Hum in tube amp

[vcelkamaja]

I have build small SE tube amp using 6L6 and I can hear undesired audible hum from it. Few details:
- it is PCB based (commercial so I belive PCB is not a source of hum): 6L6 SE Tube Amplifier Kit 10+10W (Stereo)_Power Amplifier Kit_Valve Amplifier Kit_Analog Metric - DIY Audio Kit
- toroidal power teansformer is inside the chassis (as well as PCB)
- output transformers are outside the chassis
- power supplt has no choke - just capacitors and 10R resistor
- AC heating is center grounded - ok
- I use start grounding and both input and output jacks are isolated from chassis
- I use 3 wire power cord (PE connected to chassis)

What I have observed:
- after power on hum rises and after few seconds it is audible (heating / warming up?)
- hum disapears immediately after power off
- hum is still audible even if I shorten input pin to gnd

Based on these observations what is the source of the hum?

I don't think there is an issue with ground loop or circuit (commercial PCB kit).
Two ideas coming to my mind:

1. better power supply filtering (add a choke?)
2. "better" wiring - twist pairs of wires tightly, placement of wires - signal / output transformer / heating / power

Any ideas or recommendations where the hum comes from?

 

[chip647]

Hum like you describe is almost always the result of a ground loop. It is one of the most difficult things to track down, but it is fixable. You don't need any new parts. Here is a list of possible causes:

1. A "cold" solder joint. Very easy problem to create with lead free solder. Check all of your joints for any looseness. Re-flow the ones that look suspect. I have had joints that looked perfect make lousy connections.

2. A ground that runs to two different places. Each item should have only one path to ground. Think Star ground.

3. Unplug your interconnects and short the amp's inputs to make sure the problem is contained within the amp.

4. Make sure everything that should be grounded is grounded. Use your meter to check every grounded item.

You will find it. When I started out doing this it would sometimes take weeks to find the problem. (Not so much anymore).

Good Luck!

 

[Frank Berry]

I'm guessing that you should add a choke to the power supply. According to Analog Metric, it's an option.

 

[DF96]

Is the hum 50Hz or 100Hz? Pure hum or buzz?

The circuit shows a 100R resistor in the PSU. You say 10R?

It only claims 85dB S/N, which is not particularly high.

The power rail decoupling looks OK, provided they have got the grounding right.

Some hum may be expected from the heater supply, as the PCB can't do twisted wiring.

 

[Frank Berry]

"It only claims 85dB S/N, which is not particularly high."

That S/N is at 1kHz. They don't give a figure <1kHz.

 

[DF96]

S/N sometimes includes hum in the noise figure. It shouldn't, but sometimes it does.

 

[roline]

I tried to not use an inductor in the power supply only to have to add it to get the hum out... C-L-C(B+)-R-C (2nd stage)-R-C(first stage) Also be sure your heater wires are tight twisted and 1/2" off the PCB or away from the active components. Use a hum pot or verify 6.3v CT is connected to ground, checked. Watch your input to the first stage and verify no ground loops, shielded cable or tight wound.... Verify not a ground loop on the power ground and incoming signal ground, another source....

 

[jrenkin]

I think it is important to know if it is 50hz or 100hz (60/120 in USA). 50hz= ground loop, 100hz = power supply. If it is the ps, us a choke, increase caps or add an RC stage if you have the headroom. Otherwise, find that ground loop, and twist!

 

[fotios]

I bet that is a power supply ripple noise.
What is the difference between the terms: "HUM" and "BUZZ"?

 

[DF96]

Hum is a sine wave. Buzz is a sawtooth, with lots of harmonics.

 

[vcelkamaja]

Thanks for all your comments.

I think it is 100 Hz hum (didn't measure it, just compared by ear with tone generator) and it didn't appear immediately after power one but a few seconds later (warming up?). But it disappears immediately after powef off (no fading out)


I believe I have all grounds connected right (no grond loops).
The resistor in power supply filter is 100R, not 10R as I mentioned before.
So do you think replacing this resistor with choke will help?

Thanks

 

[TubeHead Johnny]

I had a similar problem a few years back involving a SE guitar amp. The OPT was picking up hum electromagnetically through the Power Transformer. I had to move the OPT to the opposite side of the chassis to fix it. SE designs are much more susceptible to this than PP. An easy way to find out is to just disconnect the P/S secondary and run the transformer w/o a load. If there is electromagnetic coupling you will still hear hum through the speakers. I think in your case the best thing to do would to install the amp and P/S on two different chassis.

 

[vcelkamaja]

Still haven't found a solution...

Well, after a couple of months I'm trying agan to sovle this issues with hum. A few observations:

- the hum is 100 Hz
- it doesn't appear immediately after power one but rather a few seocnds lates - maybe when heating of tubes gets to the right temperature?
- it disappears immediately after power off
- I have tried choke instead of power resistos but there is no difference
- I have tried hum pot parallel to CT resistors (connected to GND) but it doesn't help (it influences 50 Hz hum but not 100 Hz that I'm getting)
- I have twisted all pairs of wires tightly
- all grounds are star grounded - no loops
- I have tried shielding of power transformer but again no difference

Any other suggestions?

 

[jrenkin]

I had this on an amp that was not supposed to have this, but did due to my choice of rectifier tube and the required low value 1st capacitor. I solved it by adding an additional filter stage. Worked great. Now CLCRC. with a low value R, you lose very little B+.

 

[TubeHead Johnny]

Well if there's a delay, it's not electromagnetic coupling. You can rule that out. Do you have a schematic for the P/S? That would help us a great deal!

 

[vcelkamaja]

So you think that CRC filtering - 100u-100R-200u is not enough?

I have tried CLC filtering as well - 100u-10H-200u but there was no difference.

 

[TubeHead Johnny]

How bout a photo?

 

[vcelkamaja]

I'm trying to attach some pictures. The circuit is realized on PCB - see the first page with link to supplier's web page.
All pictures show top plate of amp with output transformers open but there is not difference in 100 Hz hum if the top plate is open or closed.

 

[N4BBQ]

I do not like that power supply design. It reeks of 100Hz/120Hz ripple. Add the choke, re-measure the ripple, then go from there.
---
Schematic is found by digging here.

6L6 SE Tube Amplifier Kit 10+10W (Stereo) - Analog Metric - DIY Audio Kit

 

[vcelkamaja]

Well, I tried to move from CRC to CLC by replacing 100R resistor with choke (I think 10 H / 250 mA) but there was no audible improvement.