hum measurement

[dobias]

I read 1-2mv ac at the OPT speaker connections & 1-2dBC (1" from the speaker throat) without any input signal.
Is there a better way to determine hum with just a meter ?
Is zero hum possible with a 2A3 amp?

[trobbins]

Is that with volume and tone set for max output, and input shorted, and highest speaker impedance setting?

If you have feedback loop then you could always disconnect that.

Are you trying to trim the hum?

[dobias]

trobbins,
There is no volume or tone control on this basic stereo amp (Moondog 2A3),
I took new readings this morning while it's quiet:
59dBC ambient, 61dBC, with amp on, taken 1" from the speaker throat.
13mV ac using the 16 ohm tap.
Can a 2A3 amp have zero mV & zero dBC hum ?

After thinking about it, I realize I may have caused the hum problem with my DIY amp. I have a separate PS for the B+ but added four filament transformers close to the tubes in the cramped-for-space amp chassis. I'll be rebuilding the PS to relocate the filament transformers & rectifiers now in the amp.
Anything else I can do?

[DF96]

To get zero hum you need to use a battery DC supply in the middle of a field, well away from any buildings or mains wiring.

To get low hum you need to take care with component and wire placement, and the grounding scheme. This assumes a good design; with a poor design you can't reduce hum. Do a search on here, as this topic comes up quite regularly.

[tomchr]

Originally Posted by DF96

To get zero hum you need to use a battery DC supply in the middle of a field, well away from any buildings or mains wiring.

To get low hum you need to take care with component and wire placement, and the grounding scheme. This assumes a good design; with a poor design you can't reduce hum. Do a search on here, as this topic comes up quite regularly.

And by careful component placement, he means the placement of the various components within the amplifier chassis/circuit. Not the stereo components (CD, amp, etc).

Zero hum is not possible in a 50/60 Hz environment. But you don't need zero. You "just" need inaudible.

I find a spectrum analyzer or wave analyzer to be a better tool for measuring hum as it tells you the frequency as well as the harmonic content. A good external sound card can be a reasonable substitute for those, provided, that the sound card itself doesn't contribute any significant amount of hum itself. Many of the cheap ones do...

~Tom

[dobias]

Gentlemen,
What is considered low hum measured in mV &/or dB?

[kmaier]

Originally Posted by dobias

Gentlemen,
What is considered low hum measured in mV &/or dB?

You also need to have some reference level to measure against. I use 1-watt RMS as the output reference level. As we're talking about low power SE amps, a 45 is good for 2-watt, the 2A3 for 4-watts and a 300B for 8-watts. I would use 1-watt as the reference in all of the above and consider the additional power as headroom.

In any case for my own acceptance, I prefer 80dB signal-to-noise referenced to 1-watt output. For an 8-ohm load, 1-watt is 2.83 VRMS into 8-ohms. The equivalent output noise to achieve 80dB would be 0.000283 VRMS, or 283 microvolts. For a 16-ohm load, it would 4.0 VRMS for 1-watt and 400 microvolts maximum output noise. This is much lower than what you're posting initially, your equivalent s/n assuming 8-ohms is 69dB for 1mv output noise.

Regards, KM

[dobias]

kmaier,
Thanks for the explanation. I finally know what my goal is for a minimum hum.
With a little luck, relocating the filament transformers will do the trick.

[DF96]

'Low hum' would depend on your loudspeaker and room, but -80dB seems a good place to start from. That is -80dB from your normal listening level, which may be below 1W. If you can't hear the hum, it is low enough.

[ChrisA]

Originally Posted by dobias

trobbins,
I'll be rebuilding the PS to relocate the filament transformers & rectifiers now in the amp.
Anything else I can do?

Moving it may or may not help. You are assuming the coupling is via a magnetic feild from the transformers. It could be more likely the wires inside the chassis. Longer wires make better antenna so think about the overall design. The worst thing is to have wires parallel to each other.

Is there ripple in the B+ power supply?

How is the system grounded. Using a star? is there any way at all current can be flowing in the ground. No ground loops. Are you using shield wire in the signal path?