Sources of hum.

[pre65]

It's rare for me to build a completely hum free amp, so I'm looking into ways of identifying and curing hum.

Firstly, am I right in thinking that shorting the amp input phono will remove any prior equipment (ie pre amps) from the equation ?

Second, is a DMM set on AC the best way to start hum chasing ?

If so, what is an "acceptable" limit of ac component ?

[Cressy Snr]

It's rare for me to build a completely hum free amp, so I'm looking into ways of identifying and curing hum.

Firstly, am I right in thinking that shorting the amp input phono will remove any prior equipment (ie pre amps) from the equation ?

Second, is a DMM set on AC the best way to start hum chasing ?

If so, what is an "acceptable" limit of ac component ?

Shorting the input does what you say and confines your search to the
piece of equipment in question.
As to acceptable levels of hum in a preamp, The only acceptable level is none.
If you think about it only one millivolt of AC in a preamp could be magnified by as much as 200 by the following circuitry so that's loud hum at the speakers.
The best tool for chasing hum is a scope. I have found that your average non-professional AC millivoltmeters can pick up allsorts of crap from the environment, becoming useless as it swamps the stuff you are trying to pick up.

The scope can show you exactly where in the circuit, the hum is being introduced and you can see it. It is usually a pretty dirty sine wave. I usually probe from output back through the circuit, but some do it the other way. Makes no odds.

Maybe it's time to get that scope out.

[Nick]

Don't just short the inputs out, make up some plugs with 10k resistors in, some circuits will get excited when the input is a dead short to ground. It also prevents you from using a clip lead, which will short to ground, but also act as a loop antenna at the same time.

I would aim for 20dB below what you can hear. You may not manage it, but its something to aim at.

For a preamp, phono or dac I would aim at no sign of hum or noise on a scope set to 2mv division, for a power amp, I like to aim for less than 400uv P/P at the speaker terminals.

But work it out, you know your gain (or should), you know the signal level, aim for at least 90dB S/N, and not just at peak signal, but average, which will be 10dB or so below peak.

The meter is no use, you need to learn how to use the scope if you want to track down hum at its source.

[Paul Barker]

Also has to be mentioned just how much noise comes through the power supply. When you use a power conditioner a lot of noise goes including hum. Whether you made it or bought it.

[Mike H]

It also prevents you from using a clip lead, which will short to ground, but also act as a loop antenna at the same time.

Done that on a phono pre! What a staggering amount of crap was picked up

[pre65]

Initial investigation lead me to the AK4393 DAC as the worst culprit at the moment.

So, I've installed my other DAC for a while and got the offending item on the bench.

Oscilloscope where art thou ?

EDIT. I've read somewhere (quite recently) about using a 1M resistor between live and ground on phono plugs to give noise an escape route. Or have I misunderstood something ?

[Nick]

EDIT. I've read somewhere (quite recently) about using a 1M resistor between live and ground on phono plugs to give noise an escape route. Or have I misunderstood something ?

Not sure as I dont know what you understand, but if you are making a shorting plug I would use a lot less than 1M

[pre65]

No, not for shorting plugs. It was in the ACF instructions.

Quote " Here's a cheap trick to try: at each RCA jack , place a 100K - 1M resistor, bridging input hot and jack ground. Why ? The resistor provides a path for the AC signal present at the jack, so given a choice between radiating into the chassis or going through the relatively low impedance resistor the AC signal chooses the latter path, reducing crosstalk."

[Nick]

Ok, that (I think) means that unused inputs still see 100k, or 1M. Not sure why thats a issue for noise, as it will have a path to ground at the other end of the interconnect.

[Paul Barker]

Ok, that (I think) means that unused inputs still see 100k, or 1M. Not sure why thats a issue for noise, as it will have a path to ground at the other end of the interconnect.

Could it be RFI collected along the length of the interconnect if it's low quality design?

[pre65]

My AK4393 DAC has hum.

I've got it on the bench and done a few basic measurements with the DMM.

Left output - 2.9mv DC 14mv AC

Right output - minus 7.86mv DC 14mv AC

Now, I presume the input cap on the pre will block the DC but what can I do about the 14mv AC ?

[ed]

put the scope on it and find the frequency?

may give a clue as to whats causing it

[pre65]

The scope has been energised.

Still can't find my little book that tells me how to use it though.



EDIT.

I have had to cease activity cos I don't know what I'm doing.

[pre65]

I've taken a piccy of the scope trace.

One channel output.

20mv - 10us - MTB (auto)

Strangely the trace seems to be the same when I press the AC/DC button.

[ed]

oooh...

too much hash man!

just trying to follow a wave looks like 15us which I make 66.6 khz.

It's the mark of the beast, you're doomed!

but seriously, I don't know, somebody will though, have faith, or call a priest