'93 Fender Concert (Pro-Tube) too much idle hiss

[Maddo]

A few months ago I got me a cheap '93 Fender Concert. Although slightly worn looking, it was in working condition, but more or less noisy.
I thought a full recap and retube will take care of this. I also replaced all pots, since all axles were bent (seems like it fell over on its front some time).
But now, refurbished as it is, there is still too much hiss on idle, with nothing plugged in the inputs and all pots on zero (or 1). It has that "ffffffffffff" sounding hiss, like when you max the volume on your home stereo. Reverb adds some buzz when dialed above 3, but that doesn’t bother me.
When I pull the preamp tubes in signal order, the hiss stops when I pull V5 = the last preamp tube before PI.
Disabling the reverb via footswitch has no effect on the hiss.
I found a service manual with a parts list, they did not use carbon comp resistors in this amp, the plate load resistors are carbon film.
Where should I look next?

Schematic:
https://el34world.com/charts/Schematics/files/Fender/Fender_93_super_concert.pdf

[Latole]

Replace all tubes plate's resistors.

[stratomaster]

If you have a set of alligator clips and wouldn't mind an experiment: set your amp to the clean channel and clip from the wiper of the treble pot from the non ground side of R71 to the junction of R33 & R34.

This will bypass the effects loop circuitry. 

There's a lot there that looks like it would trash the signal to noise ratio.  I haven't done any math to back that up, but it just looks messy.

If the hiss abates when you do this then you've found the main culprit.

It also looks like there's a ton of high end emphasis on the reverb signal.  The coupling cap is ~5x smaller than on an AB763 and there's a bright cap on the Reverb control.  I'd look at snipping that bright cap as it's bringing along a good bit of hiss along with the reverberated signal.

[Latole]

If you have a set of alligator clips and wouldn't mind an experiment: set your amp to the clean channel and clip from the wiper of the treble pot from the non ground side of R71 to the junction of R33 & R34.

This will bypass the effects loop circuitry. 

There's a lot there that looks like it would trash the signal to noise ratio.  I haven't done any math to back that up, but it just looks messy.

If the hiss abates when you do this then you've found the main culprit.

It also looks like there's a ton of high end emphasis on the reverb signal.  The coupling cap is ~5x smaller than on an AB763 and there's a bright cap on the Reverb control.  I'd look at snipping that bright cap as it's bringing along a good bit of hiss along with the reverberated signal.

Just to show what you said ;

[Maddo]

Just took the chassis out to make some measurements. During recap I already noticed that the balancing resistors for the heaters were put on the solder side. I thought that this perhaps was done at the factory as some kind of late service bulletin. But now I noticed that the original balancing resistors are still on the component side, and I measured no continuity on them. So it was just a lazy repair. The original resistors have no sign of burn/electrical damage. As I was soldering out the resistors, there was suddenly a spark at the solder tip as I touched a nearby solder point! So I checked voltage at the big caps and there were still 24VDC, although the amp was powered off for nearly a week. So I measured the bleeder resistors and 3 of 4 of them were also open… Also no burn marks or else. Could this be vibration damage? Anyway, looks like I just discovered a large pile of work - will now check every critical resistor in this thing.

[Latole]

Just took the chassis out to make some measurements. During recap I already noticed that the balancing resistors for the heaters were put on the solder side. I thought that this perhaps was done at the factory as some kind of late service bulletin. But now I noticed that the original balancing resistors are still on the component side, and I measured no continuity on them. So it was just a lazy repair. The original resistors have no sign of burn/electrical damage. As I was soldering out the resistors, there was suddenly a spark at the solder tip as I touched a nearby solder point! So I checked voltage at the big caps and there were still 24VDC, although the amp was powered off for nearly a week. So I measured the bleeder resistors and 3 of 4 of them were also open… Also no burn marks or else. Could this be vibration damage? Anyway, looks like I just discovered a large pile of work - will now check every critical resistor in this thing.

You wrote ;  "So I checked voltage at the big caps and there were still 24VDC, although the amp was powered off for nearly a week."

To discharge; stanby swith must be in Play position with all tubes in.
It was ?

[Maddo]

Normally they would discharge through the bleeder resistors (C33 through R95, C34 - R96,…), but because the resistors were open/defunct, that didn’t happen.

[Latole]

Normally they would discharge through the bleeder resistors (C33 through R95, C34 - R96,…), but because the resistors were open/defunct, that didn’t happen.

You did not say if you put stanby on Play. ?

[Maddo]

The standby switch doesn’t play a role in the discharge process, since the rectifier diodes are in blocking direction in relation to the polarization of the caps.
But yes, I switched power and standby on after it sparked, but the voltage didn’t go lower than 24VDC. I had to discharge manually.

[Latole]

The standby switch doesn’t play a role in the discharge process, since the rectifier diodes are in blocking direction in relation to the polarization of the caps.

I desagree.

[stratomaster]

The standby switch doesn’t play a role in the discharge process, since the rectifier diodes are in blocking direction in relation to the polarization of the caps.

I desagree.

This amp uses a non center tapped secondary with a FWB rectifier. The standby is achieved by lifting the rectifier.  It isn't the usual HT interrupt at the screens of vintage Fenders. 

The bleeders should drain voltage regardless of standby switch position. 

I'm suspecting cold/cracked solder joints before 3 out of 4 resistors failed.  Would be worth inspecting and reflowing suspect joints before continuing the service.

[Willabe]

....  It isn't the usual HT interrupt at the screens of vintage Fenders.

All the old Fender stand by switches switched off the B+ before the OT CT fed, not at the screens. 

[Maddo]

But what would be the audible effects if all of the resistors R95-R98 fail? Or in my case, all but one?

[stratomaster]

If the caps are in good health, there would be no audible effects.  However that is a huge caveat.  Those resistors form a voltage divider that is intended to ensure the voltage across each series cap is roughly equal.  This would happen naturally if the capacitors were identical.  In practice this isn't the case. And all bets are off as the capacitors age--particularly the IC caps in Fender amps from the 90s on.

Series caps are used because the excepted voltage exceeds the rating of a single cap.  It's an easy way to use two lower cost components to get a higher voltage rating at the expense of 2 resistors. If one of the caps or resistors fails in such a way that the voltage rating of one caps is exceeded then the whole totem pole is likely to fail and would cause excess ripple on the supply and/or a big draw down of HT. You can set your meter to AC and measure the ripple.  You can also measure DC voltage drop across each cap to check for gross mismatch between the two. These two measurements would be very valuable for troubleshooting.

If you indeed have a component issue and not just bad joints, I would think something is wrong with your caps (or at least was wrong with them before the recap) as it is very strange to have resistor failures at this location. 

[Maddo]

To bring this thread to a conclusion:
Meanwhile I repopulated the whole pcb with the finest components, like higher wattage metal film resistors, ultra fast diodes, new opamps, z-diodes…etc. Like 90% of all components replaced (at least everything in signal path). Hiss still there. I have to accept that it is inherent to that amps design (too many tube gain stages).
So the only easy remedy I see is a master volume. Due to a jumper wire (JP3) between preamp and PI it is really easy to implement a master volume pot here. I already successfully tested this with a 1M log pot and a 100pF bright cap. The hiss is barely noticeable at 1/2 master vol, but it can still get pretty loud with the channel volumes.
Only task left is to find a nice location to place the pot finally without disturbing the aesthetics.

[stratomaster]

Did you try this yet?

If you have a set of alligator clips and wouldn't mind an experiment: set your amp to the clean channel and clip from the wiper of the treble pot from the non ground side of R71 to the junction of R33 & R34.

This will bypass the effects loop circuitry. 

There's a lot there that looks like it would trash the signal to noise ratio.  I haven't done any math to back that up, but it just looks messy.

If the hiss abates when you do this then you've found the main culprit.

It also looks like there's a ton of high end emphasis on the reverb signal.  The coupling cap is ~5x smaller than on an AB763 and there's a bright cap on the Reverb control.  I'd look at snipping that bright cap as it's bringing along a good bit of hiss along with the reverberated signal.

Just to show what you said ;

[Maddo]

@stratomaster:
I have to admit I forgot to test this. But thanks for the reminder, I will try this tomorrow before I make any further steps.

[Merlin]

I suspect a classic case of the hissy reverb mix resistor R54. It can possibly be bypassed with a 10-22pF cap as in most Fenders, to shunt high frequency hiss.
Make sure the FX loop is set to max headroom, to remove R39 from the signal path.

[Maddo]

@merlin:
Thanks, I will try that, too.
I already was wondering about the 10pF cap you see in nearly all similar Fender designs, like the DR or even the '83 Concert II.
But why they forgot it in mine?
For R54 itself I already tried different values: 1M 0.5W MF, 2M2 0.25W MF and 3M3 0.25W MF; but it barely had an audible effect relative to the noisefloor.

[Maddo]

Tried all the suggestions:
- 10pF cap over R54: no effect
- jumper wire from R71 to R33/34: no or minimal effect
- jumper wire from R71 to R33/34, V3 pulled: no or minimal effect
- reverb pot bright cap desoldered: no effect
- snubber caps (220-470pF) over plate resistors on V1 and V5: takes a little edge off the hiss

(All tests are regarding idle noisefloor with no cable in input, all pots zero'd, reverb tank disconnected)

So, a master volume is still my preferred fix for this, if no one has any better ideas left.

[stratomaster]

Have you tried strategically grounding grids and bypassing plates with a large cap to isolate the offending stage?  Turning the volume down essentially grounds the grid is V1B, so you can safely rule that tube out.  Pulling V5 kills the reverb recovery and all the preamp feed to the PI, so all that tells you is the hiss is originating somewhere between V1B plate output and PI input.   That is a lot of unexplored circuitry.  In later reissues it's common to see added gridstoppers at all preamp stages (10k if memory serves) from the factory and some go so far as to cut traces and use shielded coax for further reduction of background noise.  I wouldn't suggest shotgunning a ton of gridstoppers but won't rule out that one or two strategically deployed wouldn't help the issue. 

Further exploration and troubleshooting is needed before masking the problem with a MV.  I refuse to accept that it's just a noisy amp and nothing can be done to improve it.

[Maddo]

You mean grid stoppers like in this amp?: https://el34world.com/charts/Schematics/Files/Fender/Fender_pro_reverb_pro_tube_guitar_amplifier_schematic(1).pdf

On V1 und V5 this would be easy to implement, because the sockets are chassis mounted with separate leads.
Iirc I already strategically grounded the grids, but only on V5B it went quiet. But I used a 100nF cap as the probe's tip. Should I test again with full grounding?

[stratomaster]

100nF is fine. You can do that with the plates too and essentially AC ground the signal out of the plates. 

As far as gridstoppers that's close, but if you look at something like the 68 custom series they just added them everywhere--needed or not. Reverb and tremolo included. 

[Maddo]

Took another round of grounding, tested every plate and every grid on the preamp, and the only notable effects were:
V5B - plate: complete silence
V5B - grid: loud mains hum

Also disconnected NFB to proof that there isn’t something coming from there.
But that also made the hiss a bit louder.

Soldered a 10k grid stopper at  V5B, but it also with no effect.

[stratomaster]

Took another round of grounding, tested every plate and every grid on the preamp, and the only notable effects were:
V5B - plate: complete silence
V5B - grid: loud mains hum

This is very curious.  The only thing I can think of that would cause this is if your cap still had a charge from a plate when you attached it to the grid. 

This the why I prefer to ground grids and bypass plates. 

Before you add your MV I'd be curious if grounding the grid directly causes the amp to go silent as well.  I know that's a rather high impedance node already with the reverb mixer preceding it.  I have noise problems with AB763 type amps with a similar topology.  You could probably get away with a much larger gridstopper, but at that point it'll likely audibly affect the high end.  Keeping that grid connection as short as possible and perhaps shielding it would be my next steps, along with rolling tubes in V5.

[Maddo]

Yesterday I installed the MV; but also one thing I’d like to try is new tube sockets. Although I cleaned and re-bent them, V5 is sensitive to tapping (loud popping). I tried different tubes there: JJ, Sovtek, TAD.
I will try to install pcb mounted sockets for V1 and V5, perhaps these 2 inches of offboard wiring to the chassis sockets are the culprit.

[Maddo]

To end this thread, in the long term the MV solution was a very good conclusion. Made the amp very usable at home, no drawbacks so far.

I connected it in place of a jumper wire just before the PI input. At the pot is a small perfboard with 22nF blocking cap and a 47pF bright cap. The cable’s shield is only soldered to the pot casing and signal ground to minus of a filter cap.

[joesatch]

i have one of these amps. i gutted it and put in an ab763 circuit. made it a super reverb.