Raoul Duke wrote: ↑Wed Mar 25, 2026 3:37 pm
Temped in an 82k and got -51.6 straight into the wall at 122vAC (no current limiter). I read that -30 to -50 is where I should be, and I think with a 10k pot wired “fail safe” and 27k resistor to ground I’m close?
I think so.
Raoul Duke wrote: ↑Wed Mar 25, 2026 3:37 pm
I know “center” on an adjustment pot is a moving target and I’ve been reading up on bias circuits because I never get them right the first time - but the more I read, the more turned around I get. Maybe bias is just “get close enough” and a certain amount of trial and error is the fine tuning?
I would agree with that. There is enough variation in the components that at least one iteration is required.
Understood, thanks Martin. I always feel like I could get closer the first time if I understand it better, but I forget that there’s just a lot of variables that can’t be quantified until everything’s together and the tubes are in.
Got just about everything ready to do the complete power amp test but wanted to get some opinions on PI lead dress.
In this case, the V3b plate lead falls on the far side of the OT impedance taps, so I picked it up and flew it over that bundle.
The V3a plate lead has to cross the NFB lead, so as I saw in actual 183 gut shots - I pushed the NFB lead under the board then cranked it over to the taps where it follows them into the switch at the 8R tap as customary for my 50w builds.
With the V3a plate lead, I flew it out and over the cathode lead then down to the pin.
I kept the grids as short and direct as possible.
Though it’s hard to see - in 3D everything here has at least 1/4” separation from everything else around it. I’d be satisfied with this in any other circuit I’ve had experience with, but I’m always just a little more apprehensive with “D” circuits only because there’s a lot going on in them - so just wanted a second (or as many) opinion just in case I’m missing something.
IMG_0531.jpeg
Thanks!
You do not have the required permissions to view the files attached to this post.
You have the right idea. Crossing leads at 90 degrees is best to minimize coupling so you might go more in that direction at the expense of some separation.
Thanks Martin, the V3a plate crosses the cathode at pretty close to 90 degrees, just about a 1/4” above it - it’s hard to tell from the photo angle. Would pushing it down and making it a hard 90 be better or is this a “try it and have a chopstick ready” situation?
Leads that are up in the air have smaller capacitance to chassis (ground) and so might pass stray signals on to the tube. Grids are most sensitive, of course. You can try it and see what happens. There's a pretty good chance it's fine.
Yesterday we got some decent weather, so I started on the cabinet for this amp - then built the foot switch.
Today I had some time to start testing the complete power amp, so I ran through the “no tubes” part:
* limiter flashes as expected
* heater voltage good
* bias pot sweeps -42 to -54
* B+3 on PI plates: 466
* B+1 and power tube plates: 478
* bleeder s work as they should: down to single digits in <3 minutes
Put tubes in-
* got the start flash
* got heater glow
* got standby flash, but the bulb started to increase in brightness after a few seconds; not crazy, but noticeably brighter than start up. I should mention that there was no red plating or burning smells, as well as no weird noises.
Shut down and checked everything over. Nothing seems amiss, try again - same result. Of note - these are JJ E34L tubes.
Switch for a known good pair of GT EL-34M, go through start up, standby flash increases slightly as tubes start to conduct, warm it up, measure everything, set bias (at apx. 3/4+ pot travel), monitor, everything is good.
PI plates are about 8v difference with trimmer centered using a Tube Depot certified balanced triode JJ ECC83S.
Two things:
1. Should I let the JJs warm up more/was I being paranoid because of the glow?
2. To get the bias pot more toward the center - I would change the 27k tail to a smaller value, correct? Unless the answer to #1 is “yes” let them warm up and the glow will subside, then re-bias and see where I’m at.
My understanding of those E34Ls is that they withstand higher screen voltage; is more current draw part of that? Obviously the amp is healthy to this point, I ran it for 20 minutes or so - just curious about the tube difference.
WWYGD?
Last edited by Raoul Duke on Fri Mar 27, 2026 9:39 pm, edited 1 time in total.
It's time to put the light bulb limiter under the bench and plug the power cord directly into an AC outlet. Set the bias pot for most negative voltage on pin 5 of both output tubes. Now set the meter to measure DC millivolts and connect to pin 8 of a power tube. Watch for any excessive voltage.
Understood Steve. Follow that process and try the JJs again. Worst case scenario is that they will start to red plate - or am I risking other damage as well?
I’d prefer to use the JJs if possible - had the GTs for something else; but if the JJs are a risk I’m ok with the GTs.
Also, answered my own question about bias range. Decided to change the dropper to bring the pot closer to center. All good there.
Ok, did as Steve suggested: put the JJs in, made sure the bias pot was at full negative voltage, speaker load connected, DMM set to mV and connected to the bias sensing resistor on the highest drawing power tube. As soon as I threw the standby switch, red plate on V5 with mV reading in the 400s; V4 was pulling apx. 8.5mV.
Seems to indicate a bad tube in V5 socket. Just to be sure, remove both tubes and verify that the negative bias voltage on pin 5 is the same for both output tube sockets.
Thanks Steve! Done as suggested and both tube sockets are getting the same voltage. Put the GTs back in and everything balances and biases as it should. In the home stretch now - a few more board connections and two more heater runs and I think it’ll be ready to fire up as a complete unit.
Got everything connected and decided to light her up.
Went through the limiter first: no shorts, everything stable.
Took the limiter off, everything fired up as it should. Here’s the stabilized voltages after about 20-30 minute warm up:
AC at the wall: 122
Heaters: 6.6
B+
1: 485
2: 483
3: 455
4: 335
5: 327
Quiet as can be. Only problem I have is that I wired the rear panel OD and PAB switches upside down and the foot switch is reversed; I make them with OD on the left and PAB on the right - this one is reversed even though I wired it exactly like my last two.
For the switches - I mistakenly wired them like 102 slide switches then tried to correct it by looking at pictures of my 102. I guess I’ll take them apart and re-wire tomorrow. Here’s what I did:
IMG_0357.jpeg
And here’s the whole thing:
IMG_0359.jpeg
I can’t be too unhappy with this result, lol. Just these small adjustments and I can start breaking it in and finish setting up the FET and PI.
You do not have the required permissions to view the files attached to this post.
Thanks ange!
Haven’t really been able to fully explore and understand her yet - but I can say that initially she does not sound like the other three ODSs I’ve built, which was what I was hoping for. I see great potential.
The switch thing was bugging me so much that I just had to fix it tonight. Happy to report everything is working as it should now after I re-wired the switches per the 183 schematic and layout. Fixing the toggles fixed the foot switch as well. Had a hunch it might.
Tomorrow I’ll start tinkering with the FET and PI balance - unless the weather is good - in which case I’ll likely work on the cab more. My exciting life, lol.
