I built Tubelab SE mono as discussed here. It's been fine until recently. I hear buzz from the speaker. It remains even when IC cable is disconnected (only amp and speaker connected). I went through the "check out" process and all the voltages are normal. I connected the speaker cable to other speakers and the buzz followed. What could possibly be causing this?
Then it's not picking it up through the input wiring and it is less likely to be a ground issue (but still could be). If it's a 120Hz buzz and not a 60 hum, then it is probably a power supply related problem. Do you have a scope with high voltage probes? If so, take a look at the B+ and B-.
Is the buzz from one channel or both? If it is a single channel, try swapping the 5842's. I have seen oscillation cause a buzz in one channel.
No.
It's more of 120Hz buzz.
It's a monoblock amp. I swapped 5842 with another TSE but it does not follow the tube. What puzzles me is, it buzzes one afternoon, then I use it again later that evening and it won't buzz. I'm wondering if it has to do with the power from the outlet. I use TrippLite surge protector.
I have a dimmer in other wall outlet but it was off when the buzzing came back. My other TSE (stereo) doesn't have any of this problem. I'm wondering if it's the power supply circuit. Could any caps or resistor going bad cause such noise? It was working fine for 2 years. 😕
It is posible that a capacitor could be developing a high ESR condition.
Watch your line voltage to see if the buzz is prevalent under low voltage conditions. Also are there any new additions in your house that tend to distort the line voltage? Flat panel TV's and computers without PFC are big offenders.
The filament regulator feeding your output tubes may be going into dropout during low line conditions. Dirty power makes things worse. Several years ago I measured the distortion on my line voltage at 4%. Now it varies from 8 to 14% with flat topping very obvious in the evenings. This makes the B+ voltage vary even more than the line voltage does.
If this is the case adding more capacitance may help.
I haven't added anything electronic for a year. AC from the power socket is at 123 to 125 VAC when this amp buzzes.
Higher C4 value for higher B+?
It may just be a matter of heat cycling as well on the caps. C6 and C7 are near R6, which can get rather toasty. Are you using a choke in place of R4? To check if the filament regulator is dropping out, measure the voltage at the rear-facing side of R3.
With all the tubes in place, B+ = 380, B- = -280. It's a monoblock but is B- unusually high? If so, what harm could that do?
More observation, it seems consistent that it's very quiet when I turn it on but as it warms up, the buzz returns. It takes about 10 minutes or so and it slowly gets louder up to a point and stays at that level. Does it sound like any of the semiconductors?
That sounds right, but I don't have a PCB in front of me. Measure both sides relative to ground and report the higher of the two.
Put a meter on the 10 ohm resistor in the plate lead of the output tube. Maybe the current is creeping up due to a gassy tube.
I have a bias meter installed from those leads which I can monitor real time. It's stable at where I set (70mA).
R3 lead near R24 = 7.87VDC
R3 lead near C5 = 6.67VDC
One thing I haven't mentioned is, if it makes a difference, input tube (5842) plate voltage won't go above 169VDC when checked from coupling cap lead facing input tube to ground. I turned the trim pot (R20) all the way up but that's as high as it gets. I changed R3 to lower value but still same.
Last edited:
The culprit has been found. Embarrassing to say, it was me. 
When I assembled the board 2 years ago, somehow wrong value resistor was installed at R25. It was supposed to be 20K Ohm but what got in was 200 Ohm. On Tubelab website, it is now updated as 30K or 35K Ohm depending on B+ voltage. I happened to have a spare 27K Ohm resistor so I swapped it out and the buzz is almost completely gone. Also, 5842 plate voltage now reaches 174 VDC after about 15 minutes.
I'm wondering, could this have damaged other parts after about 400 hours of use with undervalued resistor in R25 (or R14)?
When I assembled the board 2 years ago, somehow wrong value resistor was installed at R25. It was supposed to be 20K Ohm but what got in was 200 Ohm. On Tubelab website, it is now updated as 30K or 35K Ohm depending on B+ voltage. I happened to have a spare 27K Ohm resistor so I swapped it out and the buzz is almost completely gone. Also, 5842 plate voltage now reaches 174 VDC after about 15 minutes.
I'm wondering, could this have damaged other parts after about 400 hours of use with undervalued resistor in R25 (or R14)?
You were probably giving the FET a real workout to get the grid bias on the output tube up, but if it has been working all this time I would not worry about it.
- Status
- Not open for further replies.