Huge difference in bias readings from tube to tube on matched sets

[w302nv]

Ok, so I got the bias probe out and put a nice set of matched RCA Black Plates in my 67 Super Reverb....It just got it back from a shop a few months ago..I was never thrilled with the tone......I thought I would just swap the tubes first and see if anything improved.....I dont know how it behaved before it went in as it was almost a basket case...

The plate voltage is the same (~460) but one socket draws ~2x as much current as the other (initially 16mA and 33mA didn't get to adjust...just shut it off)...this holds true no matter what tube is in the socket and holds true across different sets (other matched RCAs and GEs)...

Both of the 470 Ohm 1W resistors (Screen...right?) measure right around the same and are in spec, and both look new....
The 1k5 1/2 watt resistors (Grid...right?) are at 1k9 and 2k and look to be original to the amp....drifted too far?

Other kind folks have suggested the coupling cap...(would a .1uf 400v Polyester Film like a MKT1813410405 work as a replacement?) Could a bad socket cause this? What about the 220k, 100k, or the 8k2 resistors?

I am a total n00b but can solder effectively, and know how to keep from getting killed or smoking the amp...just dont know exactly where to start or what to measure. Any help is appreciated...in the meantime I am going to re-tension the socket as it seemed to be a bit loose...more so than the other one.

[Tom Phillips]

...The plate voltage is the same (~460) but one socket draws ~2x as much current as the other (initially 16mA and 33mA didn't get to adjust...just shut it off)...this holds true no matter what tube is in the socket and holds true across different sets (other matched RCAs and GEs)...

First. There is no need to immediately turn the amp off. 16/32ma at 460 plate voltage won’t do any damage.
OK. So no matter which tubes you install the idle current is low at one power tube and higher at the other and the SAME socket position is always low and the other is always high. Right?
If so, this tells us that the problem is with the amp.
Or...Is there any chance that you have a 1968 SR with a bias balance control?
What does the tube chart list for model number and rectifier tube type?

Both of the 470 Ohm 1W resistors (Screen...right?) measure right around the same and are in spec, and both look new....
The 1k5 1/2 watt resistors (Grid...right?) are at 1k9 and 2k and look to be original to the amp....drifted too far?

Good checking. These parts are not causing your problem.

Other kind folks have suggested the coupling cap..

A leaky coupling cap could cause a problem but let’s not guess. Mote later when you verify my first question.

What about the 220k, 100k, or the 8k2 resistors?

They would not cause the described problem.

It sounds like you are comfortable with making voltage measurements with the amp turned on. If you are sure you can do that safely then take the power tubes out for now, turn the amp on and measure the voltage at pin 5 of each power tube socket. Please report the values you measure. The voltage will be negative with respect to ground. Note that the plate voltage will be higher under this condition. Under ideal conditions with a BF circuit, the voltage on each tube socket will be equal.

[Tom Phillips]

One more thing. You said the amp was almost a basket case when it went to the shop. Can you post photos of the current condition inside?

[w302nv]

OK. So no matter which tubes you install the idle current is low at one power tube and higher at the other and the SAME socket position is always low and the other is always high. Right?
If so, this tells us that the problem is with the amp.

Right..the problem stays at the socket (one closest to the rectifier) and does not follow the tubes, and exists across multiple "matched" sets of different 6L6GCs

Or...Is there any chance that you have a 1968 SR with a bias balance control?
What does the tube chart list for model number and rectifier tube type?

This is definitely a 1967...of note may be that it has a later PT (someone suggested it was from 1973) and has the requisite virtual center tap off the lamp socket thanks to the last tech that worked on it.

Pics to follow (Thanks Tom)

[Tom Phillips]

Right..the problem stays at the socket (one closest to the rectifier) and does not follow the tubes, and exists across multiple "matched" sets of different 6L6GCs

This is definitely a 1967...of note may be that it has a later PT (someone suggested it was from 1973) and has the requisite virtual center tap off the lamp socket thanks to the last tech that worked on it.

The new PT is not an issue. However, you never know what you are going to find in old amps that have been through a basket case phase.

Sounds like we need to troubleshoot the bias circuit. It will be interesting to know the voltage on pin 5 of the power tubes. My guess it that they are not equal. Could be caused by someone messing with the circuit. Could be a coupling cap from the phase inverter that is leaking DC.

Tom

[w302nv]

Hi Tom,
Both are at -49.3VDC running on standby...should be the same either way right? Can/should I measure with full voltage applied?
Here are the pics. Hope they help...cant find the "good" camera so had to use my phone.

Before:




After:

[Tom Phillips]

Hi Tom,
Both are at -49.3VDC running on standby...should be the same either way right? Can/should I measure with full voltage applied?
Here are the pics. Hope they help...cant find the "good" camera so had to use my phone.

From the before & after photos I see that the amp was cleaned up quite a bit. Looks like some heavy duty chemical was used because the labels on the blue molded caps seem to have been cleaned off during the process.

IMHO it appears that he bias supply is working. Make the measurements again with the amp in operate mode. If the voltage on pin 5 of one of the tubes shifts significantly then it would indicate that one of the 0.1uF coupling caps between the phase inverter and the power tubes is leaking DC. Let me know what you measure. Also measure the screen voltage on pin 4 of the power tubes.

Regards,
Tom

[w302nv]

From the before & after photos I see that the amp was cleaned up quite a bit. Looks like some heavy duty chemical was used because the labels on the blue molded caps seem to have been cleaned off during the process.

If I remember correctly, he said the board was removed...the labels are still there they are just black because of something the tech put on there to prevent the board from drying out.

IMHO it appears that he bias supply is working. Make the measurements again with the amp in operate mode. If the voltage on pin 5 of one of the tubes shifts significantly then it would indicate that one of the 0.1uF coupling caps between the phase inverter and the power tubes is leaking DC. Let me know what you measure. Also measure the screen voltage on pin 4 of the power tubes.

They are still at -49VDC..the screens are at 515 & 512VDC...All voltages are consistent on the other pins (3,4, and 6 are the same...1 and 5 are the same) from tube to tube

Any chance the bias probe could be skewing the results? It works just fine on my Princeton Reverb....It is an Amp-Head Dual Bias Tester-MPD I Guess I can answer that by using just one side of it at a time....

Edit: Things stay the same no matter what

[g1]

I'm a little suspicious of those 1K5 grid resistors. 30% off and who knows how far off when under voltage. Maybe not enough to create the discrepancy you are getting but I would certainly replace them.

[MWJB]

If the 100K & 82K PI plate resistors were drifted you would have a big mismatch at pins 1 & 6 of the PI tube, these should be within 10-15vdc of each other.

If the 220Ks were seriously drifted then the negative dc voltage would by uneven.

Yes, the 0.1 MKT metallised poly caps are fine (yellow jobs from RS?).

Have you checked the OT? Check DC resistance from each power tube plate to centre tap. Do you have a variac/low voltage AC source to check turns ratio?

[w302nv]

If the 100K & 82K PI plate resistors were drifted you would have a big mismatch at pins 1 & 6 of the PI tube, these should be within 10-15vdc of each other.

Would this impact bias current? If so is it OK to measure from the tube side of the 12AT7? With the power tubes out and off standby with full voltage applied? Sorry for the dumb question...remember, I'm a n00b

If the 220Ks were seriously drifted then the negative dc voltage would by uneven.

Good to know, they were both at -49VDC

No worries there then

Yes, the 0.1 MKT metallised poly caps are fine (yellow jobs from RS?).

Yep...from Mouser but the same thing

Have you checked the OT? Check DC resistance from each power tube plate to centre tap.

This is what I was afraid of....if you dont mind can you describe this procedure...Im not sure if it makes any difference, but this amp has the original OT but a PT from 1973 with a virtual center tap (2 100Ohm resistors off the lamp socket a la Silverface).

Do you have a variac/low voltage AC source to check turns ratio?

Yep but I should probably Google how to do this before I even bother you for more info

Thanks!

[Tom Phillips]

They are still at -49VDC..the screens are at 515 & 512VDC...All voltages are consistent on the other pins (3,4, and 6 are the same...1 and 5 are the same) from tube to tube [IMG]file:///C:/DOCUME%7E1/Tom/LOCALS%7E1/Temp/msohtml1/01/clip_image001.gif[/IMG]

Interesting. No help there. Do you have a bias probe with two different sensor heads or do you move the sensor head to each tube socket in turn?

(The screen voltage is reading high because the tubes are out of the sockets.)


We need to find the root cause of you apparent large discrepancy in idle condition plate current. So far. All the part values you have mentioned do not appear to be capable of causing the problem.

[w302nv]

Thanks to everyone for the insight, I really appreciate the help. The bias probe has 2 "heads" and 2 switches that switch between head 1 and 2 and between bias current and plate voltage. This is just a weird problem...the amp never sounded very good after I got it back good (now I have an idea why)....as the bias is adjusted up and down the difference seems to stay at roughly half (or double depending on how you look at it).....any advice on testing the OT?

[Bruce / Mission Amps]

If you check the bias probe (I never use one) and it is working correctly it can only be a few things that can cause this..... I read you've checked all those.
I didn't read if you cleaned, de-ox'd and retensioned the sockets. Make sure you do that.
Anyhow, I've had this happen a few times over my career and when all the easy-quick stuff fails, and it sounds like you have failed to remedy the situation with the easy stuff, replace the tube sockets. I know, it's pain in the butt.
Just because the socket has the same bias voltage delivered to each socket does NOT mean the tube in question is seeing the correct voltage.

Added: Yes a bad OT can do weird things to your idle current ... but so can an ultrasonic oscillation on one tube that you can not hear but the tube is drawing "apparent" DC idle current, when in fact it is a lot of AC (ultrasonic audio) power going out the tube.....

[Tom Phillips]

...The bias probe has 2 "heads" and 2 switches that switch between head 1 and 2 and between bias current and plate voltage....

If you swap the position of the bias prob heads between the tube sockets does the high/low current reading move too?
Does your bias probe appear to work correctly in your other amps?
Just trying to cover all the possibilities here.

Tom