Hi all,
Just a quick question regarding cathode bias voltage and, in particular, stability.
After catastrophic failure caused by an overused output tube (runaway cathode voltage). I rebuilt the amp and installed meters to indicate what the cathode bias voltage was on the KT66's (as pictured).
Generally, they do vary slightly (albeit always 1v apart) but occasionally, the left channel (blue) will skip very quickly between 38v - 39.5v. It sort of cycles from 38v then jumps quickly to 39.5v then back down, then back up. This happens over a few minutes and all the time the right channel (red readout) is stable. After a few minutes, it becomes stable too.
Happened today shortly after switch on, yesterday after about 2hrs use.
Schematic attached.
Just a quick question regarding cathode bias voltage and, in particular, stability.
After catastrophic failure caused by an overused output tube (runaway cathode voltage). I rebuilt the amp and installed meters to indicate what the cathode bias voltage was on the KT66's (as pictured).
Generally, they do vary slightly (albeit always 1v apart) but occasionally, the left channel (blue) will skip very quickly between 38v - 39.5v. It sort of cycles from 38v then jumps quickly to 39.5v then back down, then back up. This happens over a few minutes and all the time the right channel (red readout) is stable. After a few minutes, it becomes stable too.
Happened today shortly after switch on, yesterday after about 2hrs use.
Schematic attached.
Attachments
No actual question was stated, but if these are new tubes, just burn them in at least 100 hours before more bias checking.
If they are well used tubes, and vary somewhat similarly, it could be your AC line.
If you really want to isolate the problem, ground the output tube grids and observe the bias current.
If problem now is gone, there is LF crud from the first stage causing the erratic behavior, which
indicates better supply filtering is needed, or possibly the feedback resistor is the culprit.
If they are well used tubes, and vary somewhat similarly, it could be your AC line.
If you really want to isolate the problem, ground the output tube grids and observe the bias current.
If problem now is gone, there is LF crud from the first stage causing the erratic behavior, which
indicates better supply filtering is needed, or possibly the feedback resistor is the culprit.
OK, well the question was whether the varying cathode voltage was normal, especially as it only affects the one channel.
Tubes, at a considered guess, would have over 100hrs but not by a great margin,
AC line could be ruled out as on 1 channel is affected.
I did however, yesterday, install another 6n1p. After a couple of hours, one channel skipped voltage as described in 1st post. Then, before switch on today, I changed the 6n1p back to what I was using originally. The voltage skip occured again at switch on, but went after a few minutes and has not returned.
Therefore, as you say, LF crud from 1st stage (or specifically an old 6n1p)
Thanks.
Tubes, at a considered guess, would have over 100hrs but not by a great margin,
AC line could be ruled out as on 1 channel is affected.
I did however, yesterday, install another 6n1p. After a couple of hours, one channel skipped voltage as described in 1st post. Then, before switch on today, I changed the 6n1p back to what I was using originally. The voltage skip occured again at switch on, but went after a few minutes and has not returned.
Therefore, as you say, LF crud from 1st stage (or specifically an old 6n1p)
Thanks.
Also consider the possibility of bad tube pin oxidation, bad sockets, bad solder joints, etc.
Also consider that the coupling capacitor to the grid may be going leaky.
I did upgrade/rebuild so the coupling caps are new (doesn't mean they can't leak though).
Will continue to monitor, no more fluctuations since replacing the 6n1p though..
Will continue to monitor, no more fluctuations since replacing the 6n1p though..
OK, so another night of stability with the previous 6n1p which, as rayma suggests, indicates a bad tube that caused the fluctuations.
I'm glad I put these volt meters on the output tube cathodes. Visual indicators that something is wrong, not only with the output tubes but also the driver.
Thanks for the advice
I'm glad I put these volt meters on the output tube cathodes. Visual indicators that something is wrong, not only with the output tubes but also the driver.
Thanks for the advice
Spoke a little too soon...
OK, so yesterday, big 'POP!' and the meters were reading:-
Right channel (red) should be 38v; was reading 24v.
Left channel (blue) should be 39v; was reading 0v.
I took the bottom cover off, started to run through variac and got up to operating voltage and, with the amp upside down, both channels were reading correctly for over an hour.
I think, (THINK), the sockets are faulty and need replacing. They are very, very loose.
OK, so yesterday, big 'POP!' and the meters were reading:-
Right channel (red) should be 38v; was reading 24v.
Left channel (blue) should be 39v; was reading 0v.
I took the bottom cover off, started to run through variac and got up to operating voltage and, with the amp upside down, both channels were reading correctly for over an hour.
I think, (THINK), the sockets are faulty and need replacing. They are very, very loose.
Last edited:
If you still have erratic Cathode current after replacing sockets then:-
Have you considered that there could be an internal output transformer fault? Such as shorted turn arcing to Ground occasionally, something like that, with perfect silence you might hear noises coming from the output transformer?
A bit drastic but you could physically swap the output transformers and see if the fault stays with the transformer or stays with the channel.
Have you considered that there could be an internal output transformer fault? Such as shorted turn arcing to Ground occasionally, something like that, with perfect silence you might hear noises coming from the output transformer?
A bit drastic but you could physically swap the output transformers and see if the fault stays with the transformer or stays with the channel.
This is drastic (!)
We'll see what happens when I replace the sockets and re-flow all solder connections.
Thanks
We'll see what happens when I replace the sockets and re-flow all solder connections.
Thanks
Yes drastic, what I'm thinking is what you said about the original catastrophic runaway Cathode voltage failure - was that the left blue channel by chance?
It's essential there are no dry joints and good sockets of course.
The input 6N1P valves cannot cause those very low frequency DC bias steps you are seeing because the 0.22uF capacitor blocks DC. I would look very carefully at the 220K and 3K Grid resistors, if the KT66 Grid went open circuit that would turn the valve on affecting the DC bias slightly or greatly.
When the left blue Cathode went to 0V the right red channel was affected suggesting that either a lot of current was being drawn from the HT B+ that the HT B+ line dropped voltage or even went faulty.
The risk is burning out another KT66, especially if you are leaving the amp on for hours.
Another thing you could try is running the amp with only one channel working (pull one of the two KT66 out in turn). That might give you more certainty about which channel is faulty.
Be careful with those high voltages.
I would be putting a DVM on the B+ line and scope probes on the KT66 Cathodes, whatever you are comfortable with.
The DC bias jumps you see are quite likely to be accompanied with smaller random DC bias shifts that your meters don't register.
The photo of the amp - it looks very smart piece of kit!
It's essential there are no dry joints and good sockets of course.
The input 6N1P valves cannot cause those very low frequency DC bias steps you are seeing because the 0.22uF capacitor blocks DC. I would look very carefully at the 220K and 3K Grid resistors, if the KT66 Grid went open circuit that would turn the valve on affecting the DC bias slightly or greatly.
When the left blue Cathode went to 0V the right red channel was affected suggesting that either a lot of current was being drawn from the HT B+ that the HT B+ line dropped voltage or even went faulty.
The risk is burning out another KT66, especially if you are leaving the amp on for hours.
Another thing you could try is running the amp with only one channel working (pull one of the two KT66 out in turn). That might give you more certainty about which channel is faulty.
Be careful with those high voltages.
I would be putting a DVM on the B+ line and scope probes on the KT66 Cathodes, whatever you are comfortable with.
The DC bias jumps you see are quite likely to be accompanied with smaller random DC bias shifts that your meters don't register.
The photo of the amp - it looks very smart piece of kit!
Yes, you are right, it was the blue channel that had the catastrophic runaway voltage.
This was at a time I was using the amp during covid lockdown, 12hrs a day + everyday for 2yrs (!). This, and having the cathode bias resistors mounted on wood not heatsink/chassis. (Live and learn)
Replaced output tubes, only, everything OK for a couple of months then the rectifier flashed and popped. Changed that, then new rectifier flashed and popped again. Problem was the filter cap before the 6n1p plate (33uf). Replaced that and rectifier, (again!) then all was fine.
Until of course, 0v on blue channel as decribed. With another pop, quite a loud one. Odd it was OK when testing for an hour, upside down. Made me think loose sockets and they were very loose. The 3k grid resistor is on the pin of the very loose socket.
Thanks - it is gorgeous looking amp I hope I can get it up and running reliably again.
This was at a time I was using the amp during covid lockdown, 12hrs a day + everyday for 2yrs (!). This, and having the cathode bias resistors mounted on wood not heatsink/chassis. (Live and learn)
Replaced output tubes, only, everything OK for a couple of months then the rectifier flashed and popped. Changed that, then new rectifier flashed and popped again. Problem was the filter cap before the 6n1p plate (33uf). Replaced that and rectifier, (again!) then all was fine.
Until of course, 0v on blue channel as decribed. With another pop, quite a loud one. Odd it was OK when testing for an hour, upside down. Made me think loose sockets and they were very loose. The 3k grid resistor is on the pin of the very loose socket.
Thanks - it is gorgeous looking amp I hope I can get it up and running reliably again.
Hope it is the sockets.
I'm sure you will get it working reliably again.
It's all about working carefully, being patient, thinking about how to test things and never giving up.
In this situation in the past I have made up a pair of high impedance headphones with a croc clip for ground and a probe via a 4K7 resistor and 1uF capacitor in series. Its great for probing around listening for noise etc.
I'm at the bored stage waiting for the all night election results session lol.
Good luck.
I'm sure you will get it working reliably again.
It's all about working carefully, being patient, thinking about how to test things and never giving up.
In this situation in the past I have made up a pair of high impedance headphones with a croc clip for ground and a probe via a 4K7 resistor and 1uF capacitor in series. Its great for probing around listening for noise etc.
I'm at the bored stage waiting for the all night election results session lol.
Good luck.
Hi all,
Well, after rebuilding the amp with new sockets, I tested the voltages of the amp yesterday and all was well.
I then listen to it for an hour and a half.
This morning, switched it on, and after only 10-20 seconds or so, there was a loud pop and the cathode on one channel dropped instantly to 0v followed by a hum noise on that channel.
This is same as before I rebuilt it. Works great first time, the second time it pops and cathode voltage goes to zero on one channel?
Everything seems to be working but obviously not, Any ideas? ,im sure it's not a faulty joint or bad socket and it only affects one channel, also I changed the KT66'S around, so, as it affects the same channel, its not that either.
Well, after rebuilding the amp with new sockets, I tested the voltages of the amp yesterday and all was well.
I then listen to it for an hour and a half.
This morning, switched it on, and after only 10-20 seconds or so, there was a loud pop and the cathode on one channel dropped instantly to 0v followed by a hum noise on that channel.
This is same as before I rebuilt it. Works great first time, the second time it pops and cathode voltage goes to zero on one channel?
Everything seems to be working but obviously not, Any ideas? ,im sure it's not a faulty joint or bad socket and it only affects one channel, also I changed the KT66'S around, so, as it affects the same channel, its not that either.
Last edited:
Zero V on the cathode means No Current through tube.
If the socket is good now, only few causes remain:
If the socket is good now, only few causes remain:
- B+ disconnected from OPT, wire broken or solder joint
- OPT primary winding broken between B+ and UL tap
- Screen disconnected from OPT, wire broken or solder joint
- screen resistor open or solder joint, may have seen overload during runaway event
- there are two in series, 100ohm and 1.2k, take them out and measure them under mechanical stress
Great, thanks, I'll go through those in order.
Could I add to that list the possibility of a dry earth joint at the cathode?
The meter that I have installed at the cathode, when it goes to 0v, flickers.
Also, really oddly, when the amplifier is upside down (testing) the problem hasn't occured yet (not saying it wouldn't, just that, as yet, it hasn't.)
Dry joints can be heat sensitive, sometimes work when warm/hot, then not when cold,
Could I add to that list the possibility of a dry earth joint at the cathode?
The meter that I have installed at the cathode, when it goes to 0v, flickers.
Also, really oddly, when the amplifier is upside down (testing) the problem hasn't occured yet (not saying it wouldn't, just that, as yet, it hasn't.)
Dry joints can be heat sensitive, sometimes work when warm/hot, then not when cold,