Thoughts on Jimmy Page’s Super Bass Marshall?

[WhopperPlate]

A lil birdie told me:

Super bass spec

Treble channel 560p 270k v2 grid

.68uf cathode bypass v2

150p treble bleeders on each plate of PI (with 47p snubber)

220k bias with 6550s

[Roe]

A lil birdie told me:

Super bass spec

Treble channel 560p 270k v2 grid

.68uf cathode bypass v2

150p treble bleeders on each plate of PI (with 47p snubber)

220k bias with 6550s

560p 270k v2 grid could be standard values for the channel mixer? Or is there a big grid resistor in addition with a bypass cap?

[WhopperPlate]

560p 270k v2 grid could be standard values for the channel mixer? Or is there a big grid resistor in addition with a bypass cap?

Nothing additional …just channel mixer… bright channel of course

Also, 47k nfb off spkr jack

[cdemike]

A lil birdie told me:

Super bass spec

Treble channel 560p 270k v2 grid

.68uf cathode bypass v2

150p treble bleeders on each plate of PI (with 47p snubber)

220k bias with 6550s

560p 270k v2 grid could be standard values for the channel mixer? Or is there a big grid resistor in addition with a bypass cap?

Nothing additional …just channel mixer… bright channel of course

Also, 47k nfb off spkr jack

Installed these over the weekend. I had my doubts, especially with the phase inverter plate bypass caps, but this sounds really good (assuming you like the Page 1973 thing). Caveat: as mentioned earlier in this thread, my amp makes 50w, so not all of this will work the exactly the same way as in a 100w, though I likewise adapted this spec to make the amp behave as much like a 100w amp as I could; the only place I deviated from this spec was moving the 47K to the 16R tap to get the same magnitude NFB signal reaching the PI.

I ended up using this as a starting point to dial in a sound that I thought sounded closer to the unedited July 1973 MSG soundboard recordings that I'm using as my tone bogeys. Where I landed sounds extremely close to me, and doesn't require radical control settings:

Bass-spec with the following changes
-V1: Cathodes are split, so Rk was adjusted to the equivalent 1K5 versus the shared 820R. both fully bypassed and 3uF sounded good for Ck values. It currently has 3uF, but that's just what I had in last. I'm going to leave it that way for a few days to see how I like it, though I suspect I'll reinstall the 100uF that was there before.
-15nF coupling cap off V1 (should make a trivial difference versus 22nF; just went with this because I had a NOS Philips available in 15nF...)
-4n7 bright cap
-470k/500pF mixers (normal channel volume set to zero). The 270K's sounded good, but with 6550's it was a little too thin; I suspect this would sound very good with KT88's. More on this below.
-680nF on V2; Rk = 1K (didn't notice a major difference with 820R)
-Bass-spec tone stack; mid pots on the high side of tolerance helped
-Plate bypass caps on each side of PI (I didn't have 150pF on hand, so I went with 120pF on the inverting side and 100pF on the non-inverting side); also retained the 50pF snubber cap
-82K with 6550's
-Loosened filtering: 16uF on each preamp node; 64uF on PI (sane as MSG100), 40uF on screens and reservoir nodes (chosen to get 60V sag from my power supply)
-Dave Weyer-inspired increased screen grid resistors (2k5 on my amp vs his suggested 2k2)
-Doubt it impacts the tone much, if at all, but I also have a low-impedance bias circuit to keep the 6550's as happy as I can; they're also biased per the contemporary Marshall practice of biasing them as 25w tubes rather than the full 35w

I swapped the KT88's I had in before to some GE 6550A's, and I was surprised how different those sets sound. Some of it is just different sets sounding different (particularly since I'm comparing new production KT88's against NOS 6550's), but the feel also changed substantially. My experience matches others' observations that 6550's felt much firmer and clearer than KT88's. The difference wasn't stark enough that changing settings couldn't add back some fog and sag to compensate, though. The load lines look happier with the 6550s, so I'm keeping them installed.

[WhopperPlate]

increasing cap to .22 with 82k would get closer to the target frequency. 91k would be even closer.

Not sure about the bright cap in original, but I assume stock super bass value

power filtering I am almost certain is typical 50/50 cans throughout, maybe 32/32 in the pre, but might just be a 50/50 as well.

always felt above 1k on the screens was too compressed imo

also, one more thing...PI is 12au7...

[cdemike]

I might be mistaken, but I don't think most Super Basses came with bright caps from the factory (schematics only show the 500pf treble peaking cap over the mixer resistors). Agree about the filtering -- mine has the loosened reservoir and screens nodes to compensate for it being a 50w vs 100w amp, and I landed on those values based on testing based on a goal sag figure of ~60-80V. This amp sags typically 60-65V.

My testing mostly lines up with Miles' belief that Page's attack envelope suggests lower preamp filtering. That said, I did try a few different configurations. I consider the most likely scenario that which is most similar to stock or would have required the least work given how most modifications I'm aware of in amps from this time period being fairly conservative. AFAIK, artists started amp mods that radically reworked their amps en masse starting in the late 1970s/early 1980s. Swapping the same preamp filter can cap wouldn't be a ton of work, so swapping a 32uF x 32uF or 16uF x 16uF can for the preamp seems reasonably plausible. Alternatively, the simplest mod I could think of to get a lot more preamp sag would be to make it follow the JTM45 topology with a shared preamp filtering node. It'd be minimal work since it'd just be soldering in a jumper across the B+ dropping resistor separating the V1 and V2 nodes while also clipping the lead to either side of the 50uF x 50uF preamp can. This also sounded "right" in terms of approximating that rubbery attack envelope. I suppose you could just leave the preamp node as 100uF by just jumpering the 10K preamp dropping resistor and not clipping one section of that can cap, but I didn't try that and can't speak to how it sounds. 50uF with the normal JMP B+ rail topology sounded good, but it didn't do the rubbery attack envelope thing in my amp as well as lowering the preamp filtering.

I briefly A/B tested the 1K screens vs the 2K5 setup but I didn't test it too much since I decided I could live with a slightly more compressed amp if the trade off is that it'd be easier on these expensive big bottle tubes.

Not convinced about the 12AU7. I did a lot of tweaking with on-hand parts prior to picking up a 12AU7 that I'd potentially have sitting unused for years if it didn't sound right in this amp. To that end, these were the PI routes I explored:
- Substituting a 5751 (load lines show marginally increased bias voltage at similar operating conditions, which the tubes I have on hand at least also bore out);
- Substituting a 12BZ7 (same logic as 5751)
- Increased bias resistor value
- Reduced grid leak resistor values
- Substituting a 12AT7
- Reducing the plate resistors to get substituted 12AT7 closer to center bias (I went with 47K plate resistors per 70s Fender practice)

None of these sounded right to me, which leads me to think the hollow-sounding snarling overdrive character comes from beating up a 12AX7 in the phase inverter slot. Demos of the Sun Dragon amp seem to me to be the best evidence one way or another, but unfortunately I've only been able to find a handful of clips of it: https://www.facebook.com/watch/?v=9279278032109805; https://www.youtube.com/watch?v=M2T6t5nuE3M; https://www.youtube.com/watch?v=cvExfR6eNog; and https://on.soundcloud.com/z9J6zKRCJCl4Co89Yc

I looked high and low for other demos, as well as of clips of other amps using a 12AU7 as a LTPI but that's really all I came up with. All of those clips sound good to me, but to my ears those clips don't do the 1973 Page "thing," whereas I think the Royal MSG100, for example, sounds pretty similar, albeit too smooth, too compressed, and with a little bit too much low-mid emphasis (the lead-spec tone stack contributes at least in part to those attributes, IMO).

[cdemike]

I have a little time later this week and I remembered I have a spare ECC99 that I might try to give the general 12AU7 concept a try. ECC99 is different enough that I don’t think it’d be anything other than a feeler, but reviewing the load lines shows it should have a somewhat similar operating point with 47K plate resistors and a 1K5 bias resistor producing a voltage gain of approx 9. I had been considering using that tube as a 5687 equivalent in a West Coast style build but this seems like a better way to pilot a similar phase inverter than what I tried previously.

[cdemike]

Sorry to light up this thread with small updates. I had a chance to record some very brief demos with the ECC99 phase inverter as well as some with the 12AX7. These are quick and dirty; sorry in advance for the sloppy playing. Mic positioning was the same in each test, and I adjusted settings to compensate for the changes between the amps. Unsurprisingly, getting similar levels of drive required putting the volume near 8 with the ECC99, whereas the 12AX7 clips were recorded with the volume around 3. Listening back on these I'm realizing I probably have too much treble for the 12AX7 settings, especially since they're comparatively lacking snarl.

https://on.soundcloud.com/5kpqul7sXesYmrHKIF

[WhopperPlate]

fwiw 12au7 IS what is in the Super Dragon.

correct on bright cap, stock value = nothing, I just personally can't attest that is definitely the case in the super dragon.

one thing to remember as far as compression goes, an echoplex ep3 will change the character substantially. I myself never play a super lead without one. the JP amp without one is indeed pretty tight and stiff.

[cdemike]

My apologies if that first reply read as condescending regarding the bright cap. Thanks again for the intel -- definitely useful to have a sense of how the Super Dragon would play with that low-gain, high headroom PI, and I think I may have been talking out of turn if I tried the 12AU7 on its own without the other changes since based on what I'm finding I think a a plain Super Bass with a 12AU7 would be supremely unforgiving and stiff. With the changes like the V2 cathode bypass cap (and the other stuff on my amp that appears to be unshared by the Super Dragon like the loosened filtering), it felt stiff but still playable. Those recordings were with the bright cap, and the volume was at 5, so still a major part of the sound. Without the bright cap, the amp was way too wooly, and it didn't clean up very well with the volume set high enough to get the right levels.

Totally agree about the EP3 preamp. I played those with a Chase Tone Secret Preamp. I'm working on a DIY tape echo with a EP3 preamp, and the amp actually has an EP3 preamp built in that I didn't use in the clips since I received counterfeit JFETs from Amazon -- they're even labeled 2N5457 (I know this kind of thing has been documented before, but it still blows me away that something so niche and cheap would attract enough attention for this kind of scam to be a thing...). With this spec amp, it sounds more like the soundboard recordings set up like the early versions with the 20nF source bypass cap, IMO. I experimented briefly with hybrid 56K/500pF tonestack that sounded pretty good with the preamp set to the later spec without a bypass cap.

The broader signal chain for those recordings was LP > Secret Preamp > Amp > Weber MASS > G12H-55.

Something I noticed that I forgot to mention was that I had a pretty difficult time getting the ECC99 version to clean up appropriately. I originally planned to try to record Boogie Mama for these demos since the E played on the D string with Page's guitar in the middle position sounds so distinct, so I thought it might help shed some light onto what Page's setup was in 1973. I decided against it, though, since I couldn't get the amp to clean up enough, even with my guitar set to 2 or 3, and my guitar has 10% taper pots. That might point to the operating point being wrong for the PI stage, though, so maybe that's some insight into how they configured that phase inverter?

Edited: reading back on this I realized the part in red is erroneous despite actually building it this way; edited for clarity for any future readers.

[WhopperPlate]

no apologies necessary, I didn't detect any distasteful tone, my wording could have easily been interpreted as suggesting an alternate value bright cap. my bad.

I had the chase tone, tried my hardest to make it work, but I prefer the real deal by far. the old ones punch in the low mids.

as far as JP amp, its a stock Marshall LTPI configuration.

[cdemike]

I'm also less than happy with mine since I'm finding it contributes a lot of high frequency noise (could just be it picking up interference), though I'm not ready to point the finger at anyone/anything specific at this point since I haven't read any similar concerns elsewhere. In any event, I'd be curious if you or anyone else has experimented with other JFETs. Between using a similar JFET with a far-from "ideal" operating point and he the unique L-pad mixer/volume control it seems like it should do the EP3 thing right, so if it's not the JFET is the first place I'm looking. 2N5484 seems like it may be similar enough, and I also have been able to find TIS59's for sale, but I don't think I'm willing to buy dozens of expensive NOS TIS58's among the remaining NOS stock given how it's been out of production for so long, which I'm reading as making the remaining stock having a large share of out-of-tolerance rejects from when TIS58 were new-production.

Any chance you have any information as to the values in the PI? I drew up some load lines for straight-substituting a 12AU7 into the stock 82K/100K + 470R setup (assuming tail resistance remains 15K) which don't look great: bias voltage is around 2V, which is hardly an improvement over 12AX7 for half of the wave form. For reference, the green and blue load lines are the 47K/1K5 that would be common to the later-spec Major's driver.

12au7 ltpi loadline shared rt = 15k.png

[WhopperPlate]

10k tail . Stock Marshall values .

[cdemike]

Sundragon redid their site once they released the Nymph, and I'm noticing that the new site has manuals posted which are quite informative. Specifically of interest, it alludes to the amp having a bright cap: "Because of its unique circuit, the amp will become brighter as you lower the Volume control and 'fatter' as you turn up it up."

https://static1.squarespace.com/static/ ... ual_v2.pdf

[Mark]

Sundragon redid their site once they released the Nymph, and I'm noticing that the new site has manuals posted which are quite informative. Specifically of interest, it alludes to the amp having a bright cap: "Because of its unique circuit, the amp will become brighter as you lower the Volume control and 'fatter' as you turn up it up."

https://static1.squarespace.com/static/ ... ual_v2.pdf

Great link, I hadn’t heard of Amperian speakers before, they are made in the same location as Warehouse Speakers (or perhaps made by Warehouse Speakers?)
The speakers seem to exclusively use the 55hz cone.

https://amperian.com/pages/speakers

The link confirms the 12AU7 in the P.I. stage.
I thought Page used KT-88’s and not 6550’s.