Low Output On A "Blackvibe" Build

[AHeck]

This build is several months old and has behaved very well, but all of a sudden, it's lost volume and has an unfortunate distortion at low gain.  Initial signal voltage is .1VAC at 500 Hz, and all controls are set to 50%. OT is 4K Primary rated at 40W.
Thanks for helping if you get a chance.

[sluckey]

Tubes first. PI doesn't seem to have much gain. Maybe start there.

[AHeck]

I'd better put the correct schematic up, the PI is actually an ECC81. 

[sluckey]

Tubes first. PI doesn't seem to have much gain. Maybe start there.

[PRR]

Pull the output tubes, one at a time. One may be dead.

[sluckey]

I'd better put the correct schematic up, the PI is actually an ECC81.

You show 3.5VAC coming into the PI and 5VAC coming out of the PI. That's only a gain 1.4! You have -45V on the 6L6 grids. Those 6L6s want to see over 40VAC to sound loud.

Start with the PI.

[AHeck]

Thank you,  as soon as I get home I’ll try another 12at7 and see if that helps.

[WimWalther]

I believe there's an error in the schematic.. a missing connection dot.

[PRR]

I believe there's an error in the schematic.. a missing connection dot.

I agree that it should be dotted.

Especially since there are other crossings that are un-dotted and un-hop-overed.

But it is very unlikely to show those voltages if that crossing is un-connected. The 'wrong' voltage on grids is what you get with that connection and a finite-impedance (~~10Meg) meter. Without the connection the grids would collapse to near zero.

[AHeck]

Right, that connection should very much be dotted, good looking out.

[AHeck]

I measured my heater voltage and it is reading 2.2VAC on both legs.  Is this reading off because I've elevated the filament center tap to 60VDC or is my PT fried? I'm going into the power tubes with 53V signal on one tube and 68V signal on the other but only getting an AC reading of 7VRMS at the plates and 1.6VAC at the transformer secondary. Thanks.

[WimWalther]

I measured my heater voltage and it is reading 2.2VAC on both legs.  Is this reading off because I've elevated the filament center tap to 60VDC or is my PT fried? I'm going into the power tubes with 53V signal on one tube and 68V signal on the other but only getting an AC reading of 7VRMS at the plates and 1.6VAC at the transformer secondary. Thanks.

The common factor here appears to be measurement error, being that none of those figures seem quite right.

I don't wish to insult your experience or intelligence, but are you sure your meter is set correctly, and you've chosen proper ground / common points?

Also, what is the AC voltage between filament pins on the sockets? There's an odd tendency around here to measure each leg separately.

[tubenit]

I have gotten incorrect readings before on my voltmeter when the battery was low.


With respect, Tubenit

[AHeck]

The common factor here appears to be measurement error, being that none of those figures seem quite right.

I don't wish to insult your experience or intelligence, but are you sure your meter is set correctly, and you've chosen proper ground / common points?

Also, what is the AC voltage between filament pins on the sockets? There's an odd tendency around here to measure each leg separately.
[/quote]
I'm sure that measurement errors are happening here on my end, they were with the heater, which is apparently fine. I've gone back and taken measurements again and confirmed everything else.  I was hoping you might tell me what seems most out of whack so I can try again, maybe correcting an error in the process.  This may be wrong, but I'm measuring AC signal at the power tube plates by probing  the tube pin and using chassis as ground and my meter set to AC voltage, all other signal voltages I've gotten with an oscilloscope with chassis as ground.  Please let me know if I'm doing this wrong; I'm here to learn.  Thank you.

[sluckey]

Fixed quote

The common factor here appears to be measurement error, being that none of those figures seem quite right.

I don't wish to insult your experience or intelligence, but are you sure your meter is set correctly, and you've chosen proper ground / common points?

Also, what is the AC voltage between filament pins on the sockets? There's an odd tendency around here to measure each leg separately.

I'm sure that measurement errors are happening here on my end, they were with the heater, which is apparently fine. I've gone back and taken measurements again and confirmed everything else.  I was hoping you might tell me what seems most out of whack so I can try again, maybe correcting an error in the process.  This may be wrong, but I'm measuring AC signal at the power tube plates by probing  the tube pin and using chassis as ground and my meter set to AC voltage, all other signal voltages I've gotten with an oscilloscope with chassis as ground.  Please let me know if I'm doing this wrong; I'm here to learn.  Thank you.

[AHeck]

Thank you Sluckey. 

[sluckey]

I'd better put the correct schematic up, the PI is actually an ECC81.

You show 3.5VAC coming into the PI and 5VAC coming out of the PI. That's only a gain 1.4! You have -45V on the 6L6 grids. Those 6L6s want to see over 40VAC to sound loud.

Start with the PI.

I'm surprised you have not responded to this post.

[AHeck]

I'm sorry.  I did replace the tube and I'm getting, at this moment, 49VAC and 45VAC at the grids on the 6L6s, but I'm only getting about 20VAC at the plates of the power tubes.  Everything in peak to peak.  I've tried three sets of tubes.

[sluckey]

I'm getting, at this moment, 49VAC and 45VAC at the grids on the 6L6s, but I'm only getting about 20VAC at the plates of the power tubes.

You have a healthy signal at the grids. Don't be concerned about the AC voltage on the plates. The output tubes are all about current, not voltage. What is the DC voltage on pins 3 and 4 of each 6L6? What is the current through the tube?

[AHeck]

It was 445 and 447 with 60mA through each tube but that's high, right?
So I adjusted the bias and got these numbers...

Pin 3-433 Both
Pin4-435  Both
45mA and 42mA

[sluckey]

That all looks good too. Is the speaker connected to the correct speaker jack?

[AHeck]

It has one speaker out switched to 8 ohms pushing two 16 ohm speakers in parallel, but it does have a switched attenuation circuit on the way to the speakers which I bypassed with no improvement to the low output volume.

[AHeck]

This is the plate of one of the 6L6s.  A perfectly good sine wave goes in and this comes out.

[sluckey]

See that USB port just below the bottom/right corner of the screen? Stick a thumb drive in that port and save the screen to the thumb drive. Then upload that pic instead of taking a pic with your phone. The pic will be very sharp and we can read everthing on the screen.

Your pic would look similar to this...

[PRR]

Don't 'scope plates, especially power plates, carelessly. Aside from blowing the 'scope or the probe, the HIGH! signal voltage on the added 'scope lead will radiate back into low-level stages and cause oscillation, howl-around.

Any audio-frequency action on the plates WILL reflect into the speaker secondaries at safer low voltage.

(Yes, >10kHz may not couple but let the hi-fi guys fret that. In guitar we can and often prefer to damp supersonics with like 500pFd across driver plates.)

[shooter]

WHAT he said^^^^^
always take your PA "plate" signal from the speaker side of the OT, the signal will be REAL close to the plate without loading, induced nonsense from test equipment, amplitude will be different, calculated power stays about equal since you've converted voltage gain of the PA to power at speaker.

[AHeck]

So, for what it's worth, the OT primary (Lower) and secondary (Top).

[shooter]

now you can use a 4 function calculator on the secondary, probably need a math simulator for the primary 

[WimWalther]

now you can use a 4 function calculator on the secondary, probably need a math simulator for the primary 

So he sees LF AM on the input (pri) side, and a nice clean sine on the output (sec) side?

Maybe I'm not quite grokking this.

[sluckey]

So he sees LF AM on the input (pri) side, and a nice clean sine on the output (sec) side?

Maybe I'm not quite grokking this.

The low freq component is the 120Hz power supply ripple and should appear same phase on each plate. Common mode rejection properties of a push/pull power amp removes that ripple so it is not passed on to the secondary.

[WimWalther]

The low freq component is the 120Hz power supply ripple and should appear same phase on each plate. Common mode rejection properties of a push/pull power amp removes that ripple so it is not passed on to the secondary.

I understand CMR well enough, I just didn't expect such a large ripple component. Then again, I can't really interpret the vertical scale correctly, so the signal might be quite small in relation.

95% of all my experience is with OTL amps, so I'm really not that used to viewing transformer amp signals.

[shooter]

I'm really not that used to viewing transformer amp signals.

I came here comfortable measuring big voltage, 1st time I seen 700vac at the PA plates, I asked "Ummmm??"  staying on the secondary became the default answer.  your test equipment will love you for it, CPR becomes way less prevalent 

[WimWalther]

I'm really not that used to viewing transformer amp signals.

I came here comfortable measuring big voltage, 1st time I seen 700vac at the PA plates, I asked "Ummmm??"  staying on the secondary became the default answer.  your test equipment will love you for it, CPR becomes way less prevalent 

Was this maybe meant for the OP?

I was just commenting on the scope. screenshots he posted.

[acheld]

So, for what it's worth, the OT primary (Lower) and secondary (Top).

These signals don't make sense to me.   Where exactly are your test probes inserted?  Picture or schematic would be helpful.

[AHeck]

I’ll get some pics up but my ground is chassis and for the primary/plate measurement, I probe pin three of the tubes and for the secondary I probe the speaker out.

[acheld]

On the primary screenshot, it appears your scope may be set up for a 100X test probe ??  Those are not so common -- I'd expect to use a x1 probe in that position, as you should be seeing less than 20VAC.  It's also possible that the scope setup does not match the probe you are using.

[shooter]

I believe he's measuring PA plates, 100X probe IS the probe of choice

[acheld]

I believe he's measuring PA plates, 100X probe IS the probe of choice

You're quite right if that is where he is measuring.  Of course. 

I'm questioning where and how the measurement is being taken.

It's clear I don't have much to offer here . . .

[AHeck]

I believe he's measuring PA plates, 100X probe IS the probe of choice

You're quite right if that is where he is measuring.  Of course. 

I'm questioning where and how the measurement is being taken.

It's clear I don't have much to offer here . . .

I'm sorry I guess I didn't explain well enough, thanks for commenting.  I am measuring using a 100X scope on my oscilloscope probing pin three of the tube directly AC coupled to find the  signal voltage at the OT primary. And apparently, the circuit doesn't like it.  I'm just trying to figure out why I've got such low output with such low overhead all of a sudden.

[AHeck]

So to sit at about 40W output, 40W=v^2/8, or something like 18VRMS at the secondary, and the Voltage at the primary would have to be higher yet because of the turn ratio of the transformer and the way the amp is operating, what class it is operating in.  I'm obviously not getting the required voltage at the speakers.  If the OT is faulty, could that cause a lack of gain in the power tubes if not just plain fail to correctly "transform the voltage and current?  I'm thinking that I may ought to isolate the OT and test it.  I've been under the impression though that OTs are not usual suspects.  Any thoughts?

[AHeck]

This is the scope measurements when I feed the OT 20Vp-p at 1K hz.  Am I wrong to expect more like 320mV at the secondary? 

[shooter]

it's a step-down transformer, the turns ratio will get you a ballpark idea of "what to expect"
if it's a 30:1 then 20vdc/30 = .67vdc

[glass54]

Sorry Shooter but its AC 
"20vdc/30 = .67vdc"
I apologise for such a petty comment. You desperately need morning coffee!
Kind regards
Mirek

[shooter]

was 4AM!
drinking a green monster now

[AHeck]

It's been a minute, but I was waiting for the delivery of a replacement OT.  It arrived and I installed it and the issue is fixed.  The amp is putting out ~17VRMS into 8ohms and sounds glorious once again.  Thanks to everyone who helped.

[WimWalther]

It's been a minute, but I was waiting for the delivery of a replacement OT.  It arrived and I installed it and the issue is fixed.  The amp is putting out ~17VRMS into 8ohms and sounds glorious once again.  Thanks to everyone who helped.

Rock on! ;-)

Good work, glad to hear it. 17Vrms is a healthy 36W @ 8R, just about right for a pair of 6L6GCs.