Suggestions for eliminating selector switch noise?

[BuzzK]

I've got a modified Eico 2080 with a Broskie (Glass Audio) 3 position stereo input selector switch. It is is not a shorting kind: I get a pop when I change switch positions.

The input stage is a 12AX7, capacitively coupled. The selector switch "feeds" a balance and volume control before going to the input tube (very similar to the Eico HF-87 input).

I remember something about putting 1 MB resistors directly on each switch input to "drain" away any charge. Any other suggestions?

Thank you in advance.

 

[TriodeLuvr]

I remember something about putting 1 MB resistors directly on each switch input to "drain" away any charge.

I've seen that in many preamps, but never understood why it was necessary. No source should have a DC charge on its output, except perhaps momentarily when it's powering on or off.

The input stage is a 12AX7, capacitively coupled.

Can you post a schematic? The input cap might be accumulating and holding a small charge when the tube first conducts.

Jack

 

[BlueVUMeters]

Is the broskie switch a different type than the original Eico unit?

I have the same broskie 3-position switch in my Aikido preamp and haven't had any pops when switching inputs.

That said, I did have an issue with a Goldpoint volume attenuator in the same preamp. I was getting pops around 25% travel on just one of the wafers, and was able to isolate it to a mechanical issue with that specific wafer. I sent the part back to Goldpoint and he replaced the problematic stack on the attenuator. Now the pop at 25% is gone, but there is a new one right around 5%. It's obviously just a manufacturing defect causing the contact to bounce at that specific position.

So it could be possible you just have a defective switch? Quality control isn't really what it used to be, these days.

 

[Lavane]

Would small value decoupling caps help? I just remember a couple SS amps that would build up a very small amount of DC voltage on a filter/speaker selector board. Without the decoupling caps I would get a snap when pushing the filter switches.

 

[BuzzK]

Well, my memory ain't so good. There was no coupling cap. Here is a schematic. Not sure that the 470K grid leak resistor is the best choice here.

 

[BuzzK]

Is the broskie switch a different type than the original Eico unit?

I have the same broskie 3-position switch in my Aikido preamp and haven't had any pops when switching inputs.

That said, I did have an issue with a Goldpoint volume attenuator in the same preamp. I was getting pops around 25% travel on just one of the wafers, and was able to isolate it to a mechanical issue with that specific wafer. I sent the part back to Goldpoint and he replaced the problematic stack on the attenuator. Now the pop at 25% is gone, but there is a new one right around 5%. It's obviously just a manufacturing defect causing the contact to bounce at that specific position.

So it could be possible you just have a defective switch? Quality control isn't really what it used to be, these days.

Yes. The original selector switch wasn't needed and the Broskie unit replaced it. I don't think it's a defective switch.

 

[gadget73]

A cap between the wiper of the level pot and the 470k grid to ground would eliminate any chance of DC leakage issues causing noise, either from the tube or from the source.

 

[BuzzK]

A cap between the wiper of the level pot and the 470k grid to ground would eliminate any chance of DC leakage issues causing noise, either from the tube or from the source.

Thanks for the suggestion. What do you think . . . 0.1 or 0.22 or . . . ?

 

[gadget73]

0.1 into 470k has a -3db of about 4 hz so that should be sufficient I'd think. 0.22 would work fine too.

 

[BuzzK]

Thank you!

 

[BuzzK]

UPDATE: No improvement with the installation of 0.15 µF caps. However, I think I may have found the culprit: the line input grounds are isolated from signal grounds with 10 ohm resistors. There is noticeably less noise when I switch from phono to the line stage -- but not in the other direction. Will take them out and see . . .

 

[peterh]

Loading all unused inputs will help bleeding away any DC. Even better, use a switch that grounds all
unused inputs.
The loading could be 100k as a starter or even less if the sourc(es) can cope with the load..

 

[BuzzK]

In this iteration, the RCA socket grounds were isolated from the chassis -- and from each other. Hooking all of the grounds together solved the pops on all settings except switching into the phono stage. No problem switching out. I'll have a bit more exploration to do.