AB165 Bassman Refurbishment & Modification

Okay, this is a friend's Bassman head which was nearly unusable - lots of ripple/hum and thanks to the death cap it would shock him occasionally when he completed a ground circuit by touching a microphone. Death cap indeed! He gave me carte blanche to modify it as I saw fit, stating that he was only interested in a great clean sound. I knew he was considering buying a Twin Reverb, so after I got the amp fixed up I resurrected the Blackface character of this amp (more like the AA864 Bassman) and he was thrilled with the result. He gave me permission to post the "in progress" photos I took on the condition that I include my original captions.

What follows is the email I sent, in its entirety.

--------------
Hey, here's what's gone on so far.
* Took out 'death cap'
* 3 prong plug (will have to redo with thinner gauge cord)
* polarized accessory outlet flipped & rewired for correct polarity
* fuse moved to hot wire
* new 2A slo-blo fuse (I think the fuse in there was 15A)
* new pilot light bulb
* Changed the filter caps
* Changed the bias cap
* Changed the cathode bypass caps
* Took out FUCKING STUPID hi-cut caps that were on both channels
* Started testing coupling caps & replacing leaky ones - only 1 cap so far!

1. yep.

2. blurry shot of chassis.

3. & 4. they printed up a shitload of these and just kept using 'em. This is actually an AB165.

5. The only other marking I see inside the cabinet.

6. gutshot - that green wire is the new ground wire.

7. guts

8. guts - those cardboard tubes are the cathode bypass caps, and that red circular ceramic capacitor (just right of center... like I like my democrats) was bad as well.

9. they put a brass grounding strip behind all the pots. It would make more sense if it only contacted the chassis at one point, but whatever. Probably made assembly easier. these are the normal instrument controls.

10. bass instrument controls.

11. output, standby, power, fuse.

12. power, fuse, ground (does nothing), accessory outlet, power cord - doesn't fit the strain relief, so either I ream it and keep this thick-ass cord or hunt for a thinner one.

13. what the power switches looked like stock. The death cap is hiding.

14. why have I not been capitalizing consistently? anyway, yeah, this here looks fucking terrible and stupid to me, but it worked fine for 50 years so who's stupid now?

15. the bias board. 6 rectifier diodes, a bias cap and a bias diode hiding somewhat behind the cap.

16. the dog house. who's in there? is it stu?

17. nope, it's some big-ass filter caps. Notice the one on the left has reversed polarity? I did. It's supposed to. They put two of those big bastards in series, which gives you half the capacitance but double the voltage tolerance. So if you somehow generate 700V these would be like "cool, here's 35 uF."

18. you know, if you're advertising your electrolytic capacitors as being the fancy new "dry" ones, you know the ones that don't contain acid sloshing around.... damn you old. Also that 16 uF is on the schematic and layout as 8 uF, but apparently 16uF was always used.

19. underside. for my own reference. and my hand.

20. ooh, cauliflower!

21. NOT AS PRETTY, I KNOW. The black one on the end? Yeah, reverse polarity. These are also 100 uF (giving you 50 uF instead of the original 35 uF) but this is better, doesn't strain the PT or diodes, gives you better filtering, more punch, and seriously no one makes 70 uF caps. Electrolytics are +/- 20% tolerance so it probably doesn't matter at all. SUCK IT.

22. Hey, that looks the same!

23. New bias cap. So much tinier. This eyelet board is kind of a pain to solder to / desolder from, so shit ain't going to be perfectly symmetrical all the time. Eat a dick.

24. New cathode bypass caps.

25. New cathode bypass caps part two: Son of Cathode Bypass Caps. AAARGH SHITTY SOLDERING AT THE BOTTOM DON'T LOOK FUCK

26. New cathode bypass caps part three: The Return of the Son of Cathode Bypass Caps' Ghost - in 3-D!

27. Holy shit did they gob the solder on putting this bitch together. The two ....solders... on the left were done by me, the two on the right by some anonymous Fender/CBS employee who hated lead and wanted to use it all so there wouldn't be any left for anyone else. Ditto their use of solder flux, leaving nice wet-looking halos around each eyelet.

28. New coupling cap! But what value is it?! (0.1uF, 630V because I'm hardcore like that)

YEP. Not done yet, and I haven't found the bad cap in the normal instrument section, but the bass instrument section is fucking tits.

STILL TO DO:
* put in a proper bias control instead of the "bias balance" it has now
* proper blackface-era negative feedback
* take out redundant feedback caps around power tubes
* not sure if I'll do this, but depending on how it sounds I might change the phase inverter & power tube coupling caps to the typical blackface values. Normally it's a small coupling cap to the phase inverter which cuts the bass there and then large ones to the power tubes which lets the bass through later. As it is, the values here are backwards - letting the bass into the PI and cutting it before it gets to the power tubes. This might make it distort sooner. So I don't know. If I wanted to be a jackass I could just be all "...the power amp of a Twin!" but I'm not. Get off my lawn.


Told u I was hardcore.
--------------

I fixed the things I mentioned in my "still to do", lowered the coupling cap to the PI and left the output coupling caps alone, and had to replace a pot or two and it was done. I don't think I took any more pictures though.

The AB165 is an interesting schematic - when you keep in mind Leo's plan was always "as loud as possible, never driven to distortion" then a lot of the seemingly-silly mods make sense. A global NFB loop on the previous Blackface Bassman wasn't "enough" NFB apparently so there were additional feedback caps from the plates to the grid. They kind of killed the "sparkle" and made the distortion tone worse.

While I'm talking about the NFB in this amp, on the AB165 they actually return it to the grid of the first LTP triode instead of the usual - attaching it to the tail. I converted it to the usual, which required switching the feedback resistors to 820R/100R and flipping the leads on the primary.

The AB165 also adds a virtual-earth mixer stage, which is a clever way to get separated outputs and I figured the local NFB around that tube was part of the amp's character so I left it alone.

The original "bias adjust" pot is a good idea if you're in the late 1960s and factories are still cranking out reliable tubes. Current production tubes are less spectacular - you can order them matched (on plate current draw I imagine, not transconductance though that would be a treat) but you can't assume anymore that just plunking in a new set of 6L6GCs (or some Russian military equivalent sold as a 6L6/5881-WTF) will be biased in the safe operating range.

Plus, "biasing an amp" has taken on its own cult status as THE ultimate operational factor of a guitar amp. If you hear a difference between 60% and 65% idle plate dissipation, great, more power to ya. The cynic in me suspects that the easier something is for guitarists to change themselves, the greater the perceived tonal reward - especially if there's no math involved. Brands of tubes, brands of capacitors... whatever floats your boat.

On the other hand, I have a sizeable stash of NOS tubes that I'll defend to the death, so I'm as full of it as anyone else.

This amp was a ton of fun and I owe my friend a debt of gratitude for trusting me to work on it and giving me permission to post these photos.

On the bench

It looks like it's going to take a while for me to post details of everything I'm working on, but here's a quick list:

• Stromberg-Carlson AU-36B (100W beast)
• Bogen CHB-35A - 30 W of glory, if I can ever settle on a schematic. Planning on making this my main gigging amp.
• Crate GT50 Stealth - Combo with 4x6V6 output section and a fried PT. Might try to avoid point-to-point for this one. Probably going to be a classic Marshall, like a 4010 or 4001.This one I'd likely be willing to sell.

I've also been looking at modding a Silverface Deluxe Reverb for a customer. Been pretty interesting so far pushing the schematic around; really excited to start tweaking the bugger. 

And if $600 falls in my lap from the heavens I'm thinking about throwing together a 5E3 just for fun, maybe flip it for a tiny profit that I'd roll into future amp parts. 

Stromberg-Carlson AU-36B

This is a beast of an amp that my dad resurrected in the mid 90s. ~100W from four 6L6GCs, two 5U4G rectifiers, 6SJ7 pentode input, 6SC7 floating paraphase phase inverter, and a 6N7 driver stage for each phase after the PI.

As you can imagine from the size of the transformers, this beast is quite heavy.

Guts (so far.)

The original schematic for an AU-36 with AU-36B corrections.

The most recent revision I've typed up.

So! A whole lot of potential here, but a decent number of restrictions too, many of which self-imposed.

• I don't want to drill any holes in (the front of) the chassis. I will drill two more holes in the rear of the chassis so I can use 4/8/16 ohm cabinets. Right now it's only set up for 4 ohms, which is all I need, but eventually I think the versatility will come in handy. I could just drill one hole for an impedance selector switch, but switches are easy to bump and I don't like putting heavy currents across them if I can avoid it. 
•  This leaves me with only 4 potential knobs: bass, treble, and two gains.I tried a Baxandall stack with the two knobs on opposite ends of the chassis, but the only place in the circuit that I can put the tone stack is right after the 6SJ7 and the output impedance of that pentode is just way too high. Instead I've decided to adopt the FAC switch from Orange for a bass cut. It works really well, but the values here are still too large so I'll probably lower C4. Even though I used a shorting switch it still pops, so I'll have to put a small capacitor across it. 
• For the treble knob I was originally thinking about using a Vox style "cut" control, but I recently saw a vintage Gibson schematic of a floating paraphase inverter with a single passive treble shunt to ground on the first triode of the PI, as it feeds the second triode (and thus both phases of the output tubes.) Pretty clever! That'll also save me the trouble of putting in a series of coupling caps.
• I'm not going to bother with the master volume control - it's making way too much gain, but I think I'll try turning the 6N7 stages into anode followers instead. No idea if that'll cause problems with the global NFB loop, but there isn't a whole ton of NFB anyway so I might as well scrap it. Hey, this amp's from the 50's, it might as well sound like it! 
• The morph control is fantastic and everyone should have one. You can design the pentode to have a low screen voltage leading to an easily overdriven stage that adds crunch and a little compression, and then dial it back to a more typical triode for a more classic sound. I took that from Merlin's preamp book. 

Currently the amp's having distortion troubles - I think a resistor (or twenty) is failing intermittently and I've got to probe and scope around to find it/them. Almost all the resistors are original and apart from whichever one is failing the amp sounds like frying bacon. So yeah, vintage carbon comps are not great. The humdinger on the heater works exceptionally well though, and the overall hum and noise is manageable.

I freakin' love this amp, but I'm not going to gig with it or sell it so it's somewhat of a lower priority than the Bogen conversion which I'll mention in another post.

A Blog so I Stop Bothering People on Facebook

So! This seems like an idea. I've been obsessing over tube amp design for quite some time now, so I thought something like this would be a better idea than constantly shellacking slightly revised schematics to my facebook wall. I'm also branching out into repairing and modifying amps, so I'll keep this updated with whatever's currently on my bench.