Separate names with a comma.
Discussion in 'Shock Brother's DIY Amps' started by mrriggs, Jun 28, 2020.
This is a brilliant design. Thank you for sharing the results!
I neglected to mention the need for a series resistor. The 220k series resistor in the amp schematic is actually part of the Treble control, I simply moved it to the other side of the clipper so that it could do double-duty as a current limiter for the tube diodes. As a test, I shorted out this resistor and drove the clipper to full square-wave output. Sure enough, the tube diodes were overloaded and started oscillating.
It isn't necessary to put the resistor in the signal path. You can put it in the clipper leg so that it has no effect on the circuit with the distortion switched off.
On second thought... you probably do want the resistor in the signal path. A 2P2T switch could be used if you want it out of the circuit when the distortion is switched off.
Very interesting that you liked the symmetric soft clipped shape more than the asymmetric! Makes me want to build one and experiment. Good point on the low voltage supply. Could just use a wall wart supply low voltage B+ and not have to worry with batteries, right? Then you could adjust the asymmetry like your original.
If you're just messing around with it then a wall-wart would be fine. Heck, you could do the same adjustable bias Asym control with the AA batteries too. A 2M resistor across 3 Volts will only draw 1.5 microAmps. Those batteries will die of old age before you have a chance to drain them.
I made a couple more changes to the screen regulators to eek a bit more power out of it. Capacitors were added to stabilize the voltage going into the screen resistors. Also changed the voltage divider to bump up the maximum screen voltage from 130 to 170 Volts. Yeah, that exceeds the datasheet maximum of 150 Volts but I've heard that these sweep tubes are pretty forgiving. That brought my dissipation at idle from 70% to 85%.
Last week, before any changes to the screen regulators, this amp would put out 8 Watts of clean power. Now it's cranking out 16 Watts. Not bad for a pair of power tubes running at 200 Volts.
The kids' grandma was visiting this week so I wasn't able to crank this thing up, until today.
The changes to the power section have completely transformed this amp. The cleans are glorious, the cold-clipper sounds much tighter and the power amp distortion is to die for. What's really odd, though, is that the hard-clipper which sounded tight and articulate before, now sounds fizzy and weak. I'm going to call that three steps forward and one step back.
Oh, I double checked the datasheet for the 6GT5 tubes and it says the maximum screen voltage is 220 Volts, not 150 Volts. Watching the screens with a scope I can see that the load line is hitting "below the knee" and pulling massive screen current when the power tubes are pushed to distortion with the power knob on 10. This exceeds the screen dissipation, at least on the bench with a repetitive signal and dummy load. To determine whether this will be an issue in the real world with a guitar and speaker, I am employing C.A.D. (catastrophe aided design). The screens are rated 3.5 Watts so I hooked them up with 2 Watt flame-proof resistors. If the resistors ever burn up then I'll make revisions.
Couple more changes.
I never used the #2 input so it was rewired for a single input. This thing has some wicked hiss with the Gain and Volume dimed so I tried the Valve Wizard trick of a smaller (metal film) grid stopper with added capacitance at the grid. Hard to say, just working from memory, but it seems to be a bit better. Still plenty of hiss, though. No biggie. The hiss and mondo feedback (with semi-hollow guitar) are part of the experience. It's a quiet amp when not trying to be obnoxious.
The second jack was reconfigured as an input for the power amp. I previously built a Brownface pre-amp with Harmonic Tremolo which I've been plugging into the signal return on a solid-state Randall head. Now I can plug the BrownVibe into a real tube output.
When I said before that the hard-clipper sounded fizzy and weak, I musta' been in a weird mood. This has been my #1 amp for months and I never switch off the hard-clipper [in the 3 Volt P-P setting]. All you gotta do is turn up the master volume and turn down the gain to get big clean tone. Set the gain around 5 or 6 and you'll get a wide range of sounds with only the guitar volume. Very dynamic, better than the power stage distortion.
Speaking of the power stage... The novelty of the power-scaling output stage has worn off and I always have it cranked for max power.
The bass/mid (contour?) control is absolutely perfect. Crank it to the right for Fendery mid-scoop cleans, crank it to the left for smooth crunch. Sticking it straight-up makes my Danelectro U2 come alive. After giving away my 5F1-clone the U2 has sat unused because it just never sounded good through any other amp. Now I got my old friend back.
Another tweak to the circuit. Added 28 Volt 50 Watt Zener diodes across the cathode resistors on the power tubes.
The tubes bias about 27 Volts through the resistors with the screen voltage turned up all the way. These power tubes pull some serious current so the bias would substantially drift the harder you pushed it causing nasty crossover distortion. I wanted to keep the cathode bias because it helps balance mismatched tubes and the power-scaling feature wouldn't work without it. The 28 Volt Zeners were picked to kick in when the amp is pushed in full power mode. Crossover distortion is CURED.
Where do you get 50 Watt Zeners? The junk drawer, of course.
Power transistor (MJE15034) with resistor (100 Ohm) between base and emitter, Zener diode(s) (12V + 16V) with cathode to collector and anode to base. I wire-wrapped the components directly to the transistors. Bolted them to the chassis with plastic screws and mica insulators then wired them parallel to the power tube cathode resistors.
Neat trick. 50W might be overkill though.
No doubt. I just said 50 Watt because that is what the transistors are rated. Granted, that's assuming they are bolted to a more substantial heatsink than just the chassis.
I calculate 7.25 Watts for my amp with these current hungry sweep tubes. A standard 6V6 power stage could get by with less.
Could you provide a schematic showing how you actually implemented the "Zener" function? Sounds a bit more like a voltage regulator the way you described it. Once again, clever!
I am assuming it does not draw current until the voltage is exceeded, and the cathode resistors will still have current through them. I realize you are not saying they need to be 50W, just thought I would mention it in case someone reads it and thinks they need a 50W zener in the position. Not that I have a thing against a large safety factor. Actually all for it. Mind you I take equipment beyond they normal operating capacities. I am funny that way.
Correct. I picked a voltage above the idle voltage so the Zeners do nothing until the amp is pushed. My first attempt had the Zeners working at idle and it exacerbated the mismatched tubes (which is why I switched to separate cathode resistors in the first place).
For a regular amp without power-scaling and with matched tubes you could completely eliminate the cathode resistor and capacitor, making it a fixed-cathode-bias.
Ha ha, I totally forgot to account for the current still flowing through the resistor. In that case it'd be dissipating closer to 6.25 Watts if pushed to full square-wave output.
Well, for all intents and purposes, it IS a voltage regulator. The premise of a power-Zener is this...
You'd use something like this where you need greater power handling and/or tighter regulation than a lone Zener diode. I left out the diode between emitter and collector since it would never be forward biased in this application. You could even leave out the resistor between emitter and base. It's only there to tighten up the regulation.
Since this application doesn't require tight regulation you could use a voltage divider instead of a Zener to set the voltage.
Take it a step further and add a trim-pot to give yourself an adjustable fixed-cathode-bias.
On first glance I did not think of the resistor also, I was going to mention it as a 'poor man's fixed bias', not that there is much needed to make a fixed bias network. I like seeing people go off the beaten path in their amps.
Boo-Ya! Now we got this thing cooking.
The output transformer had a bit to much impedance for these tubes. The load line was hitting "below the knee" and pulling massive screen current which was really limiting how far it could be pushed. Simply hooking the 8 Ohm speaker to the 16 Ohm tap would have taken it too far the other way and pushed it into the red. I've been running two 8 Ohm speakers [in series] parallel to a 16 Ohm speaker for an 8 Ohm load. Swapping out the 16 Ohm speaker for another 8 Ohm speaker gave me a 5.4 Ohm load (I measured the actual impedance of the cab). Drawing it out on a load line shows it hitting in the middle of the knee.
I put together a 5.4 Ohm dummy load and fired up the scope. It is no longer pulling massive screen current. Even when pushed to full square-wave output, it won't exceed the rated screen dissipation. Max power output was up to 18 Watts.
The load line said it should be higher so I dug through the tube box and found a pair of 6GT5s that not only run stronger but are dead-nuts evenly balanced. I did have to tweak my "Zeners" to 31 Volts to allow the cathode resistors to properly set the bias.
The end result of all this... Over 21 Watts of clean output! That is from a pair of power tubes running at 205 Volts. I ran it at full output for a half hour straight, in a dark room, and there was no sign of red-plating. Got it all done about the time the kids were going to bed so I could only play through it a little bit. Golly-gee it sure sounds good. It's also nice to have the piece of mind that I can crank it and not worry about melting the screens.