The Donut Princess Build

[mirage_indigo]

I know it looks like a rat's nest, but it actually went much better than I thought it would. I can see why people really like the turrets over the eyelets. I'll let it sit for a bit then go back and do the pre-tube wiring double check and trace. I'd already fired up the PS board, so I know that works (or worked anyway).

What worked well:
- Ampgarager's are the Best! Thanks especially to Steve Luckey and Rob Robinette (and everyone else I learned so much from).
- Turrets are the bomb. Never using eyelets again.
- Toroids are awesome (so far).
- The preamp tube in the middle of the circuit card worked smashingly well from a layout pov. Way fewer wires across the chassis.
- Regulated DC heaters means you can get away with a lot of slop in your heater routing.
- Planning planning planning! Invest the time to draw your layout fullscale with wiring up of everything.

Things to do differently:
- I populated the board then did the hookup wiring once it was installed. It worked pretty well, but next time I'm going to try it Hoffman method where we install the board then do hookup to pots and sockets then do components last. The basic reason is the topology of the amp is much less likely to change than individual component values.
- My layout still leaves too little space between the front pot array and the row of stuff along the front of the preamp board. I need to bring the main board towards the center of the amp another quarter inch. The good news is the front part of the board is almost all ground nodes, but you still don't want these double grounded to the chassis. I got it in there withot shorting but it's _really_ tight under the pot cases.
- I installed the main preamp 22uF over top of the other components as planned, but forgot to account for the height of the turrets themselves. You can see it tweaked a bit kattywumpus in the photo to fit down below the plane of the bottom of the chassis better. Every time I do one of these I'm making fewer basic mechanical errors in layout. By the time I work up to the SSS, I'll be golden.
- You just can't cram that many leads into a turret. Two tops. It does seem though that I can lay them out 3/8" apart (especially if they're the same node electrically) and I'm fine.

PXL_20201002_022044908.jpg

[Colossal]

Well done and a good build report. Turrets are great. For those junction points that require four components to meet, like with a long tailed pair PI, I often enlarge the hole of the central turret just slightly with a drill bit. Works well. You have to select the right bit, of course, and be careful.

[mirage_indigo]

That's a great idea, Redrilling the turret literally never occurred to me. I assume you just drill out the top half? I really coulda used that trick with the big hexfred diode leads.

[Colossal]

That's a great idea, Redrilling the turret literally never occurred to me. I assume you just drill out the top half? I really coulda used that trick with the big hexfred diode leads.

Yes, exactly. Just select a (good, sharp) bit slightly larger than the existing hole in the turret, but obviously, smaller than the outer diameter I bore it out a few millimeters deep; enough to hang parts in comfortably.

[mirage_indigo]

It's ALIVE!!!! (almost) worked right the first time I fired it up (although I'd already debugged the power supply board which most assuredly did not).

It sounds great, super punchy with the MV at max, almost too much. This is somewhat to be expected with the plastic filter caps, main cap at 47uF, SS rectifier, etc. With the bootstrap style pre-cathodyne MV at about two thirds, it's more mellow and sounds more compressed like what I'd expect is stock princeton. With the preamp wide open and using the MV as the main volume, it sounds HUGE through my 15" weber P15n. I was running with the preamp wide open and the MV as the only volume control and it was nice. Super super quiet even at peak everything. Initially the reverb didn't work but I'd inadvertently reversed the cables. Def not a lot of gain of tap, but overdrives nicely with Strat kinman regulars.

What worked:
- The Pre-cathodyne bootstrap MV works flawlessly with a470k grid stopper. And it adds a ton of utility to the circuit. Highly recommended.
- Reverb works great. Very nice sound, no noise at all, less splashy than my DR at max, quite a usable range with the 100kA mix pot. Could likely use a dwell control. I think the KOC self-split + hammond 125A may be my reverb driver of choice from here on out. It is kind of bright, might increase the 560pF coupling cap from the feed side of the 3.3M to get a bit more balanced sound. But it's quite nice the way it is.
- The capacitor coupled bias worked really well in conjunction with the bridge rectified mains. The 250K-B adjust pot seems to give a good range from 0 -> 50+mA per tube with 6V6s.
- My tone controls are stock Fender TMB, but I can see why a lot of folks just solder in a 6.8k mid and call it good. The mid interacts with the bass pretty substantially. The treble is unusably bright unless the preamp runs hot with the volume wide open. The MV really opens up a lot of territory in terms of trade off between bright/punchy and compressed.
- The trem most definitely works, although I'm currently getting _way_ too much trem signal to be useful beyond very minimum setting. I double checked my layout and schematic and I think this is because I swapped the stock 1M output oscillator coupling resistor for 250k. I'm going to swap it back and see if that helps tame it down a bit. I didn't end up adding the Mosfet driver, most assuredly don't need it for 6V6GTs. Also I mechanically installed the switch upside down, but I think I have enough slack in the wire I can rotate it.
- The toroid works great. Gets slightly warm, but not anywhere near hot even under max volume.
- It's light and compact! Whole amp + tank weighs maybe 20lbs? Need to build the head cab still. I think I can get the head in under 25lbs to match the external 15" sapele speaker cab.

Problems:
During the initial power on with the 25W bulb in place a couple things happened:
- The regulated heater circuit dropped out completely, down to 1.8 VDC. Which meant the tubes didn't really conduct at all. This meant the power supply was almost completely unloaded and I was getting crazy high DC values on the B+ node (500+) even with the series 25W bulb. Still not sure why, but it might have actually worked without the 25W limiter. I pulled the power tubes and it went up to 3V DC which still wont get the preamp turned on. The other main problem is that there was really no DC path to ground without the tubes conductive, so the B+ would stay hot for a long long time. Even with the power tubes pulled, I was still not getting enough voltage to light up the preamp heaters.

I ended up bypassing the regulator and bridge completely and installing a pair of 100ohms and just running the thing stock to see if it would work. Worked great, no noise even with my routing the heaters right underneath the main circuit card. I might futz around with that circuit later to figure out if I can get it working with a low-dropout regulator. AS it stands now, it looks like the regulated heater circuit in the schematic is operating too close the the margin to really work with a light bulb limiter. I don't know that it works _without_ the limiter but more research its required.

- Without the tubes, the lack of DC path to ground is bad and I need to put in a 100k resistor someplace to drain off B+ so it doesn't float off to the heavens. My caps are all rated 630, but still.

Bias/DC values:
Initially the amp sounded _terrible_ but then I remembered I'd set the bias at max cold. Once I got it biased up to 25mA per side things sounded so much better. With moderate bias, I'm getting about 380V on B+ (which is less than stock but about what I expected with a 320 v RMS output from the PT. Stock BFPR is 340.

The bias point of the amp also seems to affect the headroom of the preamp a great deal. When I biased it up to 35MA a side, I was getting only 140v on B4, which is really a huge drop over stock (I'm using 1K-15k-15k as the dropping resistors). When biased to 20mA a side I'm getting more like 210v B4 This may be from using only a 50VA PT, but I'm going to drop the 15k preamp dropping resistor down to 10k each and see if that gives me better headroom but still acceptable buzz.

Changes to make on this one:
Drop the 15ks to 10k in the power supply.
Change the 250k trem coupling resistor back to 1M.
Strap a 100k across B+ to ground someplace to give the power supply something to do when unloaded and a discharge path.
Flip the trem switch right side up.

Changes to make for next build:
Tweak the layout for better clearance around some of the caps.
Figure out a separate passive regulated heater for the preamp then run the power tubes off one winding and the preamps tube off of DC.
For two more lbs, upgrade to the 100VA transformer, mostly since you can get an actual Antek center-tapped secondary at 350-0-350, which means you could be much closer to stock voltages.

[mirage_indigo]

Update: I had the Intensity control wired backwards. It's pretty hairy up at the top end, so I'll still make the 250k -> 1M cap change. Should post some samples at some point this week.

[mirage_indigo]

After action report:
This thing sounds GREAT! I've been living with this as my only home amp for a bit and I love it. Matching head cab out of sapele photos over in the cab/speaker forum. Sounds just insanely great clean. I am playing it through a 1x15 cab with a weber Mass attenuator, loud, but there's more than enough bass there to be fun. So it's basically a loudish, clean Princeton without that much gain. The cleans really do sound great with a depth to them I don't hear on a typical princeton from the SS rectifier + over filtered power supply. Sounds more like a mini-JM wonderland without so much bass. Overdrives well, but def sounds like a Fender BF.

What's good:
- Very light weight. I think it came out at around ~18lbs for just the head + cab + full size reverb tank. Very portable.
- The KoC push-pull reverb driver sounds amazing with a full size 3 spring tank. Zero noise, which is in itself kind of shocking. This will be my reverb driver of choice from here on out. See above for the working schematic with a hammond 125A. I just bought a Jazzmaster, so there is a lot of reverb in my future. It's a rule, play a Jazzmaster, overuse reverb.
- The tremolo is great. I still have the Intensity pot wired backwards, but meh. The amp can get a little weird at max bias setting.
- I'm not sure I'd put the mid control in there again.
- The cathodyne "boostrap" master volume works well.
- Since the PT does not have a center tap, the bias supply is "capacitor coupled" this also worked great.

What could be better:
- It can get a little muddy when pushed hard, but still sounds great.
- I am getting some residual hum at 60hz which I suspect is a result of the last minute changeover from DC to AC heaters. I didn't plan for this. so was somewhat haphazard about the heater wire twisting. You can't hear it at all when you play, but it's there and seems to not respond to the MV, so it's likely in the power amp section heater wires.
- The DC regulated heater circuit utterly failed. I suspect I was too close to the voltage margin for the part and what I need is a "low dropout" regulator. I could never get this working and eventually just threw up my hands and went with a 100 ohm fake CT to ground. I thinkI might try to wedge in a regulator board (just bridge into CLC filter) in there somehow, but mechanically it's a tight squeeze. I used a similar regulator board in my last build and it worked well and was easy to debug.

[TUBEDUDE]

Before you try all that, try biasing up the heater string with ~ 30 volts or so D.C.
This eliminates pesky hum.

[mirage_indigo]

That's a good idea, I think I have a low impedance DC node around someplace. I suppose I can just string a voltage divider across the B+ node and grab 40V off the middle.

[mirage_indigo]

Now with samples!!!

A sampler of sounds from a 62 AVRI Jazzmaster: https://youtu.be/RWV6gxOqGqg
The sound of the KOC push/pull reverb driver in action: https://youtu.be/ww9vLuN_1ik

[TomAmp]

Hi

I’m really interested in this PP reverb! This is great work, congrats on the build! The videos are awesome

The PP reverb seems less drippy than a standard Fender. On the other hand it is much cleaner, isn’t it?

How does it compare with the modified Fender reverb that Dumble uses where he builds in localised negative feedback / negative feedback? I suspect it is somewhere in the middle?

How would one practically implement a dwell control on the PP reverb? Like a PPIMV after an LTPI? That may not be great for tone at lower send levels?

Lastly an explanation of the PP circuit would be brilliant if possible please. I’m not sure what the 1uF and 100nF + resistor networks do!

Thanks for the inputs please!

[TomAmp]

Not much luck let me be more specific

Do you think this simplified arrangement that I sketched would work just as well?

I can definitely see some similarities with the Hiwatt 70s PI (attached) which looks simpler.

Do you see any merit in adding NFB like the SSS 002 has? I suppose that's easy to test with a couple of alligator clips...

Any views appreciated please - thanks!!

[mirage_indigo]

Hey Tom, I just saw your question, I've been off building pedalboards.

Disclaimer: I lifted the design more or less stock from Kevin O'Connor The Utlimate Tone books as mentioned in the thread.

The Basic P-P driver reverb driver sounds less "splashy" than a stock fender, but it's stil very much in that territory. There is no NFB in the Kevin O'Conno design. The Dwell control would be a pot to ground in the place of the bottom 1M pot on the input side. It will absolutely look a lot like a "long tail pair" phase inverter since that's exactly what it is.

So it sounds good, but the main reasons you do it are:
a) it partially rejects power supply hum since it's push pull and
b) you can use a smaller reverb transformer since you don't have unbalanced DC through the core.
Basically the same as the reason you go to any push-pull amplifier vs single ended. The input signal is the difference between the two grid inputs, so the caps and resistors shape the rolloff of the reverb signal into the Reverb transformer.

The downside is it takes more parts, requires two preamp tube sides and it requires a center tapped transformer primary. But since these aren't that uncommon, we're fine.

The Hammond 125A works really well here since it's cheap, plentiful and small. Only 3W. Use lugs 2-4. I also had ten of them lying around.

I was able to mount all the parts (and the transformer for the next iteration) on a 5" x 3.5" daughter board that only had the connections: Power In from B+, Ground, Reverb In from preamp, Reverb Out to Tank. I put a hole in the middle and mount it around the tube in the chassis.

So it makes a nice, compact way to add a reverb output. You still need the extra preamp stage for recovery on the other side of the tank.

[mirage_indigo]

Also in your schematic, the signal into the left grid gets taken off the wiper of the dwell control.

As far as NFB, I didn't need it. Letting the reverb just kind of saturate and go non-linear is part of the fun assuming you don't get a lot of buzz.

Which Dumble schematic specifically?