Strange issue with 1968 vs 1969 Marshall Super Lead build

[shakti]

I am wondering if those of you who have owned/tried both 68 and 69 Super Leads could chime in with your knowledge. Definitions first; by 68 I mean the type with the filter caps under the chassis (typically 50uF mains, 16uF screens, 32+32uF preamp). By 69 I mean the type with all caps on top of the chassis (typically 50uF mains, 50uF screens, 50+50uF preamp) and the output transformer rotated an moved farther away from the power transformer.

My own experience with these amps is limited to clones, except I have owned and played many 1970-72 Super Leads, as well as a converted 1970 Super PA.

I first built my 68 clone to minute detail, using NOS mustard caps and Iskra resistors, RS silver mica caps etc. You know, the works... Merren transformers. I also used NOS filter caps, Erie 32+32 for the preamp and 2xErie 100uF for the mains filter. It sounded great, but always very gainy and quite compressed with little attack. I just figured it was supposed to sound like that, being a Van Halen type amp and all, but I kept tweaking it to have a little more clean headroom and punch. I increased the screens filtering to 50uF (2 x NOS Erie 100uF caps in series) and used the 490V tap on the power transformer. Negative feedback is 47k @ 8 ohm.

Then a little while later, I decided to build an all out 69 clone. The amp was built using almost the same components and circuit. The opnly difference is WIMA TFF 0.022uF caps in place of mustards and WIMA MK series caps for the two 0.68uFs (being true to a very specific batch of Feb/March 69 SLs). Same Merren transformers. I used new filter caps on this one though; all ARS 50+50 caps.

Here's the problem; the two amps now have virtually identical specs as far as the circuit. The 68 now has a variable feedback pot, but even messing with that does not change its fundamental tone and feel much. The 68 also has 32+32uF preamp filtering, but apart from that, these two amps are exactly the same, circuit-wise. I am using a self-contained bridge rectifier on the 69, and four individual diodes on the 68, but the type of rectifier is fundamentally the same, with the PT center tap between the two mains filter caps.
No volume 1 bright cap in either amp.

Even when the two amps are virtually the same, they sound quite different. And more importantly, they *feel* very, very different. The 69 has a fast, clean, immediate attack with a lot of punch. The 68 is much more forgiving, in fact, it's rather soft sounding. It has much more compression, and does not clean up well at all. Here's a short video I made where I A/B them through the same cap. Same amp settings. Shitty sound, but towards the end of the video you can hear how one amp just does not clean up:

I should mention that in the above clip, the 69 actually had 6550 power tubes, but the difference remains consistent regardless of tubes. In fact, I have tried almost everything to level the playing field:
1) I put in the same type of tubes in both amps, biased the same
2) swapped preamp tubes back and forth, no change
3) I swapped the pre- and power amps between the amps. The difference in punch was less noticeable then, but the lack of clean-up and headroom seemed to follow the power amp (i.e. from phase inverter and out)
4) I have tried putting the 69 choke into the 68 amp - no difference
5) I have tried to replace, one by one, all the filter caps in the 68 with new production caps. No difference to speak of. Admittedly I have not done them all at the same time. But the old caps all measure fine for capacitance, ESR and leakage.
6) Replacing all the components in the phase inverter - no change
7) Replacing the 560pF cap in the tone stack - no change
8) Idle voltages in the amp are all very comparable

I should also mention that the 68 has always behaved impeccably - no hum, no extraneous noise, no unstable operation, no squeel even with all controls on 10. However, in my A/B testing I did hear some artifacts in the breakup at certain settings/combinations. Here's another clip which demonstrates a harsh breakup which happens particularly with one guitar with the 68. This nasty distortion isn't as obvious with any other guitars FWIW:

[shakti]

I bought an oscilloscope to try to get to the bottom of it, thinking perhaps there is some sort of parasitic oscillation going on. I can't really see anything on the waveforms that looks like a parasitic though, but I am a complete novice with an oscilloscope. The waveform does seem to start to distort in the phase inverter with a 0.25-0-30 mV sine wave and volume up to 2-3 (all tone controls on 10).

I am slowly pulling my hairs out over this one, and am open to any suggestions. Is it normal for two amps to vary this much? Or is there something wrong with my 68? It sounds killer cranked, but I am really missing some clean headroom, attack to the notes, punch and clean-up.

FWIW; my 69 sounds much closer in terms of punch, feel and attack as well as clean-up and headroom/breakup to any of the 70-72 Super Leads I have tried. Those of you with experience with both 68s and 69s - how much do you feel they vary in attack and headroom at cleaner settings? What about punch?

Finally, I have the OT secondaries on the 68 entering the chassis closer to the PT just as on 68 originals. However, I have them running straight up towards the rear of the chassis, running between V6 and V7, then bending 90 degrees towards the impedance selector and speaker jack. In doing so, they are both running closer to the PT, as well as parallell to the heater AC wires for some distance. On the 69, the orientation is different, so the secondaries are shorter, and are run between V5 and V6 straight to the impedance selector. Thus they are not as close to the PT, and not parallell with the heater wires. Could AC in the heater wires or OT somehow affect the output and response of the amp? If anything, the 69 has *more* hum/buzz than the 68 even if they are both very quiet by Marshall standards (using a modified local star ground system which works beautifully in all my other amp builds).

Sorry for the long and rambling post which I tried to make as clear as I could. I am open to all suggestions. Thinking about a complete reuild of the 68, but I would love to be able to diagnose the problem, if there is one, more precisely.

I'll try to get some photos of my amps uploaded somewhere.

[shakti]

Don't know why the Youtube links don't work, so here the addresses are.

! No longer available

! No longer available

[Ugly Distortion]

And more importantly, they *feel* very, very different. The 69 has a fast, clean, immediate attack with a lot of punch. The 68 is much more forgiving, in fact, it's rather soft sounding.

I'd start by tossing into the rubbish bin those 40y/o Eries and recapping the '68 copy with fresh e-caps, all fresh e-caps including bypass.

[shakti]

Well, I did try to do that, albeit one by one. And as mentioned, they measure fine, and were slowly re-formed on first fire-up. I won't rule out the possibility that the cumulative effect of the old caps could cause these symptoms, but replacing both main filter caps with new F&T 50+50 caps did not change it to any appreciable degree (except for making it sound just a hair colder and less "interesting"...hard to put into words. It did not improve clean-up or punch).

I should emphasise that the amp seems to have just a touch less output than my 69, but most importantly heaps more compression and less dynamics.

[Ritchie200]

First of all it is very apparent that you have WAY too many Marshalls sitting around and I would be glad to let you store a few here - they would be in good company and I promise to never turn it on...

Boy that sure sounds like blocking distortion but if there is no change with the feedback pot, it would not be the typical issues associated with that problem.

Double check the values on the 1M Iskra's on the phase inverter.  Or did you change those out with no difference?

Jim

[shooter]

bought an oscilloscope

If your problem is more than corksniff nuance you should be able to side by side the two walking the scope from in to out, note any changes in voltage, look close at the peaks of the sine wave, look for differences like flattening, sharper than original peak, the trace *seems* to un sync, there should be a notable difference, somewhare

[shakti]

First of all it is very apparent that you have WAY too many Marshalls sitting around and I would be glad to let you store a few here - they would be in good company and I promise to never turn it on...

Boy that sure sounds like blocking distortion but if there is no change with the feedback pot, it would not be the typical issues associated with that problem.

Double check the values on the 1M Iskra's on the phase inverter.  Or did you change those out with no difference?

Jim

Ahem...yes, I have been known to hoard Marshalls... They are all clones though, but all built with as many NOS parts as I could find. Love to hear the fine differences between all the different variations over the Plexi theme, so I have basically a 66, 67, 68 and 69 100-watter, as well as a block logo JTM45 (Radiospares OT) and a 67/68 50w (aka Bass). And also a 2x12 18w. They're not going anywhere though...

I have thought about blocking distortion yes - because the amp has always sounded very gainy, but with a kind of sluggish feel, almost like it's sagging a little bit. You know the sound of Neil Young's tweed Deluxe cranked to the max...this is nowhere close to that, but it has a little bit of the same quality to it. But disconnecting the feedback loop did not change it, neither does varying the amount of feedback with the feedback pot. I have a 47k resistor in series with a 250k pot, connected to the speaker jack, which means it starts (typically) with 47k @ 16 ohm and feedback is essentially lowered from there. This way I can cover all the ground of the 68-73 Super Leads, which went from 47k @ 16 ohm (speaker jack) on the 68s to 100k @ 4 ohm for the later ones.

When I wrote that I changed all the components in the phase inverter, that's not exactly true...I changed all the caps in the PI, but not the resistors. Those Iskras hardly ever drift though, but it's something to check out, definitely.

[Ritchie200]

You cant be married and if you are, she's a keeper!

The only other thing I can think of is OP tranny?  Or did you pair the preamp on the 68 to the power amp of the 69?

Jim

edit - Just re-read your post.  I see you did do that and it followed the power amp section.  Yeah first check those resistors and then maybe do a tranny swap?

Jim

[shakti]

A transformer swap is an obvious thing to do, but I hesitate because it's a bit of a hassle to pull the OT from the 69 build, which is super clean and nice...and the OT has those mounting brackets on it like the original.

I made a thread over at the Plexi Palace where the OT builder is also contributing. Check it out...with a new video added:

http://vintageamps.com/plexiboard/viewtopic.php?f=1&t=110976&p=1053595#p1053595

[shakti]

bought an oscilloscope

If your problem is more than corksniff nuance you should be able to side by side the two walking the scope from in to out, note any changes in voltage, look close at the peaks of the sine wave, look for differences like flattening, sharper than original peak, the trace *seems* to un sync, there should be a notable difference, somewhare

It's definitely beyond corksniffing level. Just haven't had time to have them on the bench side by side so I could compare the traces. I couldn't see anything obvious on the 68. *Maybe* some slight blurring of the peaks of the sine wave starting after the tone stack, but then again...at what level would I begin to see some distortion of the waveform in any amp? As mentioned I am a complete novice with oscilloscopes. I don't have a proper signal generator, so I just used an online tone generator. What is an appropriate input voltage? I compared with a guitar signal and ended up setting the online tone generator so I got about 30mV going into the amp (measured before the 68k input grid resistor).

[shakti]

You cant be married and if you are, she's a keeper!

Forgot to say; yes, she's a keeper! And so are the amps... 

[HotBluePlates]

I'm wondering if the ratty sound could be an oscillation. You were saying one amp doesn't clean up... Is that at any volume level?

If it's a wiring issue leading to oscillation, it would go a way toward explaining why the problem following the power section used.

[shakti]

Oscillation or blocking distortion are two hot leads, yes. But I was unable to detect anything serious on the oscilloscope. I need to go back and test more thoroughly. After reading around on the net, it seems that 30mV was a very low signal level to apply, so I should try it with a hotter signal. It was however plenty enough to cause a very loud signal through the speaker, and enough that the screen voltage dropped well below 400V when volume was turned up to 7-8.

If I turn the guitar down to maybe 1 it seems to clean up, but it does react quite differently from the 69. I have tried to move around wires to see if it does anything, but no success yet. I am going to rebuild the power section in any case. These amps typically had a mounting bracket on the OT which I have used on the 69 but not the 68. Not that I think it matters at all, but I am going to put the bracket on the 68 as well for appearance, and in the process I will rewire the OT and run the secondaries in the more typical fashion between V5 and V6.

[shakti]

An update;
I got a tip on another forum from transformer guru Chris Merren to try this trick with an AM radio:

I have been trying to figure out what this phenomenon signifies. What I *think* it may mean is that there is a very high frequency (like in the 100s KHz region) modulation going on, as that is where AM radio signals are happening. This sort of very high frequency noise can come from rectifier ringing, happening when the small capacitance of the rectifier diodes works together with the inductance in PT secondary wires and sets of a very high frequency ringing. This can spread capacitively to other areas of the amp, and/or picked up in nearby wires acting like antennas. Another thing could be the switching noise from the rectifier diodes when they reverse bias which can spread via the heaters or other nearby parts.

Original 68SLs did not have any snubber-style caps on the diodes, but later amps did. My 69 clone does not have them, but does not have the sound artifacts of the 68 clone.

[shakti]

And another thing; I may have spoken too soon about there not being anything special to see on the oscilloscope. I made a few videos which show the waveforms in different parts of the amp and with different vol settings. Take a look:

[Ritchie200]

Were you able to check those two resistors?  Any "imbalance" could cause the artifacts you are hearing. 

Jim

[shakti]

Have not gotten to that yet. What I did do was an OT swap with the 69, which didn't change anything appreciably. I have just received some parts, and intend to do a partial rebuild of the whole amp, starting with the power amp and output section, then working backwards through the preamp.

[Willabe]

Were you able to check those two resistors?  Any "imbalance" could cause the artifacts you are hearing.

Have not gotten to that yet. What I did do was an OT swap with the 69, which didn't change anything appreciably. I have just received some parts, and intend to do a partial rebuild of the whole amp, starting with the power amp and output section, then working backwards through the preamp.

Why not try what Jim suggested 1st before a rebuild? Much easier, it's only 2 resistors. 

He has a lot of experience with Marshall's.   

[shakti]

Will do, but the scope plots in the videos above indicate that something is happening also *before* the PI. I suppose there could be more than one issue going on here; bad cap in the tone stack was one suggestion (it's not the 560pF, I checked that with a replacement), imbalanced grid leak resistors in PI is another, and finally rectifier noise leading to parasitic oscillation or high frequency interference is yet another. So I want to weed out as many possible factors as I can. Besides, I had intended to clean up the OT wiring and add the mounting brackets anyway.

[HotBluePlates]

An update;
I got a tip on another forum from transformer guru Chris Merren to try this trick with an AM radio:
...
I have been trying to figure out what this phenomenon signifies. What I *think* it may mean is that there is a very high frequency (like in the 100s KHz region) modulation going on, as that is where AM radio signals are happening. ...

Merren had you create a form of a "search coil".

Because the radio picked up noise exactly, and only, when you plucked a guitar string, it is unclear if you were just picking up a radiated field from the output transformer or if something more was happening. Try that test again but with the continuous test tone you used for scope work.

If you get an actual oscillation and want to see this noise on the scope, you have to use your 1kHz test signal, but shorten the time base radically to go from ~1ms to ~1µs. Any longer timebase closer to 1ms (1kHz), and the oscillation will just look like fuzziness or lack of focus of the trace. In other words, you are zooming very close in on a small portion of the wave generated by the 1kHz test tone.

I don't think I saw either grid stoppers or a cap from phase inverter plate-to-plate. If you suspect oscillation and have some pF-range caps handy, start by adding a cap from plate-to-plate at the phase inverter. You might want to kick off with a too-big value like 250-500pF, just to see if the noise issue stops (yes, your amp will sound very dull). If better, start reducing the cap value until either normal amp tonality or the oscillation returns. If oscillation returns first, you might be stuck with the higher cap value & a duller sound.

You can also try grid stopper resistors (either approach is "right" though grid stoppers might be better). You might start with at least 1.5-2kΩ per pin 5 of each output tube, moving up if the noise doesn't stop. When the resistors get to a high enough value, they will dull the amp's sound as well.


The videos showed a lot of time devoted to concern that a late-amp signal was a different absolute-phase than an earlier signal. This is a non-issue, and is 100% a consequence of having reactive components (coupling caps) in the amp circuit. You will be unable to hear any difference in absolute-phase unless you're playing your guitar through multiple amps at the same time, in which case opposite-phase/polarity is the only major concern.


I saw the "distortion kink" when you turned the amp volume up high enough. What we didn't know is what the peak signal level is when that happens, or its relationship to the bias voltage. Strange waveforms *will* happen when the peak output of the phase inverter exceeds the value of the bias voltage. That is, if the output tubes are biased at -45v and the output of one side the phase inverter exceeds ~44-45v peak.

This is because the output tube will start passing grid current, which will tend to charge the coupling cap and shift the bias voltage (i.e., "blocking distortion") and the apparent load to the phase inverter drops suddenly during that grid current causing the phase inverter to distort more/differently.

[shakti]

Thank you for a very detailed reply!

I can try to zoom much closer in on the waveforms to see if I can detect any noise.

I do have grid stoppers on the output tubes; 5.6k for each output tube. Exactly the same as in the 69 build.

I also do have the "fizz" cap across the PI plates, but I took it out during troubleshooting to see if it might have been faulty, and forgot to put it back. I rebuilt the bias circuit and PI today so the cap is back in (actually another NOS RS 50pF cap). Nothing has changed with my rebuild. I could try a higher value fizz cap, but I am pretty sure I clipped in another higher value earlier, and it didn't change the fundamental problem. I tried new coupling caps through the whole PI, but no difference. As mentioned earlier, the 1M grid stoppers in the PI were measured, and they are well matched and good values at about 970k each. I didn't replace those or the 470R or 10k, but all were measured and well within spec.

As mentioned it was my first time with a scope, so I was unsure whether the phase shift was significant or not. I put my 69 build under the scope and it is doing the same thing, so that is obviously not a problem. The 69 does, however, not distort the waveform in the same way. But I am using a 12AT7 in the PI slot currently, so that may account for it. I am building my way backwards through the preamp and hope to find it eventually. Could be a faulty part *somewhere*. The amp doesn't sound terrible by any means, but *something* is not right.