Real TrainWreck mojo!

[big_teee]

So how about this layout?
We need something we can all look at and analyze.
http://site.triodestore.com/TWEXPRESSLAYOUT.pdf
How come no 68k grid blocker on the input?
If 10 people built this circuit, it would be like 10 guys building a marshall.
They would probaby all end up tweaking the same circuit a little differently.
The Classic tone PT has the 300-0-300 and a lower tap for the brown sound.
It looks like it would be a fun build and should fit on a 17 inch marshall chassis.
T

[Chuck H]

That's pretty close. The first thing I notice is the lack of power tube grid stoppers (which should be right on the tubes and floating). Trying to look up actual TW chassis shots now turned out to be an exasperating exercise in driving down the information super highway. I've taken to calling it digging through "the information pile" since there is low priority on organization and a high degree of infiltration even with specific subject matter. I do have a six year old CD around somewhere with a lot of chassis shots from a couple of old TW's. I'll try to find it in the interest of accuracy.

[daz]

I noticed something strange in that triode layout. Is there some reason why on the front 3 filter caps the negatives are tied together while the read 3 have 2 tied together and one has it's own seperate ground wire. Seems like whatever the reason it's gotta be nit picking to the nth degree. Guess those are the kind of eccentric things he did, eh? Probably had his reasons but i can't imaging that making any difference. It's not like it's grounded to a different place then the other 5.

[NateS]

While I don't doubt he kept some of the secret sauce secret, Ken would talk freely if you got his ear. I doubt that there is any willful obfuscation going on in the wrecks. I've also never seen lead dress cause anything but problems when it's not done properly. When it's right, it just works. I suppose you could introduce feedback with a carefully routed wire.

As for the choke vs BFR, the typical unclamped choke introduces some ringing in B+ visible at startup, which the BFR doesn't do. I'm not certain it's impact at audio frequencies under load, but 5H/50u is upper infrasonic and it likely has some impact on dynamic response. My instinct suggests the transformers are not what they seem.

[jazbo8]

I noticed something strange in that triode layout. Is there some reason why on the front 3 filter caps the negatives are tied together while the read 3 have 2 tied together and one has it's own seperate ground wire. Seems like whatever the reason it's gotta be nit picking to the nth degree. Guess those are the kind of eccentric things he did, eh? Probably had his reasons but i can't imaging that making any difference. It's not like it's grounded to a different place then the other 5.

No, grounding makes a significant difference and it is a common practice to keep the high current ones separate from the low current signal ground. It does not guarantee good sound, but it is highly recommended. As for the TW schematic, I think many builders are working off the one found here.

Found a clone demo using many of the original parts that KF used, but it sounds nothing like the GK's videos, well the playing style is quite different, but still... I hear very little bloom or swelling in the clone.

[Chuck H]

+1 on the filter caps. Word to the wise daz! The most significant effect is lower noise. Other things that the filter grounding arrangement can affect would include stability and even safety. Always separate your low current filters from your high current filters. In fact it's best to ground filters at the same ground point where the circuit they're filtering is grounded. There are plenty of amps around that do ignore this in the interest of simplicity for design or assembly, but trust us on this.

Jazbo8, I can't remember my Amp Garage log in info. I'm not inclined to start a new account since I hate doubling up on that sort of thing. I hope to remember at some point. In the meantime... Can you post a link that doesn't require membership for those of us that are either forgetful clowns (like me) or aren't subscribed to that forum.?.

[jazbo8]

Here ya go...

https://dl.dropboxusercontent.com/u/...0%28RLW%29.pdf

[Dave H]

I noticed something strange in that triode layout. Is there some reason why on the front 3 filter caps the negatives are tied together while the read 3 have 2 tied together and one has it's own seperate ground wire. Seems like whatever the reason it's gotta be nit picking to the nth degree. Guess those are the kind of eccentric things he did, eh? Probably had his reasons but i can't imaging that making any difference. It's not like it's grounded to a different place then the other 5.

Do any of the TWs or clones have a problem with buzz? I hope the ‘mojo’ isn’t in the ground wiring because that layout looks like a prime example of how not to do ground wiring in an amplifier. For example (if I’m reading it correctly) the CT of the B+ winding is connected to the chassis at the bottom right near the mains input but the first filter cap -ve is connected to chassis just below the filter caps which makes the rectifier current pulses travel through the chassis to get from transformer CT to first filter cap -ve. The pre-amp is also connected to the chassis many times so how can you be sure that the current pulses won’t get to the pre-amp? I know you wouldn’t want to change the layout too much but if there’s a buzz problem I’d at least run the B+ winding CT wire straight to the first filter cap -ve terminal to keep the current pulses out of the chassis.

[daz]

Well, not much you can do when you're using a 20/20/20/40uf cap can with a single ground for them all. But my amp isn't noisy. Less than many marshalls i've owned. I have thought many times about using separate caps and actually did buy some F&T's a while back. But somehow one was bad. I rechecked my wiring a million times and could not see anything i did wrong, but i never got to see how it sounded/felt compared to the JJ can. I'm a big believer that the apparently subtle things that filters can change are not as subtle as they first may seem. I should have built this last board with F&T one the main board at each node as i did when i first built this design.

[Chuck H]

The TW amps aren't unusually noisy, but probably more noisy than they need to be. Some builders report noise gremlins that they need to troubleshoot. There's a thread at AG where a nominal level of noise for this circuit is derived. So now cloners can determine if their build is more noisy than it should be!?! It's certainly not the worst grounding error ever devised by a builder. But I agree with you.

[big_teee]

On marshall type builds I do the improved grounding, and it makes a big difference on hum.
For example, no grounds on the rear of the Pots.
I see many things I would change on the grounding, and heater wire routing, but then again that may compromise the sound.
Also another thing that always bothers the hell out of me, is that unused gain stage.
I would have to hook that baby up!
T

[Chuck H]

So now we have a basic schematic (with noted corrections) and a layout. I'll try to find some actual amp porn to post just to be thorough. I've never been able to locate anything revealing via these documents, but maybe someone else will.

As far as I can determine the "secret sauce" would have to be in finite, within tolerance component swapping, lead dress manipulation and tube substitutions. Which would be a real PITA to get a handle on. This process probably wasn't spread throughout the circuit. There must be key areas that have the most profound affect. So I guess the quest, as it stands right now, is to locate what parts of the circuit would effect the bloom of the note and make the swirl of the good touch wha type and not fizzy. The note attack crunch is probably most affected by these changes in the preamp and sounds more "right" when the other things jive too.

Bloom is a function of sag. The only sagging thing about the amp is the BAR feeding the screens, the big-ish resistor feeding the PI (and subsequently the rest of the preamp) and the fact that the power tubes are usually clipping hard. So this may just be a matter of biasing by ear. Note the low-ish plate volts for this circuit. It seems likely that this is to allow for more current allowing the PS rail resistance to impose sag.

Swirl, as I've always known it, is a function of crossover distortion. Which can range in effect from wha like to slightly annoying to a buzzy mess. I've noticed that the degree of mismatch on the power tubes can greatly affect the amount and nature of crossover distortion. And, of course the bias can also. After that you need speakers that don't accentuate the worst sounding elements of crossover distortion. The OT affects this too but since that is a known quantity for this circuit we'll take it for granted. Ken was pretty specific about the speakers that should be used with these amps. A 4x12 with greenbacks.

So the right degree of power tube current, the amount (probably a small amount) of mismatch on the power tubes and the right speakers are all tied together in the bloom/swirl game. There may be some manipulation of lead dress to affect phase relationships in the PI leads? Any other ideas?

[big_teee]

Too many issues with the amp, for me to build one?
The grounding.
No fusing on the HT B+
No master volume
Don't understand the purpose of the 6 diodes on the Output tubes?
What tubes are used, 6V6, or EL34.
Don't understand the purpose of the staggered tube layout.

[Chuck H]

I'm with you! I've never built one because I would probably roll it differently based on my principals and then I wouldn't have an accurate point of reference... So it wouldn't matter. But the tone!!! I've looked at the layout and schem for longer than would seem sane trying to find genius and premeditated intention. There is some, but it may only be wishful thinking on my part. If looking at the design allowed me to think these were better than the amps "I" build it would be easier for me to "clone". But then there's the tone!!! Is it only a happy accident? Is it possible that KF just had the right sense of perspective and sensibilities to voice ANY amp? Even this one?!? That may be what it comes down to. I do know that Ken kept his personal Express as a reference point. He called it "Reality Check". So it could be that this is a slightly flawed circuit that just happens to sound amazing when voiced correctly. It wouldn't be the only amp like this. My guess is that he built the amp, it sounded good, he voiced it to maximum potential and then never veered far from the original design in the interest of tonal integrity. Many of us have built amps that had less than ideal things going on somewhere. Maybe Ken's just sounded so good that he decided to roll with it rather than confound a good thing.?.

[big_teee]

This sounds pretty decent, but to dark and syrupy at times for my taste.
The drawing says .022uf between the PI and the Output tubes, but sounds more like .1ufs.
I wonder what speakers in the marshall cab?