The land of the unusual fault

[rabski]

Today, I have mainly been.... trying to eliminate a faint but annoying hum on the new preamp output. All the usual suspects eliminated. Power supply is as clean as is realistic. HT separate per channel, valve rectified and valve regulated. Thoroughly smoothed (down to about 1uV ripple) before the regulators, then more caps after. Filaments are regulated DC and also separate per channel. A 2-box affair, with all the PSU separate. Incoming mains earth goes to casework, then a standard 'lift' circuit for the chassis ground.

The business end is connected by a pair of umbilicals and these use shielded wire. The preamp bit has an internal copper chassis, rubber mounted so again isolated from the case. So far so good. Except, a low-level hum. Everything 'volts wise' eliminated, so looks very much like a ground loop. Except it can't be can it? It's 'lifted' and the hum is there with the pre connected to the power amp, both warmed up and nothing else connected.

And... I used a pot extension shaft to mount the input selector at the rear of the chassis. The rotary switch is mounted on a bracket that is, of course, attached to the copper chassis. And the extension rod passes through a brass bush on the front plate, thereby making a perfect electrical connection between the case and the subchassis and causing a ground loop.

Took me the entire day to find that, as it simply made no sense at all. Last job today is therefore to ream out the bush a little and slide a bit of heatshrink over the extension shaft so it doesn't make a contact. At the moment, it's running 'lid off' without the extension shaft connected and I'm down to the 'ear against the speaker before you can hear anything at all' level of noise, which is fine.

Every day brings a new surprise

 

[andrew s]

Reminds me of the early days of toroidal transformer fixing to the chassis "problem" which resulted in a single turn short. Less catastrophic in your case.
Regards Andrew

 

[Tony_J]

Reminds me of the early days of toroidal transformer fixing to the chassis "problem" which resulted in a single turn short.

Ouch! That must've been interesting...

 

[MF 1000]

‘Bushes’ always cause problems

 

[rabski]

‘Bushes’ always cause problems

I knew putting 'reaming' and 'bush' in the same post was asking for trouble

 

[andrew s]

On reflection, nice meal and half a bottle of wine, adding the shrink wrap may have fixed the dc loop but left a capacitance for rf ingress so best to add some ferrite cores on the spindle to avoid that noise floor modulation.
Regards Andrew

 

[Tony_J]

I knew putting 'reaming' and 'bush' in the same post was asking for trouble

Ah yes, that old proverb - a reamer in the hand...no, that's not right

 

[rabski]

On reflection, nice meal and half a bottle of wine, adding the shrink wrap may have fixed the dc loop but left a capacitance for rf ingress so best to add some ferrite cores on the spindle to avoid that noise floor modulation.
Regards Andrew

Trust me, a bit of RF is the least of my problems

 

[rabski]

On the double entedre line, there won't be any RF ingress via my shaft, as there's a metal knob at the end of it.

 

[audio_PHIL_e]

In a recording studio many years ago, one of the chaps wired up i/p jack sockets at one end of the ally strip above the double glazed window between the control room and the soundproofed room and o/p jack sockets from the monitor and headphone amplifiers at the other end. They turned it all on and couldn't work out where this massive hum was coming from. I took one look at the sockets on this ally strip and remarked that the metal casings of the i/p sockets were forming a ground loop to the metal casings of the o/p sockets, suggesting replacing the o/p sockets with plastic ones. One trip to Maplins later, the o/p sockets were replaced with black plastic jobs and the hum went away.

 

[andrew s]

A trip to Maplins. Those were the days.
Regards Andrew

 

[rabski]

A trip to Maplins. Those were the days.
Regards Andrew

Miss it badly. We used to have one about five minutes drive from here. An absolute godsend when you realise you need the odd bits and pieces.

Used to be able to order up to 8pm from RS for next morning delivery as well, but their deliveries don't seem as consistent as they used to be either.

 

[audio_PHIL_e]

Miss it badly. We used to have one about five minutes drive from here. An absolute godsend when you realise you need the odd bits and pieces.

Indeed. many a thing has been fixed or at least got working for now with bits from maplin

 

[audio_PHIL_e]

RS

does anyone else still refer to them as RadioSpares?

 

[Tony_J]

does anyone else still refer to them as RadioSpares?

Definitely

 

[audio_PHIL_e]

Definitely

Maybe we are of similar age

 

[MF 1000]

Bring back Tandy’s …aka Radio Crap …oops Shack

 

[rabski]

To me, they've always been Radiospares. I just type it as RS, because of the website.

I actually have a few oddments in the spares pile that are still in 'Radiospares' packaging

 

[rabski]

Reminds me of the early days of toroidal transformer fixing to the chassis "problem" which resulted in a single turn short. Less catastrophic in your case.
Regards Andrew

A number of us may have done the old 'top of the casework touching the top of the torroidal mounting bolt' trick. That has some 'interesting' end results...

 

[audio_PHIL_e]

Bring back Tandy’s …aka Radio Crap …oops Shack

back in the day you could get real components from them - before they filled the shop up with cheap disco lights and radio-controlled cars.