Interstage curious

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
EmRR said:
Well, you've got too much noise without the input transformer too, right?  Swap the input transformers?

The transplants I propose can just be disconnections and wire moves, you don't have to move the whole PSU/etc.

Now that you mention it, I believe I did test this without the input transformer (simulating an instrument input).
I even went so far as to replace everything on the audio board (caps, resistors...), just in case ANYTHING was acting goofy.
We can rule out phantom power.  That's not even hooked up.  There are no switches of any kind in the signal path.  It's as basic as I can get it.

I'll get both preamps on the bench by the end of the week.  One way or another, I'll see what combination of preamp/power supply works or doesn't.  Not much point in introducing the IS and PP stages until I get this licked.
 
You guys are never going to believe this.....
(Then again, you've seen a lot more than I have, so maybe...)

I couldn't leave it alone.  Call it hunch, call it a still small voice, call it what you like, something told me to swap out the tube socket.

Replaced the ceramic tube socket that I think I paid less than $2.00 for, and replaced it with an old phenolic(?) chassis mount socket I've had kicking around for years.  I guess the pin holes are tighter on this one.

THE NOISE IS GONE!!!

THE HUM IS GONE!!!

This thing is now so quiet, I had to check to make sure the filaments were glowing, because I didn't think it was actually working until I plugged in a mic.

I guess I need to spend more than $2.00 on tube sockets!

Hallelujah!!

 
EmRR said:
Also why Gates recommended pulling and reseating tubes every 3 months......


It's one of those mundane things you wouldn't normally give a second thought, and yet it can bring the whole project to a screeching halt and have you questioning your own sanity!  (Along with other components that aren't even the problem...)

Now let's see what happens when I reintroduce the IS and PP stages.  :eek:
 
I couldn't leave it alone.  Call it hunch, call it a still small voice, call it what you like, something told me to swap out the tube socket.
It's the Groupdiy hive mind force ???

Well done!

DaveP
 
Update....

This thing wasted no time knocking me down a few pegs.

I reconnected the IS and PP stages.  Loads of hum, even with the gain pot down.  Tried grounding the PP stage in a few different places to no avail.  Removed PP stage.  Connected my first two stages (which definitely work) to the IS, and then connected the IS right to the output jack.  (I guess I just wanted to make sure it was OK.)  Still worked nice and quietly.  PP stage was still hacked together on a perf board of its own, so I completely took it apart and rebuilt it.  Hooked it back up again, and still lots of hum. 

Tried additional power supply decoupling for that stage's B+.  No difference.

A couple observations....
When I short out the IS primary, the hum disappears (along with the signal).  If I short out either side of the IS secondary, it almost disappears.

I'm thinking ground loop, but can't seem to find it.  The only spot that looked like it might be was where ground is connected to the center tap of the IS secondary.  But that's also ground for the cathode network of the PP stage.  I seem to remember early on disconnecting that center tap and things getting smoky, so I'm hesitant to try that again.  (Maybe it was just the cathode stuff I left floating last time...)

Updated schematic attached....


 

Attachments

  • FE-3_011720_as_built.jpg
    FE-3_011720_as_built.jpg
    493.5 KB · Views: 14
Tried a few things....

Disconnected the center tap of the IS secondary, but kept the PP cathode network grounded.  Still hums.
Tried feeding the B+ to the output transformer primary center tap from a different spot.  Still hums.
Tried different places to ground the PP stage.  Still hums.

I figure this particular output transformer just doesn't like having DC B+ supply running through it under any circumstances, balanced or not.

I thought about just running the B+ directly to the plates of the PP stage and then capacitor coupling those to the output transformer.  I don't think that will work in this case.

 
Went back over this very long thread and was reminded of something.  Work backwards.

I disconnected the first two stages and connected a tone generator right to the primary of the IS.  No more hum!
This tells me that the PP stage does not have a ground loop within itself anyway.

The first two stages are clean and hum-free by themselves, so no ground loops and/or hum when I just go from the first two stages to the output transformer.

But put it all together, and.....hmmmmmmmm.....

Must be something about the way I'm putting them together.

 
Your B+ supply is over engineered, you just need all those big 220uF caps separated by suitable resistors....... job done.

DaveP
 
You keep forgetting the PP stage has been hum free in many iterations.  The two just don't like being connected now. 

Agree with Dave on the additional resistance in the PSU, but that's not the whole deal, you've had that already.  Look at that book again and read the parts on grounding order.  Might be a clue there. 
 
Fought with this thing all weekend and got nowhere.

Read Langham's text on grounding.  Read Blencowe.  Read Thompson-Bell.  As near as I can tell, I am doing all those things.  (Shields connected at one end only, Pin 1 of XLR connected at chassis, shortest possible runs, mains earth connected to chassis at power inlet, etc...)

We know the first two stages as a complete circuit don't hum when connected straight to the output transformer.  We know they don't hum when connected to the interstage and we listen to the output of the interstage.  We know the PP stage doesn't hum when just a tone generator is connected to the interstage primary.  But something happens when both circuits are connected via the interstage transformer to create some sort of hum loop.

I've tried several different ways of connecting each part to 0V Ground.  Attached is a schematic showing a few ways I did this.  (Sorry for the size of the picture...)  Some methods will obviously appear redundant compared to each other, but I'm trying everything I can think of.  Maybe I'm just too dense to see the forest for all the trees.  The "0V Ground" indicated on the schematic is the ground connection at the power supply.  All transformer shields are connected to the chassis.  I also tried to illustrate how interconnecting runs are shielded.  I even tried physically moving things in relation to each other.  No difference.  I guess the good news there is that it didn't get worse.

 

Attachments

  • FE-3_012120_as_built.jpg
    FE-3_012120_as_built.jpg
    996.1 KB · Views: 8
You shouldn't need the shield connections around the interstage.  Paths should be short.  If it's not worse without, leave it disconnected until you make something else better.

Input and output shields should be to the XLR pin 1, not the chassis elsewhere. 
 
The third schematic down is the proper way to wire the secondary ground.

Sometimes the cathode cap (C4) forms a resonant circuit with the primary inductance, so try removing C4 or changing its value.

Personally, I would wire a simple circuit like this point to point, then you are free of the track resistance in your PCB.

DaveP
 
DaveP said:
The third schematic down is the proper way to wire the secondary ground.

Sometimes the cathode cap (C4) forms a resonant circuit with the primary inductance, so try removing C4 or changing its value.

Personally, I would wire a simple circuit like this point to point, then you are free of the track resistance in your PCB.

DaveP


The interesting thing here is that the circuit doesn't seem to care if the center tap on the interstage transformer is connected to ground anywhere or not.
 
EmRR said:
You shouldn't need the shield connections around the interstage.  Paths should be short.  If it's not worse without, leave it disconnected until you make something else better.

Input and output shields should be to the XLR pin 1, not the chassis elsewhere.


Yep, tried all that.

I even got really desperate and tried connecting each side of the interstage secondary to each PP grid with a separate shielded wire, instead of both sharing a shielded cable.  Didn't help.  Tried separate wires with no shield.  No dice.
 
The interesting thing here is that the circuit doesn't seem to care if the center tap on the interstage transformer is connected to ground anywhere or not.
The grids need a path to ground so it has to matter.  I take it you have checked all the secondary windings for continuity?

DaveP
 
DaveP said:
The grids need a path to ground so it has to matter.

That's what I figured, which is why I've been leaving it grounded.  Disconnecting ground was basically a "Hey, let's see what happens when I do this" move.



DaveP said:
I take it you have checked all the secondary windings for continuity?

To be honest, I hadn't, since it seemed to have no problem passing a signal, in spite of the hum.  I just checked it now.  Checking it with a digital Fluke meter, the meter jumps all over the place and never really settles down.  Checking it with a plain old continuity tester, it checks OK for continuity.  Not sure if the Fluke's antics are any indication of a problem.

Wouldn't that be a kick in the dangly bits to go through all this just to find out the transformer is no good!
 
If you're getting reasonable sound audio, the windings should be good.  It'd be thin sounding if open. 

I tend to think the capacitance to ground in the windings may be enough to avert grid problems if floated,  though there should be a path through the CT connection.    You could try 1M grid resistors and leave it floating at the CT. 
 
Well, I think I finally found the problem.  Naturally, it was a result of my own inattention.

As we've discussed, the first two stages are almost identical to the two stages of my previous project.  That being the case, in an effort to keep things neat and make things easy, I used an audio PCB left over from that project to handle the first two stages here.  It was supposed to have the added benefit of eliminating the ground loop issue I started out with when I had these stages bodged together on perf board.

So, what's with all the hum?

I started poking around again with the scope and discovered the tell-tale trace of power supply schmutz appearing on the primary of the IS transformer!  And nowhere before that.  That would put the source of the noise in the second stage.  (Which also explains why it's still there with the gain all the way down.)  Carefully studying the current schematic and comparing it to what was there, I found my error.

On the schematic, the power supply decoupling (C6 and R6) is upstream of BOTH anode resistors for the first two stages.  On my PCB from the old project, that decoupling circuit is only upstream of the FIRST stage.  The second stage was not actually decoupled from the power supply.  Really dumb assumption on my part!  Fortunately, not difficult to rework.  So now the decoupling actually takes place where the schematic actually says it takes place.

And the hum is gone!

Now to reintroduce phantom power to see if I can dig myself into another hole...
 
Brought back phantom power, and still all is quiet.  And so far, I haven't had the switch go bad, so I might have just had a couple lemons.

Brought back the pad switch (input secondary series/parallel).  Still quiet.

Decided to go for broke and threw in an instrument input.  That did introduce some hum.  Tried a couple different shielding schemes with no luck.  Might just cut my losses and skip that feature.  It's for good reason some of us have mixed feelings about those.
 
Back
Top