• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Locating hum source

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hey there,

my amp test setup is finished und luckily the first power-up went without smoke and yielded some nice music output :)

See http://www.diyaudio.com/forums/tubes-valves/180273-rundmaus-work-pp1c-without-sand.html
for details...

Now, some slight buzz/hum is audible close to the speaker. It does not sound like a clean 50Hz/100Hz sine wave, instead somewhat 'sharper', sawtooth-like.
Also, it did not disappear instantly when I switched off the mains.

Conclusions so far: not heater-related, not due to B+ ripple

I am planning to follow these steps to locate the problem:

* check amplitude and waveform of the hum via oscilloscope

* ground the grids of the input/PI stage, if the hum is induced in the input wiring/volume pot circuitry, it should disappear

* if not, ground the grids of the finals, if the hum disappears, it's coming from the PI stage

* if it still hums, ask someone more experienced ;)

Did I forget anything important? Any remarks concerning possible sources of not mains-related hum?

Thanks in advance,
Andreas
 
The best way to determine where the hum is coming from is to o'scope and see what the waveform and frequency is. If double the line frequency and sawtooth like, it's AC ripple getting into the signal chain, usually because of ground loop problems. Could also be magnetic coupling from an AC ripple choke.

If the same as the line frequency, then it's likely magnetic coupling between a PTX and the OPT.
 
The best way to determine where the hum is coming from is to o'scope and see what the waveform and frequency is. If double the line frequency and sawtooth like, it's AC ripple getting into the signal chain, usually because of ground loop problems. Could also be magnetic coupling from an AC ripple choke.

If the same as the line frequency, then it's likely magnetic coupling between a PTX and the OPT.

He has already switched off the AC mains power and listened as the amp faded out. This is a VERY good test because an amp still running on no AC power can not have B+ ripple, Heater noise or magnetic coupling between transformers.

With the mains switch off the only source of hum is the AC mains wiring inside the chassis that connects the input IEC connector to the switch. Make sure that wwire is tightly twisted and tuck in the corner of the chassis and that there are no audio signal or power supply wires even close to the mains wiring. You can test this by unplugging the amp and listening vs, switching off the amp and listening.

The only thing left is that AC hum is getting in via the "environment". The best solution to that is "correct" value grid stoper resisters that are soldred directly to the tube socket and also using shielded wires on the signal wires inside the amp. Shielding really does help (but remember to ground only one end of the shield.)
 
Hey there,

Also, make sure that there is no space between signal and signal ground conductors, anywhere. Otherwise they are an antenna for time-varying magnetic fields. Tightly twist them together. (Do the same for all conductor pairs.)

I tried to do that wherever possible, but not in all cases there's a signal ground wire there to be paired with the signal carrying one.

A picture of the build and a schematic would be very helpful.

The 'build' is currently a not-very-tidy test setup, see the following post:

http://www.diyaudio.com/forums/tubes-valves/180273-rundmaus-work-pp1c-without-sand-3.html#post3192994

Schematic can be found in the same thread, although it has been combined with a different power supply:

http://www.diyaudio.com/forums/tubes-valves/180273-rundmaus-work-pp1c-without-sand.html#post2418761

Regards,
Andreas
 
Hey there,



I tried to do that wherever possible, but not in all cases there's a signal ground wire there to be paired with the signal carrying one.



The 'build' is currently a not-very-tidy test setup, see the following post:

http://www.diyaudio.com/forums/tubes-valves/180273-rundmaus-work-pp1c-without-sand-3.html#post3192994

Schematic can be found in the same thread, although it has been combined with a different power supply:

http://www.diyaudio.com/forums/tubes-valves/180273-rundmaus-work-pp1c-without-sand.html#post2418761

Regards,
Andreas

Andreas,

You could try running the two signal grounds through wires, instead of through the chassis, back to the ground end of the last capacitor of the main power supply.

Those are the chassis gnd symbols on the schematic that are between the two 100K inout resistors and between the two 330k resistors for the second stage.

Can you post a schematic of the power supply?

If you have more than one smoothing cap for the main supply, then the first smoothing cap's ground should probably connect at or very near the transformer or rectifier ground, and the other caps' grounds should connect at a different spot, which is probably where the signal grounds from above should also connect. The idea is to not have the rectifier charging pulses sharing the same conductor path as anything else, since they will induce voltages across the conductor, which would then be arithmetically summed, to some degree, with the voltages back at the non-ground ends of whatever shares the conductive path. It is part of the concept of "star grounding", which it looks like is not implemented in your build. But doing it for only the signal grounds might help.

You also have some "enclosed loop areas". Just looking at the schematic: If you could put the 100k input Rs and the 100 Ohm Rs in a straight line, and then pull the 330Ks and 0,33 uFs from the next stage over and put them and their wiring right against the line of input resistors, and then put the two 4k7 Rs right next to each other (centered vertically on that part of the schematic) and run their four wires twisted together, straight to the left all the way to where the line of input resistors was, you could probably close up the loop area and not have antennas for the time-varying magnetic fields to induce currents in. There are probably several ways to accomplish the same thing. So there is probably a better way than I described. If you could also get the ground wires to follow their signal wires for as long as possible, it would be even better. Ideally, you might have wanted to fold the schematic in half, along the horizontal center line, and then implement it that way. But if you don't want to move tubes around at this point you should be able to contort the wiring to accomplish almost the same effect.

If we're lucky maybe DF96 or one of the other experts will chime in.

Read this thread: http://www.diyaudio.com/forums/tubes-valves/203526-thoughts-apparent-noise-ground-due-power-tx.html

Cheers,

Tom
 
Last edited:
Andreas,

You could try running the two signal grounds through wires, instead of through the chassis, back to the ground end of the last capacitor of the main power supply.

Hey there,

I took another look at the breadboard after a good night of sleep and found that in the line of 'solder iron enthusiasm' I slightly mixed up my star ground scheme, leading to the following:

Grounding scheme (current setup)

2007_001.jpg


Some connections do not feel right, I marked them with '!'. Any more comments on how to improve the grounding?

Andreas :)
 
if the hum goes away with a cheater plug on the amp or preamp....this virtual cheater plug will work without posing a shock risk. it acts like a cheater plug for normal ground loop voltage levels but once the voltage goes above the diodes forward volts...the circut transistions into a ground preventing the shock hazard.
 

Attachments

  • ground loop block.png
    ground loop block.png
    10.6 KB · Views: 148
if the hum goes away with a cheater plug on the amp or preamp....this virtual cheater plug will work without posing a shock risk. it acts like a cheater plug for normal ground loop voltage levels but once the voltage goes above the diodes forward volts...the circut transistions into a ground preventing the shock hazard.
 
Hey there,

I took another look at the breadboard after a good night of sleep and found that in the line of 'solder iron enthusiasm' I slightly mixed up my star ground scheme, leading to the following:

Grounding scheme (current setup)

<snipped image>

Some connections do not feel right, I marked them with '!'. Any more comments on how to improve the grounding?

Andreas :)

Yeah, I would definitely make _separate_ ground-return conductors all the way to the star ground point for all of the small-signal stuff.

Just as inportant as star grounding is to not make antennas by having enclosed loop area between conductor pairs. "Current flows in loops. Eliminate or minimize the area they encircle."
 
Yeah, I would definitely make _separate_ ground-return conductors all the way to the star ground point for all of the small-signal stuff.

Yes, I think I will try this. In the current grounding scheme, the ground connection from the EL34 cathode resistors with its high current (approx. 90mA) goes to the star with all the small signals and grids. I'll move this one to the middle star and report back!

Andreas
 
Just as inportant as star grounding is to not make antennas by having enclosed loop area between conductor pairs. "Current flows in loops. Eliminate or minimize the area they encircle."

One way to do that, that works well on breadboards is the "star of stars" groiunding system. You can make local star ground for each sub circuit and then conect these to the central star.

The idea is to place the AC curent source (A power suply filter cap) physically near the load (such as a driver tube) and also place the local star ground physically close to both. This minimises the antenna effect described above. Then you conect the local stars to the central. Typically you have as many local stars as you have power suply filter caps.

the diagram on this page shows how to do this, it really does minimize the loop antenna effect The Valve Wizard
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.