Two-channel > Harmonic Tremolo > Reverb > Low Watt Amp Build

[Snfoilhat]

@BenTobith thank you. I started using gimp to highlight or cut and paste copies of old schematics years ago to ask questions on TDPRI, and I never found a real schematic program I liked more. So I have the hand-drawn symbols from a clear copy of a 1960s Fender schematic and move them around the page for each new circuit I draw.

This is nearly everything. There are all the jumpers to add, and a few components associated with the reverb that will have to go on a terminal strip near the tube socket. Second thoughts have been brewing about not ordering the doghouse for this chassis

Working on the layout and seeing which parts I actually have changes the schematic, as everything iteratively inches along.

second draft, untried

 

[Snfoilhat]

After re-checking the layout and schematic, I trimmed the excess leads and began soldering.

I don't think I'll need access to the bottom of the board anymore, and so continue with the board in place.

Bonus content: I found ~13 ft of 11 1/4" x 3/4" pine in decent shape and started planning a Bandmaster or Vibro-King kind of cabinet for the summer.

 

[NTC]

Bonus content: I found ~13 ft of 11 1/4" x 3/4" pine in decent shape and started planning a Bandmaster or Vibro-King kind of cabinet for the summer.

3 x 10"? Yer killin' me. I have a weakness for 3 x 10's. If I ever get that far, I would make a 5G7. Your build in this thread is halfway there...

 

[Snfoilhat]

This kind of view reminds me of this:

From the Indianapolis Children's Museum

 

[King Fan]

This kind of view reminds me of this:

Nah, yours is way prettier...

 

[Snfoilhat]

Here are the very first sounds this thing made. After confirming that it didn't have any dangerous faults, I set my phone on the speaker cab (you'll see) and took this clip. There are some clear mistakes, some strange behaviors to hunt down including trying to figure out the hacked together trem intensity situation. Anyhow, here's a bit of payoff for everyone's attention an patience.

I'll document the troubleshooting stuff too, putting on my technician hat and treating this thing like an unfamiliar, poorly documented amp with multiple issues dumped in my lap by some rando

 

[Snfoilhat]

Initial feedback:

1.High-frequency oscillation associated with the lead dress, especially the output transformer primary (surprise!)

2. Gain and loss across the stages wants some experimentation. The breakup with pick attack is excellent, but the total volume seems wrong and the oscillation needs to be fixed before any of this is really relevant.

3. I don't want to get wrapped around the axle trying to make the 6 watt EL95 work where a 12 or 14 watt tube like the EL84 or 6V6GT would be perfectly happy. I don't think that difference matters much, and it's not like I'm sitting on a pile of valuable NOS EL95s. I think I'll give them another try in a small push-pull amp where I can get enough total current demand to make a PT happy, but this PT, despite it's low nominal voltage (250-0-250 @ 100mA), is putting up anode voltages way above the limits for this power tube. Why fight it?

4. I don't want to lose sight of my happiness that this pile of parts actually runs and the trem is 90% there on the first swing (thanks everyone in the encyclopedic Harmonic Tremolo thread). Mine does sound a little like a washing machine running in the next room ... ...

draft 3

 

[Snfoilhat]

Draft 4: Changed V7 from 6DL5 to 6V6GT and rebiased, which bumped up the total volume and sorted the power supply issues with the extra current draw. Fixed the output transformer lead dress and eliminated the major source of oscillation. The amp has more headroom now, with overdrive beginning to come on about 9-10 on each channel's volume control.

Changed the voltage dividers at each channel (Fender used a range of values in brown amps) to 100k:33k, up from 100k:12k. I could try going even higher here, but there are a couple other lossy spots.

One thing I'm looking at is that I left the cathode bypass cap for the last gain stage out of my original layout, and didn't leave a space for it on the board. It's crowded in that area. When I added a cathode bypass cap later on the amp immediately started oscillating again, so the extra gain, the poor lead dress, or the cap itself kind of flapping in the breeze rather than tucked neatly down is a problem I haven't sorted out.

There is a second, much quieter source of high-frequency noise associated with the reverb recovery, and I haven't yet addressed that. Last, at the extreme ends of the tremolo's sweep with the intensity all the way up, there is a quiet but noticeable zapping kind of sound, which I think is more of this high-frequency noise. No pops or knocks or thumps. I did bump the 4.7M (6G5-A Pro amp) resistor in the tremolo intensity to 8.2M (6G4-A Super amp), and maybe lost a little of the extreme end of the intensity sweep and also lost some noise. Installing the chassis in the cab didn't help, so it's probably not just the wifi or phones. More chopsticking to do.

Just gotta squeeze a little more gain out of it and reduce the noise.

draft 4

 

[jmp81sc]

This is a great looking build. Thanks for posting all of your information and fine tuning details. I did a deep dive into the
Fender harmonic trem and I got a lot of information from one of your older build threads.

 

[Snfoilhat]

I'm stumped.

There is a ton of noise in the reverb recovery. Reverb all the way down there's noticeable high frequency noise. It disappears with the reverb recovery tube removed. Turning the reverb up doesn't change the quality much, just makes the noise louder. I removed the OT from the chassis and jumpered it in from outside the chassis just to tease apart whether this was power amp oscillation or the power supply feeding the OT, but it didn't change anything. So I'm 90% confident it's isolated to the reverb recovery.

I shielded both grid leads, including the leads to and from the reverb control, and didn't hear a difference. There's enough slack in the cathode and plate leads on both sides of that tube to move each a fair length, and moving them doesn't change the noise.

My best clue, if it is a clue, is that the ground noise (reverb 0, ch 1 volume 0, channel 2 vol 0) didn't always have the same high frequency type noise earlier in the build. Across this whole build and revision process, the total noise floor has gone down, and I'm hopeful that if I sort the reverb the amp will really be excellent, but if a memory of a week ago is worth anything, there used to be less noise in the ground. Still, I don't know why a ground lead dress change is contributing to the noise I'm hearing. Somehow acting as an antenna? Placing some of the reverb components on the board and some of them on terminal strips beside the RCA jacks has led to some difficult choices for where to ground everything. That seems like a probable spot to look but I can't get the noise to change as I chopstick everything I can think of.

If you have any ideas I'd love to hear them -- thanks!

Here's the schematic view

 

[mountainhick]

Noisy tube? Can you sub in others?

Bad cap?

Bad resistor?

Bad solder joint?

 

[Snfoilhat]

Another 12AX7 in the same slot and the noise remains the same.
A 12DW7, which by luck puts the 12AU7-like 20 mu side in the reverb recovery while keeping the 12AX7-like 100 mu side on the final gain stage results in less noise, but doesn't eliminate it. Also makes the reverb mix really low even at 10, as you would guess.
Pulling the reverb driver tube results in the noise remaining the same.

Thanks!

 

[Snfoilhat]

I added the cathode bypass cap to the final gain stage and improved the OT lead dress again, routing the primaries away from either side of the reverb recovery + final gain stage tube socket.
I fixed a careless mistake in the James stack channel (it needed a dedicated coupling cap to keep DC off the Bass control).
I bumped one of the 10 nF capacitors in the oscillator loop up to 20 nF to slow the tremolo.
I added small caps to augment the miller capacitance of the reverb recovery + final gain stage triodes for more high frequency filtering.

I tried in this clip to show relative volumes of the noise at Reverb 1 and Reverb 10 compared to unamplified strumming and gently picking with the Volume at 10. I imagine the phone is doing some levels adjustment behind the scenes, so of course the subjective volume in the room isn't perfectly captured here. With the volume cranked the amp is reasonably loud in the room.

Though you may notice that if this is channel volume at 10, I still have some work to do with the gain structure to reduce the headroom. I'd like to sort the reverb out first.




I think the trem plus reverb at 7+ sound is just great, so I'm really hoping we can figure this out. Thanks!

 

[Snfoilhat]

Well, there still hasn't been a eureka moment. I've been reading about the reverb recovery issues that cropped up with the Fender Prosonic. The description sounds like what I've been facing with this amp, so I looked at the Tech Note meant to fix it. I wonder why the Prosonic, unlike any Fender I can think of, has a two-stage reverb recovery. Anyway the idea, if I'm reading correctly, was that it needed more separation from other circuits on the power supply. I couldn't follow the same exact change, of course (different recovery circuit) but I added another branch, a dropping resistor and filter cap, feeding just the reverb recovery and final gain stage.

I now have very little reverb noise bleeding into the rest of the amp when the reverb is a minimum, though if the noise at reverb 10 was reduced it wasn't by much.

Until I figure it out, I've gone back to playing with the 12DW7, which brings the noise down to a level I think I can live with even with the reverb at 10 (which it needs to be). Not spashy, but with some echo I prefer to hear when playing at home. Even my really cool tremolo amps get pushed aside for practice in favor of anything with reverb, so it was a must for this amp.

 

[FenderLover]

Keep at it. You'll get it. I'm happy and amazed the Vibe circuit is the only one appearing to behave!
Edit: I still want to do a straight-up 6G2 with the vibe circuit in place of tremolo.

 

[Snfoilhat]

@FenderLover I think that would be rad. You wouldn't have to bias it cool to ensure the trem worked -- a really unconstrained preamp compared to the big 6GX amps into a really cooking, smaller power amp. I didn't use your draft harmonic trem values from the other thread exactly, but I think it's clear that your arrangement of the oscillator straight into the rest of the harmonic trem can work great.

***

Still working on the reverb circuit. I want to make a demo of the overdrive sounds if I can find a moment when I can get loud in the apartment and leave this thread with a demo that doesn't sound like a broken amplifier.

Here's a teaser of the 3 x 10" Bandmaster-shaped thing coming next.

 

[Snfoilhat]

Switcheroo!

Now that I've built two amps with Bassman chassis, it's opened up the possibility of swapping cabs. Here's a sort of EL84-powered Tremolux 6G9-alike that I think works well in the new cab.

Here's the Tremoluxish's original back panels, which match the new cab better than they did the original! Here also is the Eminence GA-SC64 (12") I received when I had ordered a GA10-SC64. Instead of returning it, inspiration struck and I thought it would pair with this amp. The new harmonic trem amp is in the older pine cab and seems perfectly happy there.

Here's the thing the 10" Alessandro was supposed to go with:

It's shaping up to be a sort of 'super' AC4. I'll make a build thread down the road.

Here's a gross hifi mono amp I hope will transform into a sort of Ampeg B-15-alike. Or turn into anything to justify the stupid price I impulsively paid for it. Conversions of untested junk aren't something I've been trying lately.

Came with a full complement of old tubes which I haven't tested in any way yet. Fingers crossed.

And of course I will still make those recordings of the (mostly) fixed harmonic trem amp. I just took the oral examination to complete my Master of Science, and I'm looking forward to some music and relaxation this summer. Thanks for your time and attention, all!

 

[BigDaddy23]

As per usual, your cabinets are bloody fantastic. Looking fwd to seeing what comes of that rust bucket!

 

[marshman]

Nice work. I was curious...my personal experience with an Epiphone Valve Jr @ 5watts through the stock 8" speaker is still plenty loud enough to disrupt the notion of a conversation, I was wondering how that would pan out with a 12" speaker, it looked like your original power tube was going to make @ 3watts. Now, however, I'm nearly certain that a 6V6GT through a 12" speaker would be loud. Am I right?

 

[jumpbluesdude]

I built a Revibe that I modified to get more range out of the trem. It turned out great and I ended up selling it to a guy in Alabama earlier this year. Ironically, the Tremolux I built had a trem with one stage in it that was every bit as good as the revibe trem that had 5 and was the harmonic trem everyone raves on. Much respect for venturing out with your own circuit design. It's a challenge to chase out all the gremlins. I have a Princeton Reverb chassis that came with bassman transformers and 5 tube sockets that I put an original circuit with enough gain for Marshall territory but backs off into Fender cleans. I still haven't chased everything out of it. Anyway, here's some pics of the Tremolux and what was left of the wood I started with.