Voltage races too high

[pbman1953]

I haven't  had time to try and rid my Traynor of some noises for a while. I went back to it last night and found that I had a noisy tube and I replaced it. I re-biased and even played the amp for 20 minutes with out an issue and it sounded great. I came back to it today to do a last minute bias check and one tube flared up and died. Plus the amp blew a fuse. I took the tube out replaced it and went back to rebias again and I see the tube closest to the power side is racing at 1000 volts on my Bias Rite meter. The others are still calm at 550 or so.




 This amp was fine now V7 is out of control. The worst part is I lost a perfectly fine SED 6550, that hurts.


Any steps to take


Thanks

[John]

I would check R's 24&25, 28&29, make sure one of 'em isn't open, and also those diode looking things (dunno what they are, sorry for my ignorance).

[shooter]

If it's mine, I would take the time, install 1ohmer for each power tube, pull all power tubes and see where your bias V is at.

[jjasilli]

Tubes flaring seems like: i) tube went bad & shorted internally; or ii) loss of bias voltage; or iii) those diode things failed as John said.  Ironically, I believe their purpose is to protect the power tubes.

[pbman1953]

This is so odd. I checked all the resistors from the bias feed to each tube and they measured to what they should be. I took out the tubes to keep them safe and used my meter to see what the voltages would be. Still the one tube socket is at 1000 volts and all the others are where they are supposed to be at 554 voltages averaged.
This is so confusing because there is only one line from the bias section  then branches out to the 4 tubes.

[sluckey]

The highest voltage in that amp is approx. 550vdc. Where do you suppose the 1000V reading is coming from? I suspect your bias rite is lying.

[pbman1953]

The highest voltage in that amp is approx. 550vdc. Where do you suppose the 1000V reading is coming from? I suspect your bias rite is lying.

You could right on that but I swapped sockets to see if it would move

[sluckey]

Completely remove the bias rite and check voltages on every pin of the suspect tube socket. Do you see 1000V on any pin?

[pbman1953]

Plug "D" on the Bias Rite is bad internally according to CJ At Weber. He's sending me a replacement.




Thanks!

[shooter]

is bad internally

That's why I recommended the 1 ohm method, and as a bonus, if they ever do go bad, it'll be externally visable 

[John]

Plug "D" on the Bias Rite is bad internally according to CJ At Weber. He's sending me a replacement.




Thanks!

Well doggone it. I spent a lot of brain power on that one. 

[SoundmasterG]

Plug "D" on the Bias Rite is bad internally according to CJ At Weber. He's sending me a replacement.




Thanks!

To avoid that problem in the future, be sure that you are only using the Bias Rite to set the bias and not to play the amp. Once the bias is set, take the Bias Rite out of the equation and put the power tubes back in directly to the sockets, and then play. When you play with the tubes plugged into the Bias Rite, you can exceed the wattage of some of the parts in the Bias Rite. It is only designed to check the bias and not to play the amp while it is plugged in. If you want to be able to play while monitoring the bias, then add some 1 ohm resistors from cathode to ground as mentioned previously and monitor with a DMM.

Greg

[pbman1953]

Hey Greg,




If I ever leave the plugs connected it's only during an extreme low volume for maybe 5 minutes. But will do less in the future


Thanks!

[92Volts]

It sounds like this could be oscillation, even if the Biasrite is damaged that could have been the original cause. The only way to get voltage higher than B+ is AC current through the (inductive) OPT. Meters will react differently to changing voltage (when they’re supposed to be measuring DC) but some could display peak voltage (or total nonsense).

With B+ of 550v it’s possible for there to be peaks around 1000v in mostly “normal” (but large-signal) operation. Most cheap voltmeters (and likely the circuits used in the Biasrite) are good to 1000v. So there was a mention of exceeding the “wattage” of parts of the Biasrite but that voltage might be an issue long before things start to heat up.

[pbman1953]

It was the Bias Rite socket, because I used another socket and it was fine. In turn , Webber VST set me a new socket.

[92Volts]

It was the Bias Rite socket, because I used another socket and it was fine. In turn , Webber VST set me a new socket.

I see, I misunderstood how the Bias Rite system worked. I thought more electronics (metering) were built into the socket/adapter but it seems that's not the case.

[pbman1953]

https://www.tedweber.com/gadgets/bias-rites

[SoundmasterG]

It is possible when you have very large voltage swings as you mentioned to be an oscillation, so it is a good idea to check everything if you encounter it. I had one amp I worked on that was swinging up to 600V on a preamp tube and was an oscillation caused by the builder putting a trem oscillator stage in the same tube as a gain stage, and the wiring was routed such that it was causing positive feedback and that whole stage was going out of control.


Glad it was just the Bias Rite socket though. My Bias Rite has been very reliable over the years but I am always making sure I only use it to set the bias.


Greg

[PRR]

> I thought more electronics (metering) were built into the socket/adapter but it seems that's not the case.

My assumption was that the plug-socket was at most a small (1r?) resistor from K pin to K hole, and a connection to the P link (*maybe* with a series resistor). Then three wires to the display.

The display would read the mV across the K series resistor and call it mA. Or the volts from the P link as plate volts. (A frill would be a 1Meg inside the adapter and a 1K inside the display, so V reads in mV but is displayed as "V". This avoids dangerous voltage on the cables.)

I think Hoffman sells a kit to build this plug-socket? Resistors may not be included, but you know where to find those also. That is NOT the Weber full I & V system; you can reverse-engineer that.