1987 turret board rewire- Hiss on start up that fades

[Jaymz77]

Hi all,
I have just finished installing a turret board into my 1987x reissue.
When I first powered it up there was a faint hum (normal) then a hissing/ white noise type sound came in. I could hear signal (touching the end of a cable) but after a minute or two the hiss faded out (yes slowly faded) and I could no longer hear signal.
Any ideas what that could be? Nothing popped and I didn't let any smoke out but this weird phenomenon freaked me out a bit.


All help appreciated.
Cheers

[shooter]

did you check DC volts while it faded, anything on the DC side look/act weird?

[pompeiisneaks]

To help troubleshoot it really helps us if you can give pictures, and a schematic.  Otherwise it's just wild guessing.  Do you have a listening probe or an oscilloscope to track down where the signal is disappearing?    It does sound possible that the output tubes are biased to shutoff, possibly, but they slowly get there as they warm up?  If not, it could be a miswired section of the board, but since we can't see that and don't know the exact schematic you're using, we've got nothing to go on.

~Phil

[Jaymz77]

Shooter: no i switched it off straight away. I didn't want to let the smoke out. Plus it was late and my usefulness had run out for the night. Thats something ill check.


Here is the layout i have used. SDM layouts are pretty reliable.


One thing i noticed is that the bias pot is maxed in one direction. I may have to change the resistor value. Im using the stock marshall TX's not the metro ones in the layout.

[pompeiisneaks]

I'd just validate that the negative bias voltage is in range for what is expected.  Do you have pictures?  The layout seems right for the schematic I find for the 1987, just do a double check that you get 25k across the end terminals of the pot and 47k at the other, (meaning it should be max 72k). 


~Phil

[Jaymz77]

Cheers.
Sorry it's early here... what do you mean by the 47k and 72k bit? The trim pot is 25k, what are the other numbers for?

[pompeiisneaks]

To validate the bias values, the schematic has a 47k resistor and then the pot works as an additional 25k variable resistor.  Total at max should be that 72  (47 + 25) or thereabouts.  Just making sure you have the variance you'd need for that circuit.  If the voltages are out of tolerance as well, then that could be a problem.  If the - voltage is too low (too far negative) or too high you'll be out of bias tolerances and either at shutoff or in red plating territory.  The first request was to just make sure that you're getting the right resistance ranges between 47 and 72... Then to make sure the voltages are in range too. 

I'm not sure what the expected negative bias is as there's still no schematic to look at. 

~Phil

[Jaymz77]

Yeah right, of course. Like I said, it was early. 😊
I don't have a schematic that represents the layout. I do have a schematic for the stock 1987x, but I'm at work right now. I have compared the two and they are basically the same with some minor differences. The layout shows a 150k resistor in the bias circuit but the schem shows a 220k. I always knew that 150k seemed like an odd value and expected to have to change it.
As far as the hiss goes, I'm leaning towards a dry solder joint or 6. It was the way it faded out, along with any signal I had, that had me worried.
This could be a job for my new scope meter if redoing the joints doesn't work.
Cheers

[pompeiisneaks]

Good luck, let us know how it turns out.

[HotBluePlates]

D.C. voltage measurements, throughout the amp.

I have no good explanation, but wonder if the bias on some tube goes from normal (passes sound) to very cut-off/high negative bias (no sound).

[sluckey]

Are all the tubes lit when it fades out? May need to turn the lights off to tell.

[Jaymz77]

Thanks guys.
I'll check all these things and report back.

[Jaymz77]

Guys I'm as frustrated as hell right now. Its my birthday today, I'm getting old, i got stuck somewhere i didn't want to be today that ate into my tinkering time, i had about an hour on the amp, nothing achieved and now i have to get ready to go to dinner and won't have time to touch the damn thing until boxing day at least.
FRIGGEN HELL.
I tried playing with the bias circuit and got nowhere. After wrestling with damn components to try and get them out without messing anything else up i still can't get the right bias. Ive tried swapping the 150k for 220k and swapping the 47k for 56k, nothing seems to want to give me the right range. Its like it needs a 64k. I noticed that the bias pot on the original is 20k and I'm using a 25k. that will obviously have some effect but i should still be able to get the right bias but maybe at one end of the pots travel.


The hissing has me stumped. It does it on standby and volume has no effect.


Now i have to wait 5 days before i can do a damn thing. AARRGGHH!!!

[pompeiisneaks]

That sucks.  Yeah you should be able to dial the resistance to maximum or minimum and have a significant effect.

~Phil

[Jaymz77]

So over the next 5 days when I can't work on the amp I should devise a plan of attack.


What could be possible causes of such a bad hiss when the amp is on standby? I mean, on standby there isn't even any HT flowing in the amp, so how the hell...?


I just don't get it.

[sluckey]

What could be possible causes of such a bad hiss when the amp is on standby? I mean, on standby there isn't even any HT flowing in the amp, so how the hell...?

I've never heard of such. I suspect a wiring error. Show us your work, layout, and schematic.

[Jaymz77]

I'll see what can do but being such a small chassis will make it hard.

[Jaymz77]

Here is the 1987x schematic. There are some differences between this and the layout but nothing huge, or so I thought.

[Jaymz77]

Ok gents, I found a small amount of time today, luckily I had been studying schems all day and had a plan.
Turns out I had the ground for the bias circuit connected to the 47/56k turret instead of the junction of the two 10u caps. Silly mistake. However I was forced to change the bias layout on the fly because I actually left out a turret so I can see how it happened.
Anyway the horrible noise has gone. I still have a small amount of noise but is a lot less. Maybe bad solder joints.
The main issue I have to sort out now is that the inputs seem to be cascaded. Both inputs won't work unless I open both volumes. Not sure how I did that.
The whole point of this exercise was to do a switchable cascade mod, I seem to have inadvertently made it like that before I have even done the mod ( I thought it best to just swap the boards and get the new one working before I did any mods).


I'm nearly there, just got to find this issue now

[Jaymz77]

*Update


The tiny amount of noise I have left seems to have something to do with the presence pot. When maxed there is no noise at all.

[Jaymz77]

Ok, well the noise is basically a non issue now. I guess it seemed worse through a set of headphones from a dummy load.
The issue now is that the volumes don't behave as normal. I can plug into either input and crank that volume to 10 while standing a foot away, not much volume really. But if I then crack the other volume that I am NOT plugged into, it then gives me decent volume. And it's also like I have 2 bright channels.


Any ideas? Cheers

[sluckey]

That's got to be another simple wiring error. Check the wires that connect the volume pot wipers to the 470K resistors on the board.

[Jaymz77]

Well I'll check again but the wiring looks fine to me.
Any other possibilities?


Cheers

[sluckey]

If you wire it like the schematic or layout you posted it will work as expected.

[Jaymz77]

I traced the wiring last night. But I'll check again

[sluckey]

I have that exact preamp circuit in two amps. I can play just fine through either channel with the unused channel volume control set to zero. That's the way it's supposed to be.

With the symptoms you describe the most likely suspect is a wiring error. Keep looking. You'll find it. Post some big hi-rez pics and we'll help you look.

[Jaymz77]

Yeah I'll do that tonight. I'm confident it is something simple too.
I spent yesterday with the drawings open on my pc and kept going back every now and then to study. It was doing that that triggered me to what was causing the hiss, checked it when I got home and I was right. Hopefully this one will dawn on me the same way today.
Cheers

[Jaymz77]

Looking at the schem vs the layout (already noted that there are some differences) I can't get my head around how the circuit works around v1a. I don't know if this has anything to do with the issue or not.
Let's see if I can explain what I see correctly.

The schem shows a 100k plate resistor between v1b pin 6 and the 10k filtering resistors. This is not what I see on the layout. The layout shows no resistor between pin 6 and that 10k, instead it appears to me to be on pin 1.
I do remember when I pulled the stock amp apart that resistor was not connected the way the layout shows. I pulled the factory one off so I could replace it with a cc resistor (part of my experimentation process).
I guess in essence I have the plates of the 2 halves of v2 reversed, by the other pins are correctly wired.
Am I seeing this right?
Do you think this might be my issue?


Cheers

[sluckey]

The schem shows a 100k plate resistor between v1b pin 6 and the 10k filtering resistors. This is not what I see on the layout. The layout shows no resistor between pin 6 and that 10k, instead it appears to me to be on pin 1.

I don't know how you can see a 100K connected to pin 1 and not see another 100K connected to pin 6. The two resistors are side by side with the bottom lead of each connected together. Plain as day.

[Jaymz77]

Sorry my bad, V2 I'm talking about, not v1

[sluckey]

Change the pin #s on the schematic. Pins 1,2,3 become pins 6,7,8. Pins 6,7,8 become pins 1,2,3. That layout is based on an old 1987. The 1987X schematic has a lot of differences! If you built this amp with the SDM layout you should also get the SDM schematic. But none of this is relevant to your problem.

BTW, the layout you posted lost a lot of detail when you resized and converted to gif format. Here's a properly resized layout with all the missing detail.

[Jaymz77]

I had to resize it cos it was too big to upload.
I don't have an sdm schem. I could never find one.

[Jaymz77]

I'm not at my pc at the moment so I won't disagree, but what I thought I saw earlier was that only pins 1 and 6 were swapped, the other were the same

[Jaymz77]

Yeah you're right. The two halves are completely reversed. No issue there.
I read something somewhere about the two 470k mixer restistors creating a voltage divider that bleeds signal to earth via the unused volume pot. I would think this would result in lower volume on the used channel, as I'm experiencing.
Why I can't work out is what differences prevented this from happening in the factory 1987x.

[sluckey]

Yeah you're right. The two halves are completely reversed. No issue there.
I read something somewhere about the two 470k mixer restistors creating a voltage divider that bleeds signal to earth via the unused volume pot. I would think this would result in lower volume on the used channel, as I'm experiencing.
Why I can't work out is what differences prevented this from happening in the factory 1987x.

Sure, those two 470K mixer resistors will form a 2:1 voltage divider if the unused volume pot is set to zero. Works that way for every 1987 and 1987X ever built. Your amp is the only one that is misbehaving. It's really a simple circuit. The answer is hiding right before you. Keep looking. It ain't wired right.

I don't think anyone can do anything more without seeing your amp.

[Jaymz77]

Well I'm stumped. I'm out of ideas

[Jaymz77]

So every 1987 has volumes that both controll both inputs?

[sluckey]

Well I'm stumped. I'm out of ideas

Well then, take some pics.

[sluckey]

So every 1987 has volumes that both controll both inputs?

That's not what I said. How do you jump to that conclusion? Using simple mixing resistors to join two channels together is simple, cheap, and provides a lot of isolation between the two channels. But it's not a perfect mixer. There will always be a tiny bit of interaction between the two channels. But 470Ks provide plenty of isolation so volume control interaction is very minimal.

But the symptoms you describe indicate a problem. Under no circumstances should you be able to kill the signal in one channel by turning the other channel volume control to zero. THAT'S NOT RIGHT!

[Jaymz77]

No that's not what's happening.
I'm effectively getting 10% volume out of the channel I'm plugged into and the other 90% from the unused volume. Only the volume of the channel I'm plugged into will completely kill the signal.
I'm getting some pics, but all my leads are connected underneath so you won't be able to see them.

[Jaymz77]

Here

[Jaymz77]

Another

[sluckey]

The issue now is that the volumes don't behave as normal. I can plug into either input and crank that volume to 10 while standing a foot away, not much volume really. But if I then crack the other volume that I am NOT plugged into, it then gives me decent volume.

Maybe your mixing resistors are 470Ω rather than 470KΩ. Check it.

[Jaymz77]

Anoher

[Jaymz77]

Oh my god, you could be onto something, give me a minute

[Jaymz77]

Yep, 470ohm. Friggen idiot

[sluckey]

You aint the first. I've seen at least a half dozen problems with doing just the opposite, ie, using a 470K rather than a 470Ω on the cathode of a LTP PI. I knew it had to be something simple.

BTW, nice work on that amp.

[Jaymz77]

Thanks for the compliment.
This is weird. Volume 2 now has no effect on input 1 but I'm getting bleed through even when volume 1 is on zero. But if I'm in input 2 volume 1 still has an effect.
AARRRGGGHHH

[sluckey]

Temporarily remove the 470pF cap from the 470K mixing resistor. Do both channels behave the same now?

Bleed-thru with the volume on zero sounds more like a pot issue to me, ie, it's not really dialing down to zero ohms. Could also be a ground issue with the pot. Put a ground clip on the wiper of the pot. Does that kill the bleed-thru?

[Jaymz77]

Lifting the cap doesn't change anything.
Earthing the wipers does seem to make it better but not perfect.
Volume 1 wiper earthed while plugged into input 2 makes it bleed better. I can only hear bleed when vol 1 is cranked. It does let more hiss through tho.
Plugged into input 2 with vol 1 wiper earthed does the same but instead of hiss I have buzz, like I'm holding the tip of the input lead