I’m trying to start a build based on some of the more generic 300B designs (JE, JC Morrison, etc…) and I’ve run into my first major issue that’s had me stumped for a few days now.
I have a 400-0-400 -> 5U4GB -> CLCRC supply with a “slow, meandering” noise voltage at the input of C2 which makes its way all the way down. (about as far upstream as I want to stick my scope probe) Any idea what could cause this?
(for reference, carrier wave is the 120hz, vert scale 100mv)
I’ve tried different tubes, i’ve tried rearranging the LCR elements but i’m always getting it. The only constant has been the fact that the supply has always been tested with the “main load” being the two 300B power stages, albeit without the input stage, just the -70v on the grids via resistor to ground. (My thought is that i’ll commit to the input stage later since i’ve heard this is the biggest challenge with these amps. )
I haven’t been able to measure this voltage unloaded without the power tubes/opt… i know i’m leaving a lot of variables in the circuit but i’m not sure what to use as a dummy load for B+, I read some bad advice on this forum to use lightbulbs in series and ended up arcing over a perfectly good rectifier 😭 hard lesson that bulbs are like shorts when cold … i’m thinking of maybe cascoding a 6336 tube as a special supply testing unit
I have a 400-0-400 -> 5U4GB -> CLCRC supply with a “slow, meandering” noise voltage at the input of C2 which makes its way all the way down. (about as far upstream as I want to stick my scope probe) Any idea what could cause this?
(for reference, carrier wave is the 120hz, vert scale 100mv)
I’ve tried different tubes, i’ve tried rearranging the LCR elements but i’m always getting it. The only constant has been the fact that the supply has always been tested with the “main load” being the two 300B power stages, albeit without the input stage, just the -70v on the grids via resistor to ground. (My thought is that i’ll commit to the input stage later since i’ve heard this is the biggest challenge with these amps. )
I haven’t been able to measure this voltage unloaded without the power tubes/opt… i know i’m leaving a lot of variables in the circuit but i’m not sure what to use as a dummy load for B+, I read some bad advice on this forum to use lightbulbs in series and ended up arcing over a perfectly good rectifier 😭 hard lesson that bulbs are like shorts when cold … i’m thinking of maybe cascoding a 6336 tube as a special supply testing unit
Another view of the issue zoomed out over 7 seconds (500ms div), I don’t remember which point of the circuit I was probing, but it’s the same sort of issue just for illustration
Post the power supply circuit, because something is either unstable or noisy.
Agreed. Connect an unregulated supply to a slow scope trace and you’ll see this “wandering”
Yes. Let's see a schematic.
If you have no schematic, then tell us some more about your supply.
yeah it’s true it’s not that bad…. It’s not audible… i was planning on tapping off of c2 as a signal to cancel out some of the 120hz in the speakers (whatever the hum pot wasn’t able to take out) but that depended on a relatively stable B+, but then again….. i guess in a real world scenario, B+ would fluctuate with the music and it’s essentially be a some crappy form of NFB.
I’ve tried different rectifier tubes.
The PS resistors aren’t carbon comp, they are those aluminum chassis mounts… if you mean the bias resistor… i have no idea… it’s whatever i had lying around for testing purposes and it’s probably metal film or even worse, carbon film… lol
I’ve let the thing sit turned on for a couple of hours and it seems to have settled a tiny bit … they’re brand new m “golden voice” 300b’s that were only $100/pr… but some other 300Bs I have did the same thing.
I think i’ll just put more L in the supply, digi key has some 10H 90ma chokes for pretty cheap that i could grab for each channel instead of the 100Rs currently in place
I’ve tried different rectifier tubes.
The PS resistors aren’t carbon comp, they are those aluminum chassis mounts… if you mean the bias resistor… i have no idea… it’s whatever i had lying around for testing purposes and it’s probably metal film or even worse, carbon film… lol
I’ve let the thing sit turned on for a couple of hours and it seems to have settled a tiny bit … they’re brand new m “golden voice” 300b’s that were only $100/pr… but some other 300Bs I have did the same thing.
I think i’ll just put more L in the supply, digi key has some 10H 90ma chokes for pretty cheap that i could grab for each channel instead of the 100Rs currently in place
