Hiwatt DR103 baffler

[Sansteeth]

Hey folks,
grab a tea, a coffee or whatever beverage you need, it's a long one

Schematics and layout Warning: the preamp schematic is good up until V2, then the layout is more accurate and it reads reasonably well as a schematic.
https://hiwatt.org/Schematics/DR_Pre4Input_v2.pdf
https://hiwatt.org/Layouts/PreampLayout2.pdf
https://hiwatt.org/Schematics/DR_100wOutput.pdf
https://hiwatt.org/Schematics/DR_100wPS.pdf
Someone brought a old Hiwatt DR103 to the shop the other day, says it has a weird and ugly distortion, that appeared gradually, as in it would clip a little bit one week at rehearsal and he wouldn't really mind it, to the point where a couple of weeks later it started to happen all the time to the point that it's not playable for him anymore.
He uses a Ampeg Dan Armstrong guitar with a humbucker slid in (they have swappable pickups).
When I play it with a cheap Squier strat with weak pickups, I don't get the ugly clipping even with gain dimed, just classic hiwatt crunch.
But when I play it in the shop with a Les Paul that has a pickup noticeably hotter than his (15/16 Kohms of DC resistance, Alnico V)
Clipping occurs on the lowest notes hit hard (measuring around 1Vrms of pickup output)
occurs even when the master is really low. -> doubt the problem is anything post PI
occurs regardless of tone control pots position (max or half)
occurs on Brill and Normal channel alike (happens when they are bridged too obviously)
occurs on hi and Lo gain input of both channels
occurs even when gain pot is at 9 o'clock the way to max (obviously gets worse as we increase it)

Swapped all the tubes one by one (including a brand new set of EL34, you never know): doesn't solve the problem
Measured the voltages [edited after PRR's reply #3] :
         
V1a
Vp: 161.6Vdc
Vg: 0
Vk: 1.35

V1b:
Vp: 158.4Vdc
Vg: 0
Vk: 1.38

V2b
Vp: 200Vdc
Vg: 0.003
Vk: 1.32


V3a
Vp: 302.6Vdc
Vg: 56.5
Vk: 63.7

V3b
Vp: 162
Vg: 0
Vk: 2.84


V4a
Vp: 331.7Vdc
Vg: 63.4
Vk: 65.5

V4b
Vp: 328.7Vdc
Vg: 57.9
Vk: 65.5

Powertubes all have 446Vdc on the plates and 438Vdc on the screens, bias slightly under 50% of power dissipation (thanks to a bias supply that was installed by a previous tech).

At max gain, unbridged on Normal input, V1 doesn't clip even if I hit it hard with a signal generator (1Vrms to match hard strumming with a humbucker), I clipped the scope on the pot end of the 22nF coupling cap, it all comes out clean.

However, still at max gain, on the grid of V2 (scope clipped on the grid end of the 470k resistive mixer), I'm getting severe clipping on the top part of the signal and less amplitude compared to what comes out of the plate of V1. [scope pic 1]

With the scope clipped on the plate of V2 I get an amplitude similar to what's coming out of V1 but with the same square-ish clipping on the top part you see on V2's grid and a smoother clipping on the bottom part (the peak is rounder) [scope pic.2]

So what I'm thinking is grid-current limiting happening on V2, clamping the signal before it even gets to the tube. A leaky coupling cap upstream (the 1nF or 22nF) but none of them are leaky and measure at .8nF and 22nF respectively.
I measure 102k on the plate resistor that is dropping 135Vdc to get 200Vdc on the plate and 1k on the cathode resistor (that is not bypassed as opposed to what the schematics say) dropping 1.35Vdc.

I was also thinking it's regular blocking distortion, as in, the amp is not meant to take such abuse from humbuckers and it messes with the interstage filtering (wiiiiide band) and creates blocking distortion (with those two 1nF and 22nF coupling caps) but it has two fairly large(?) grid stoppers in between that should reduce the charging time of the caps. Also that doesn't jive with my customers mentioning a problem that got worse with time. Shouldn't be inherent to the design but it must be a part failing. But all those around V2 are pretty close to original specs, so what's going on??
Problem now is that I really zeroed-in on V2 and the grid clipping, I'm in too deep by process of elimination and I can't see the bigger picture anymore, I need some fresh thoughts because I'm going crazy, what do y'all think?

Thanks for reading the whole thing and thanks in advance for any pointers, things I've missed, shoot away!

[Latole]

".....Les Paul that has a pickup noticeably hotter than his (15/16 ohms of DC resistance, Alnico V). "

You mean ; K ohms.

[Sansteeth]

Yes! 

[PRR]

> Clipping occurs on the lowest notes hit hard (measuring around 1Vrms of pickup output)
....occurs even when gain pot is at 9 the way to max

If the VOL pots are anywhere near top, the voltage gain through 1st stage and 2nd stage is over 1,000. If you apply 1V input, this implies 1,000 Volts out of the 2nd stage, and of course it clips.

If you want "slight" preamp clipping, turn Master way low, advance VOL just to the edge of clipping. Then adjust room level with Master.

V4a
Vk: 63.7
V4b
Vk: 2.84

These two cathodes are strapped together. How can they be at different voltages? Are the drawings wrong or are the readings wrong?

[Sansteeth]

Hey PRR,
thank you for your response!

V4a
Vk: 63.7
V4b
Vk: 2.84

These two cathodes are strapped together. How can they be at different voltages? Are the drawings wrong or are the readings wrong?

My bad: on the hiwatt chassis, V4 (PI) is physically before V3 but obviously electronically after, that was too much for my brain yesterday, I corrected my first post.

> Clipping occurs on the lowest notes hit hard (measuring around 1Vrms of pickup output)
....occurs even when gain pot is at 9 the way to max

If the VOL pots are anywhere near top, the voltage gain through 1st stage and 2nd stage is over 1,000. If you apply 1V input, this implies 1,000 Volts out of the 2nd stage, and of course it clips.

Ohh my bad, I meant 9 o'clock to max, meaning from 3 to 10 approximately. A Hiwatt DR103 shouldn't start clipping so early, or if it does, it does in a world-class kind of distortion, so I'm not hunting down "clipping" (it's a rock'n'roll amp), I'm hunting down a particularly ugly kind of clipping that, according to the owner, wasn't there before and eventually got so bad that he knocked at my door.

And because a little sound clip is better, here goes:
https://on.soundcloud.com/v74ho

[tubeswell]

Did the owner guy try new tubes already?

[Sansteeth]

Hey,

Did the owner guy try new tubes already?

I did, all of them, exactly the same issue