Mesa Subway Blues noise

[goldenGeek]

Hello
I am currently trying to figure out a Subway Blues thats humming too much. I have read some posts about the .01uf mod (replace with .68uf), but I think the problem with this amp is earlier in the path. I will replace the cap anyway, but heres the problem:

There is a loud static and some hiss that gets louder with the volume knob. If I pull tube V1 the hum goes away so I think the problem must be in the first two gainstages - probably between input and before V1b. The volume pot is also scratchy. Theres also some trouble with te reverb but I'll fix that later since the noise goes away with reverb at 0. Does anybod have any clues what to check first? Replace the volume potmeter? Capacitors? Resistors? I have replaced the V1 tube with a couple of other known working ones but they were actually worse than the one originally present when I got it. I can of course replace the whole circuit up to V1b, but if anyone has any experience with this amp I'll be glad to hear from you Oh, by the way, this is probably a newer one, it doesnt have the "mojo module" installed, all the components are visible. Serial number in the 17xx-range if I recall.

This is what I've tried:
-Replace V1
-Replaced .01uf with .1uf (that was the closest to .68 I had at the time)
-Replaced Volume potmeter

I've attached a recording of the amp, turning the volume up from 0 to 10 and wiggling it at the end. The schematics is also attached.


I'm going to trade this amp for a MK3 black stripe thats absolutely amazing so I really need to get this amp in better shape ASAP

[goldenGeek]

Okay, so I chopsticked a bit and some capacitors in the input section is really microphonic. There is a loud tapping in the speaker when I tap the 20, 120 and 250pf before V1a. Is this normal or should I start here? I have never seen that particular input before, usually its just 33 or 68k into V1a, but here its multiple caps and resistors. If thats not the case I guess the next suspect is the tonestack and the caps there? The scratching on the volume, is that DC which in turn points to leaky caps in the tonestack?

[martin manning]

The scratchy volume pot is most likely a leaking cap in the tone stack, so I'd start there. You could lift one end of them to find the bad one(s). Static and hiss are also symptoms of a leaky cap, so listen for the ones that reduce that symptom too.

[goldenGeek]

So I tried some more stuff:
-replaced the input circuit with a "standard" 68k - no improvements, the microphonic when chopsticking improved though.
-replaced caps in tonestack - no improvements. However I didnt have correct values so the eq changes a little bit but no improvements in noise and or scratching in pot.

It seems like the tube socket for V1 is kind of microphonic, pin 1-3-area. Another observation: noise gets worse with Fat-switch in.

Edit. Maybe a good cleaning of tube socket and maybe re-soldering the socket is a good idea?

[FUCHSAUDIO]

Thats a pretty old amp. Have you tried fresh filter caps ?

[martin manning]

Another observation: noise gets worse with Fat-switch in.

That would suggest that the 750p fat cap is at least part of it. You could ground-out the treble pot wiper to confirm that the DC on the volume pot is coming from that side. It could also be grid current from V1b too, since that stage runs at pretty low plate voltage.

If the input network caps are ceramic, then microphonic might be more or less normal for them.

[goldenGeek]

Thats a pretty old amp. Have you tried fresh filter caps ?

No, I have not yet. I'll have to order some, or I could probably try with the 22uf/500V I have sitting on the shelf. The noise has som kind of static, but the worst part of it is that "schhhhhhh"-sound. Like theres a distortion pedal on... I think you can hear it in the sound clip in my first post.

[goldenGeek]

Another observation: noise gets worse with Fat-switch in.

That would suggest that the 750p fat cap is at least part of it. You could ground-out the treble pot wiper to confirm that the DC on the volume pot is coming from that side. It could also be grid current from V1b too, since that stage runs at pretty low plate voltage.

If the input network caps are ceramic, then microphonic might be more or less normal for them.

Yes, the 750p might have something to do with it, but it also provides some kind of "mid-boost-ish" effect as the treble cap goes from 250p to about 1000p when 250 and 750 are combined. The ones I soldered in are 220pf and 500pf silver micas that was the closest I had at this moment. The effect of the fat is not quite as noticable as it was before, but I guess that has to do with the different total capacitance.

I havent had time to go back into my studio to check it out, but I think I recall that when the treble is turned to 0 the noise is as good as gone. I'll have to double check that tomorrow though.

[Stevem]

Do you have a O-scope?

I have had this issue on 2 custom Fender modded builds from people / novist over the years and the it came down to the amp Oscilating in a small way,
And the tip off was the preamp tube socket acting like it was bad yet it was 100% good!
Moving the tube around even a little mad such a racket that you have swarn the tube socket was 80 years old!

This is easy to see on a Scope as many times it's too high of a frequency to be heard, if your Dog leaves the room when the amp is on that's a good indicator that indeed the amp is oscilating ! I kid you not!

In trouble shooting this leave the hiss issue aside for now.

Are the grid wires for the first 3 gain stages shielded , and if so are there grounds which should only be at one end solid ?

Any Ciramic disc caps in the preamp section can be suspect for having bad lead connections in the cap body.

[goldenGeek]

Yes, I have a scope but I'm not very skilled at using it. What should I look for and where? Do I use a signal (sine?) at the input or just leave the input empty?

There is no grid wires as its a PCB-circuit with the pre amp tubes soldered to the main PCB.

[Stevem]

You may have a combination of issues going on here, but as posted old filter caps can make for hum and having gain stages that interact with each other when they should not which can lead to other issue's.

Output tubes being warn and very unmatched can make for hum also .

I would grab hold of that 500 volt filter you have and jump it across the last filter in the preamp section and see what the does to change things.
We can hold off on scope checks for now untill you report back.

[goldenGeek]

I have now replaced the filter capacitors (brand new F+T 33uf/500V) and all of the caps in the EQ-section plus the caps in the input section. It seems like theres no improvement regarding noise. I'm very tempted to use the amp as a boat anchor.

[Stevem]

Ok , let's go back to square one, but first let me state that this amp the way it's built has too many gain stages (5) powered off of that one C labeled power supply filter mode , at most it should have only two gain stages , very poor design here!

Let's deal with the amps hum issue first and do the noise track down latter.

So let's find out what type of hum you have that your dealing with , be it 60 HZ from ac voltage sources or magnetic radiation , or 120 HZ from DC voltage supply's .
Low E on a six string is 82 HZ so is the hum you are hearing below the tone of the low E string, of above it?

If you solder a wire across each 100K plate load resistor one at a time to jump them out and kill all gain which one if either being jumped out kills the hum?

[goldenGeek]

I'll fire up the signal generator when I get home and try to find the frequency that is humming. I'll also try to bypass the 100k resistors - but can you describe a little more in detail what that experiment does?

Thank you for your patience and great support

[Stevem]

When you short out a plate load resistor that gain stage can then make no gain , so if for example shorting across V1A's plate load 100K resistor stops the hum then you have nailed the issue down to that small section of the preamp circuit!

Also here is another question / quick check to make of V1, set your meter to ac volts and report back on how many millivolts of power sipply ripple you read at the power supply node C ?