Would love to know more about my '73 Superlead

[neikeel]

I will not go into the cap thing - can of worms, there are reasonable views either way (the tyre/old oil analogy is fallacy, but often used) and there are levels of preventative maintenence to the nth degree (Rolls Royce replace every nut, bolt and washer when they do any work on their cars) but sounds like the people did the right tests and gave sound advice.
I can't tell but are you running 6550s - as that is what the amp is set up for (10k between bias caps, 82k grid leaks (aka 'bias splitters'). El34s tend to have more low end with 220k grid leaks, but you are happy with sound - why change it!

 

[Pete Farrington]

It would be interesting to know how much leakage current each of those 4 HT can caps is drawing. And hence it may be assessed whether the cap heating and plastic film shrinkage is internally generated.
It may be notable that of the 2 nearest the valves, the one for the reservoir node apparently has a lot of shrinkage, whereas the one for the screen grid node possibly none.

 

[PelliX]

ESR measurements. If really want to go further, signal generator + oscilloscope. It is, indeed, as some here have pointed out, a science.

“When you can measure what you are speaking about, and express it in numbers, you know something about it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely, in your thoughts advanced to the stage of science.” ​

― Lord Kelvin

Show me a mojometer that produces consistent results, and I'll consider changing point of view. Until then, I'm siding with the numbers. Music is not necessarily science, but the operation of the circuit in an amp is.

 

[playloud]

Show me a mojometer that produces consistent results, and I'll consider changing point of view. Until then, I'm siding with the numbers. Music is not necessarily science, but the operation of the circuit in an amp is.

Point taken, but I would argue there is actually a "mojometer" here: the amp. If you want to compare, say, two electrolytic capacitors and you record the results of passing one or more waveforms into the amp with each capacitor - all other things being equal* - then you have your numbers: the PCM waveforms, i.e. two strings of floating point numbers of a certain bitdepth (more precisely, these are samples of analog waveforms).

The question is really whether you can draw meaningful conclusions from such numbers. It would probably help if you had a decent sample of each kind of capacitor too, to produce reliable/significant results. The issue here is that the odds of finding someone with the equipment, time, inclination, electrical knowledge and statistical expertise to conduct these experiments are vanishingly small.

Our ears and brain actually do a pretty good job of approximating these conclusions for us, all things considered. Unfortunately, such conclusions are often subjective and not easily explained.

*Obviously this is impossible to achieve perfectly in practice, but with the right tools (constant reference input, reamp box, some form of mains regulation etc.) you can make it pretty consistent, to the point that you could quantify and enforce necessary tolerances with sufficient time and budget.

 

[PelliX]

Point taken, but I would argue there is actually a "mojometer" here: the amp. If you want to compare, say, two electrolytic capacitors and you record the results of passing one or more waveforms into the amp with each capacitor - all other things being equal* - then you have your numbers: the PCM waveforms, i.e. two strings of floating point numbers of a certain bitdepth (more precisely, these are samples of analog waveforms).

But that's not too hard to do, a signal generator costs next to nothing thesedays, and if you prefer 'real world' signal, record your preferred riff on a looper pedal for A/B testing. As for analyzing the resulting sound, one can do this in software quite easily.

The question is really whether you can draw meaningful conclusions from such numbers. It would probably help if you had a decent sample of each kind of capacitor too, to produce reliable/significant results. The issue here is that the odds of finding someone with the equipment, time, inclination, electrical knowledge and statistical expertise to conduct these experiments are vanishingly small.

True, but there are people out there who've done this (and with any luck some of that is still on Youtube). A classic example would be the 4558 op-amps in Tubescreamers.

Our ears and brain actually do a pretty good job of approximating these conclusions for us, all things considered. Unfortunately, such conclusions are often subjective and not easily explained.

I'm going to respectfully disagree there; our mind is an incredibly poor and unreliable tool for this job. Without opening a whole can of worms, our brain didn't evolve to appreciate the finer nuances of signal modulation. It's extremely good at approximating things (as you said, indeed), but that approximation is effectively comparable to downsampling of the input provided and combining it with expectations from previous results. If you've ever taken a swig of orange juice while thinking you were holding a milk carton, you know exactly that it took your brain a moment to shift gears because it was already 'thinking ahead' and expecting a totally different taste.

I've experimented with tone caps in guitars and accidentally fooled myself in the process. A key difference there is that your tone cap can dry up and short out with little or no consequence. I wouldn't risk a (vintage) output transformer. Also, consider the fact that sometime down the line, those 'mojo laden' caps will in fact die. Then what? Go hunting for more that are on the verge of failing, or embrace a component that has identical specs to the old one when it was installed in 1973? Each to their own, of course.

 

[playloud]

But that's not too hard to do, a signal generator costs next to nothing thesedays, and if you prefer 'real world' signal, record your preferred riff on a looper pedal for A/B testing. As for analyzing the resulting sound, one can do this in software quite easily.

True, but there are people out there who've done this (and with any luck some of that is still on Youtube). A classic example would be the 4558 op-amps in Tubescreamers.

I think I should have clarified better. There are obviously various statistics one can calculate to compare the recordings. For example, one can quickly calculate a DFT to compare the magnitudes in a given frequency bin. The hard part is using these derived results to make the overall case that component A is "better" (in circuit) than component B. Even if the null hypothesis is reduced to "component A is no different to component B", you have to be more specific about what differences you care about, and what significance levels etc.

If I presented the DFT example to you, you could readily ask:

1. Which input signals did you use? Are they sufficiently long/varied?
2. How many of each kind of cap did you use?
3. What does the observed frequency discrepancy tell me?

And so on.

Just to be clear - I'm making the case against using the amp itself as a "mojometer" here, and toward using the electrical characteristics of the components in isolation (ESR etc.), this being a more refined and mature science.

There's also an issue of combinatorial explosion. Say I have 10 brands of 100uf filter caps (including dual section cans), and I want to find an ideal combination to fill the 6 main electrolytic slots in a 1959 circuit.* Well then I have to consider literally a million different possible combinations!

*Seems odd to have different brands in the two series positions supplying the mains and/or screens, but a true tonehound would leave no stone unturned!

I'm going to respectfully disagree there; our mind is an incredibly poor and unreliable tool for this job. Without opening a whole can of worms, our brain didn't evolve to appreciate the finer nuances of signal modulation. It's extremely good at approximating things (as you said, indeed), but that approximation is effectively comparable to downsampling of the input provided and combining it with expectations from previous results. If you've ever taken a swig of orange juice while thinking you were holding a milk carton, you know exactly that it took your brain a moment to shift gears because it was already 'thinking ahead' and expecting a totally different taste.

I've experimented with tone caps in guitars and accidentally fooled myself in the process. A key difference there is that your tone cap can dry up and short out with little or no consequence. I wouldn't risk a (vintage) output transformer. Also, consider the fact that sometime down the line, those 'mojo laden' caps will in fact die. Then what? Go hunting for more that are on the verge of failing, or embrace a component that has identical specs to the old one when it was installed in 1973? Each to their own, of course.

Perhaps you're right here. I'm no neuroscientist but I get confused often. I guess I was more marveling at the speed of the brain's computation (generating a high-level impression from this long string of numbers, apparently instantaneously), rather than the accuracy/reliability of the results!

 

[PelliX]

I guess I was more marveling at the speed of the brain's computation (generating a high-level impression from this long string of numbers, apparently instantaneously), rather than the accuracy/reliability of the results!

Absolutely. We appear to have the most accurate directional hearing in the animal kingdom. Directional perception is done in part by the brain comparing the microsecond delay between receiving a sonic pattern by each ear. That's pretty impressive timing, even if the mechanism is self-calibrating to a degree. It's our 'self-calibration' that makes the whole thing so unreliable by adding very complex factors to the equation, I guess.

 

[Matthews Guitars]

As for the capacitors, I choose to always err on the side of caution. I have certainly seen enough examples of equipment damaged where capacitors exploded without enough warning for the operator of it to do more than flinch at the blast.

If you think they're going to give you enough warning to turn off and unplug the amp before they explode and contaminate the guts of your amp with corrosive electrolyte, I have to tell you that you may be overly optimistic. You may not hear any symptoms at all before the BANG.

I've seen inside plenty of amps where clearly the capacitors blew out before they were changed. Marshalls included, of course. The corrosion damage to the chassis and other components near the location of those capacitors is a distinctive signature.

You can do as you wish but we have some people telling you that there's mojo in those clapped out, worn out, overheated turds that were awesome high quality components 60 yers ago, and we have others telling you that they're little quarter sticks of dynamite just waiting for their fuses to light spontaneously and kill your amp and anybody near it. The truth is somewhere in the middle.

But where? How much risk are you willing to tolerate?

I'm risk averse. I see plenty of clear evidence....based on decades of electronics repair experience and not limited to guitar amps...that those need to go and I'd pull them out YESTERDAY. Their fuses are already lit.

I'll trust my experience, my judgement, and the input and advice of capacitor engineers that I have spoken to who work for the company that now owns the rights to the Daly capacitor brand name. (That'd be Kemet.) Incidentally they've allocated some funding to relaunch original-to-spec Daly capacitors for our Marshalls and some other amps. That doesn't mean they're in production yet.

I respect the opinion of some other people here on the board who have chimed in but I respectfully disagree with some of them as well. Their experiences with these things do not necessarily correlate with mine.

 

[Xisco]

Hi,
I was reading this post with interesting as I have a Marshall Super Lead 73. I would be very grateful if you could take a look to the pictures and tell me what changes have been made. I know the filter caps were changed and, possibly some others components. I guess the power tranformer too.
Thanks a lot

 

[Tatzmann]

Quite a frightening "gutshot".

 

[Matthews Guitars]

Added master volume on the back panel, added push-pull pots on both volumes, presumably switchable gain mods. At the very least.

Here are photos of an amp like yours in stock condition. What is different on yours is a modification or component change.

 

[Xisco]

Added master volume on the back panel, added push-pull pots on both volumes, presumably switchable gain mods. At the very least.

Here are photos of an amp like yours in stock condition. What is different on yours is a modification or component change.

Thanks a lot for your help. That is a surprise for me. I didn't know the amp was modded with push pull circuit. Do you think it is easily reversible? What about the transformers? Anyone seems to be original?
Regards