Altec 1567A buzz.....ideas?

[emrr]

Rebuilt 4 of these mixers as a group, they had the buzz before, the buzz is unchanged after filter and coupling caps. The buzz meets Altec spec (which isn't worth much as described), but there are other examples of this mixer that do not buzz at all. You couldn't use these with a microphone unless it was a steady state loud source.

Disconnect B+ from V1/2 and remove the tubes, the buzz is still there with all mix pots off, master at normal level. Master off, buzz gone.

It appears the buzz can be isolated to the mix resistor / mix pot area. Disconnect all 5 mix resistors, the unit is totally quiet. Reconnect any one mix resistor to V3A input cap, buzz is back. Tried changing one mix resistor to metal film, just in case, no change. Leave one mix resistor connected to the cap, wire to pot disconnected, buzz is the same.

I experimented with the ground planes, test leads between different ground points to make sure nothing changed, since the case ground connections are all rivets. They've been tried in multiple environments too, and a lot of time verifying environmental noise, quietest case ground connections for listening, etc.

http://www.triodeel.com/al1567a.gif

 

[Winston OBoogie]

It's 4:20 am here so, off to bed now. But I'll give this a good look tomorrow for what it's worth, whatever I can do to help ya be done and rid of these things

Edit: Are these using the chassis as ground return for decoupling/filter caps etc? Or have you redone the scheme?

 

[Winston OBoogie]

This is a 1567a yeah?

 

[emrr]

Are these using the chassis as ground return for decoupling/filter caps etc? Or have you redone the scheme?

I have not redone it. That doesn't sound fun. There were definitely not meant to be worked on, they were meant to be trashed. Chassis is ground return, does appear well laid out, considering.

 

[rock soderstrom]

Modify V3 from grid leak to cathode bias. Grid Leak is very sensitive to all kinds of interference.

 

[emrr]

Modify V3 from grid leak to cathode bias. Grid Leak is very sensitive to all kinds of interference.

Thanks, I tried that first actually, no change.

 

[scott2000]

Is that a connection at the HT secondary 130vac? Never seen that. Is that some kind of multiplier and negative grid bias.? Just trying to understand.

 

[emrr]

Is that a connection at the HT secondary 130vac? Never seen that. Is that some kind of multiplier and negative grid bias.?

Voltage doubler, also the 436/438.

 

[mjrippe]

Hey Doug, did you see this note about the cable in the 1567a manual? I recall having some 1567 noisier than others but I really don't remember if I ever found a solution. -Mike

 

[emrr]

Yep, saw that, none of that makes any difference.

This really appears the mix resistors are inducing buzz from somewhere, and they've been set up and tested in a couple different physical locations.

 

[Winston OBoogie]

I've been looking for clean internal shots on and off to try and understand the layout.
In the shot attached, is that cap with the "speech bubble" the main grounding point for the audio?

Looks like the earthy side of the first 4 channel's level pots go there, what else has its earthy/ground attached to that cap case?

 

[emrr]

I've been looking for clean internal shots on and off to try and understand the layout.
In the shot attached, is that cap with the "speech bubble" the main grounding point for the audio?

Looks like the earthy side of the first 4 channel's level pots go there, what else has its earthy/ground attached to that cap case?

Input stages and pots ground there. Then there's also a potential loop through the shielded wire to V3A input (I tried lifting it, gets worse), it attaches to both mixer 3 pot ground side AND the record out / input 5 ground terminal....the aforementioned input stage ground to the cap can is wired to that same input 5 ground, and on to the next filter cap case closer to the PT, then the next. The front panel has a ground wire that runs between riveted tabs on respective chassis.

Imagine again, C6 floating on the input side - then the buzz is gone. Picture the following EQ stage is all built on that front panel too, but is unaffected. Disconnect B+ from V1/2, take those tubes out, disconnect input 1-4 mix resistors from C6, it still buzzes with only mix 5 mix resistor connected. Then no no buzz once 5 is disconnected - it's a passive input. Maybe I've explained in an alternate way to help describe it.

V1/2 hooked back up - If I have a single mix pot connected, and its ground moved to the input 5 ground terminal, it kills the buzz when the pot is off, but it makes it worse when the pot is turned up.

 

[Winston OBoogie]

Thanks for the rundown, appreciate it dude.

I'll try and look at it from what you've written.

 

[emrr]

Moving V1/2/3 to an external DC filament supply drops the buzz some, but not a lot.

Back to original filament connection, every 60Hz harmonic is there evenly through 420Hz, then starts dropping until it disappears in the noise floor over 2K.

 

[Winston OBoogie]

I think the 1567a's internal heater DC supply is probably fine but, doesn't hurt to have checked so

Just to clarify: If input 5 (and any audio ground to chassis connection there if there were one) were completely disconnected, or if input 5 came via a 1:1 10K:10K isolation input transformer, would the buzz go away?

Also: trying to see what's going on in this picture which has an aluminium shield next to input 5's tag strip...

Do your units have this? Others I've seen don't.


Edit: Sorry, one more question - The return (negative) of the current loop at V3a's cathode and the bottom of the R22 3M3 grid resistor, where does that go to ground? Same capacitor uninsulated mounting plate that the 1 - 4 pots earthy side's go to?

 

[emrr]

I have not tried the iso transformer, as there doesn't have to be a connection from the mix resistor to the mix pot at all, there simply has to be a mix resistor connected to C6 to have buzz, nothing connected on the other side of the mix resistor, totally floating antenna.

Every later unit has all those weird shields. Note the mix resistors feed through holes in an shield to isolate the mix pot connections from the EQ and C6 input.

V3a's cathode / R22 3M3 connects to....oh boy.....
R22 to a tag strip terminal, which has a wire that runs back up to v3a cathode, then on to pin 8 of the output transformer octal socket, then on to C19.
Then, the tag strip terminal connects to the ground side of the EQ network and the VU meter range pot.
1-4 pots earth go to C1 case, then on to join pin 8 output transformer socket on way to C19. input ground is also routed through C1 case.

The riveted shield on the face between mix pots and eq goes to a riveted tag strip terminal near the ground of R22.

 

[radardoug]

Has it got a built in mains transformer? Is it just field from that?

 

[emrr]

Has it got a built in mains transformer? Is it just field from that?

Yes, and that is one of the ?, could we have 4 with noisy transformers, and other examples with quiet ones? The case grounds are all rivets. Some of these PT's are rusty. There's no trace of noise with the mentioned parts disconnected, with the remaining actives all closer to the PT. You would think the master gain would reveal it if cracked wide open, it naturally runs about 9-10 o'clock. Quiet as can be with the mixer section removed.

 

[Winston OBoogie]

I have not tried the iso transformer, as there doesn't have to be a connection from the mix resistor to the mix pot at all, there simply has to be a mix resistor connected to C6 to have buzz, nothing connected on the other side of the mix resistor, totally floating antenna.

My bad. I read this:

disconnect input 1-4 mix resistors from C6, it still buzzes with only mix 5 mix resistor connected. Then no no buzz once 5 is disconnected - it's a passive input. Maybe I've explained in an alternate way to help describe it.

and incorrectly interpreted it that input 5 mix R alone was the troublemaker. Should have read further. Sorry.


When you tried the external heater supply for V1 - V3, you're confident that this supply was clean?
Sounds like it was cleaner than the internal supply given your issue improved a little so, could your internal be improved at all to get you some of the way there?
Maybe the solution is a combination of small improvements that add up to being acceptable?

I doubt Altec would have added cost to the unit by incorporating those weird riveted shields in later units unless there was a reason. Maybe see if something there helps too.

Can you run a unit from an external HT supply? See if distance from a power transformer helps, or determine if the internal power transformers are noisy fuckers or not.

Grounding for V3 onwards through the output doesn't sound ideal but, as you said, not fun to redo.

Found a pic of a JBL 5210 which is supposed to be the same circuit (with one extra mic in) and the layout seems better in that regard.

 

[emrr]

There's comments around the interwebs about the filament rectifiers dumping garbage everywhere. Suppose I should change them, new better? No idea. EDIT- did - no diff

I did move the filament filter back to the PT side so there's less chance of dirty post rectifier crud bleeding into the first stages, also tried adding another filter section, no difference.

And WTF!!!!!! are there other units out there without buzz?!?!?!