter and know how to use it to measure low resistances to check a 1 ohm resistor, chances are your 2 ohm reading is off. About half of that could be the resistance of your test leads.

It sounds like once you get the bias issue sorted out and bias the amp, you will be on your way to happy tunes.

[steves3972]

ok, so right this very moment the amp is working again, the bias is back down to -41 and the cathode readings at pins 1 & 8 are 5.4 and 5.1mv.
I think I just need to double check for some loose solder joints. The other things that concerns me is volume output of the amp, seems not loud enough for a 50w amp. On both channels there is no sound until the volume pot hits about 2 1/2. I can play it on 10 and not go deaf. The amps sounds very nice other wise, very quiet. I have a 12ay7 in v1.

[Firestorm]

Assuming your 1-ohm sense resistors are fairly accurate (these really should be very good resistors: 5% resistors are just okay, 1% are better), with 491 plate volts the cathodes should likely read something like 32mV, rather than 5.4 or 5.1. Unless the drive signal is heavily attenuated for some reason, I would think the cross-over distortion from being biased so cold would be anything but "very nice."

As Jana said, it can be difficult to measure such a small resistance without a good meter and some experience at "zeroing out" the lead resistance.

Perhaps you can try to adjust the bias in small steps and dial it in by ear. Even with a 12AY7 in V1, the amp should be very loud.

[steves3972]

ok, so i'd say it's a 5% resistor as I got it from jaycar, I will be able to borrow a decent voltmeter in a day or 2 but until then lets say it is actually 1ohm.
I have added a photo of the resistor. It's colours are brown-black-gold-gold.
Would the fact that it's 5% and not 1% be enough to effect the reading on the cathode? Can I remove the resistor all together and just ground pin 8? I should also mention that the 6l6 tubes are not brand new, I actually have no idea how old they are. Should I buy some new ones?
Any recommendations?