Locating hum source

[Rundmaus]

Hey there,

I made some changes to the grounding scheme yesterday, the hum level is now significantly lower.

The remaining hum disappears instantly on switching off the mains - so it's heater hum or PSU ripple. In principle, both should be distinguishable by the frequency, 100Hz for PSU ripple, 50Hz for heaters, am I right?

Anyway, I am able to switch the mains for the HV supply and the heaters independently, so I can just try and find out which one makes the hum disappear instantly. Just forgot about that simple test yesterday, maybe it was a little too late in the evening

Andreas

[trobbins]

There are a number of hum ingress mechanisms - some have f fundamental, some have 2f fundamental. Sometimes its the harmonics of 2f that are most noticeable.

[Rundmaus]

Hey there,

I found time to get the scope out today, and found the following:

Output hum into 8ohms:


So there's roughly 1mV RMS hum into 8ohms on the output, with a frequency of 100Hz. Note the strange waveform: The hum is not sinusoidal, but looks like a full-wave rectified AC instead. There are short 'spikes' at the lowest parts of the waveform, too...

Anyone seen such a waveform on the output of an amplifier before? Ideas about the origin of the hum?

Checked the B+ rail for excessive ripple, but the 12mV RMS ripple I measured there seem reasonable:



Sorry for the low image quality, multiple traces due to slow exposure time and low-frequency rail variations...

Greetings,
Andreas

[trobbins]

Perhaps try a battery supply for input heaters - that would be my first check.

[Rundmaus]

Originally Posted by trobbins

Perhaps try a battery supply for input heaters - that would be my first check.

Hi,

forgot to mention the following:

a) Hum is instantly away when switching off the mains for the HV transformer

b) Hum waveform remains unchanged when switching off the heater mains for a short moment.

Andreas

[trobbins]

Hmmm, for a common PT I'd try a humdinger pot on the heater - the mechanism of HT rectifier waveform coupling to heater winding is still valid for (a) and (b).

But I think you have two separate transformers?

[Rundmaus]

Originally Posted by trobbins

Hmmm, for a common PT I'd try a humdinger pot on the heater - the mechanism of HT rectifier waveform coupling to heater winding is still valid for (a) and (b).

But I think you have two separate transformers?

Hey there,

in fact, the test setup uses three different transformers:

1. 480-0-480, 180-0-180, 5 (+HV, -HV, rect. heater)

2. 6.3, 6.3 (-HV rect. heater, ECC88)

3. 6.3, 6.3 (EL34)

The waveform of the hum very much looks like the current/voltage waveform I would expect between rectifier and 1st filter. I did not use twisted leads when setting up the power supply, maybe the rect.-filter connection couples into some high-impedance stuff in the amp circuit...

Andreas

[trobbins]

Were the scope shots with both inputs shorted to gnd? Have you checked some other simple circuitry 'exclusions' - eg. grounding the output stage signal inputs; replacing ccs with standard bypassed resistor. Then maybe DC heater, or humdinger pot.

[Rundmaus]

Hey there,

Originally Posted by trobbins

Were the scope shots with both inputs shorted to gnd?

They were made with the positive input shorted at the volume pot. The negative input is shorted to ground by default, as I do not have symmetric signal sources.

Have you checked some other simple circuitry 'exclusions' - eg. grounding the output stage signal inputs; replacing ccs with standard bypassed resistor

One test I performed was taking out the interlock relay for the negative rail, disabling the CCS. So no amplification in the input/PI stage. Hum was unchanged.

Can I simply ground the grids of the output stage for testing?

Greetings,
Andreas

[disco]

ECC88 cathode - heater isolation shot?
Try exchanging tubes.