Separate names with a comma.
Discussion in 'Glowing Bottle Tube Amp Forum' started by PMELIUS, Aug 7, 2017.
All of my resistance checks from the geofex page check out. Everything measures fine.
Sheesh. Nothing about this darn amp is simple.
The flash when you disconnect suggests there are no shorted windings, that's good.
I think in the RG Keen article, and then also in my setup, we are talking about applying a short burst of voltage and letting the magnetic induction in the core boost the voltage. My tester is a normally open push button...so the period of applied voltage is very brief.
Perhaps the behavior you are noting is valid, since you are leaving the battery connected for some period of time.
(I'm leaning towards your OT being good).
I suppose I should get my setup and try applying 9V to the secondary of an OT.
You've got that right!
I didn't have a normally open momentary switch handy (thought I did but couldn't find it.) I just used a 4pdt and only turned it on for a couple seconds at a time.
I'm going to conduct the test again tonight just to double check, but yeah...when I applied voltage to the secondary I got an initial (pretty bright) flash and then a very faint flash upon removing voltage. When voltage was applied to the primary, I got nothing but the very faint flash when voltage removed.
All of my resistance checks seemed good - 35ish ohms across either side of the primary. 1 ohm across the secondary. My economy DMM couldn't read the value from primary to secondary or from any lead to ground.
I'm gonna use the neon tester on another OT tonight and see what result I get.
let's assume your OT is 50:1 or even 25:1, I believe hooking the 9V up to the battery would step up the voltage quite a bit. An Ne needs 62V (I think to flash). So you are probably getting that.
I think your OT is good.
related reading here: http://www.antiqueradios.com/forums/viewtopic.php?f=13&t=168025
I took a couple videos just to be thorough. First is 9v on secondary. Second is 9v on primary. I'll take your word for it though...I was almost relieved at the thought of the OT being bad just to have finally found the real issue! Where can I go from here?!
I modified my tester to work like the one on this page ( http://www.ozvalveamps.org/repairs/trannytest.htm ) found through your link.
Got a flash when removing voltage on every winding. Slightly duller on the secondary, but ultimately I believe you are right about the OT. I think I'm going to get into modifying values beginning with the coupling caps from the PI
Here's a new development. I was checking voltages after changing my PI coupling caps to .1uf
I caught some AC (3v) on pin 3 of my output tubes and thought maybe that was peculiar. I figure maybe it's just hum as I have no input, so I crank the controls and go to check again.
See video for result. Output starts to make this weird low frequency oscillation with responsive blue glow similar to but ultimately different than my existing condition.
Does this help us pin point any others areas of exploration?
Edit: this is induced by the pop created from my probe touching the pins of the output tubes, but is also generated by any other signal input
Changed the bias resistor (2.7k > 1.5k) on V1A to bring down the plate voltage from nearly 300 to 277 and the cathode from ~2.5 to ~2. Still no where close to the 240 the schem calls for, but just another thing that was bothering me I figured I would try.
Also lifted the 25/25 caps on V1B and V2A as per your earlier recommendation with no change.
Reading back through this thread...Tonight I'm going to start digging into the changes you had suggested Commodore64. I see now what you were saying about that 220 > 470 between V1B and V2A, gonna try that first as well as trying to reconcile my PI voltages to match the schem
I think your pulsing is relatively normal. The crappy sounding part, isn't. But the lack of loud popping noises suggests to me that there isn't arcing going on.
That said, nearby computers, fluorescent lights etc can all cause oddities when they are nearby, or even on the same circuit as your wall plug. You've tried different outlets/locations right?
You've focused recently on changing PI coupling caps and biasing of your preamp tubes. Those are areas for tweaking, not for troubleshooting at this point. Those are unlikely to cause the issues you are seeing.
The 220k/470k voltage divider between plates...I've never seen that. I don't even know what purpose it would serve (connecting the plates on 2 preamp stages like that). The first 220k, I get. They are attenuating signal since there's sure to be plenty of gain in this circuit.
Do this: clip out the 470k. Just clip it out. See how that sounds before you do anything else.
Just tried it. Really big increase in gain. Used to be the amp was only clean up to about 3 where my symptoms would begin accompanied by distortion. Now it just starts dirty right out of the gate. Still have that blue in the power tubes.
I also removed everything from the outlet I have this on, with everything else presumably on that circuit shut off/disconnected. Nothing else on nearby. No change in symptom but I can try a different part of my house tonight. Previously, this has been on the same circuit as my fridge so I had a flash of embarrassment for a minute when you said that haha
This thing almost sounds like it has a fuzz pedal in front of it with just a single coil straight in. It's crazy
So that 470k/220k is some kind of negative feedback scheme, I'm guessing.
Yes, there is too much gain in that circuit. 3 gain stages, similarly biased just sounds like crap to me. I know EXACTLY what you are talking about, having done a similar circuit in my Bassman 10. You've got a ton of gain, with the tone stack right up front. Ken Fischer made this work, but his last gain stage was a biased cold with no bypass cap. And it was also fed by a tiny coupling cap (.001uF): http://www.blueguitar.org/new/schem/trainwreck/wreckxpr.pdf
I put a similar circuit on a scope and it was square wave city. I posted the 'scope images in my Bassman 10 thread. I could NOT get a good sound out of that circuit. Check it, those scope pics were with 12AY7 in all stages (so lower gain than 12AX7): http://www.tdpri.com/threads/ul-bassman-mods.644789/page-6#post-7611925
So as long as we're just spitballin'...
Clip the cathode resistors on both your preamp stages. See how it behaves. BTW, I did all that on my circuit, still didn't like it.
That thread is crazy! I see a lot of similarities to my issue and some notable differences as well in your amps' topology.
I was trying to adjust my PI grids to match the schem by changing the tail resistor from the advertised 22k to ~50k, and while I got my grid voltage near the schem, it drives my plate over 310v and my cathode over 120v at idle. Just another thing to add. Probably going to take it back to the 22k.
If I lift all of my cathode resistors, won't that make the amp not pass signal?
Whoops, yeah. I meant cathode resistor bypass caps!
Definitely a reduction in gain, clean up to about 4 now instead of 3. Not a huge difference. Made the amp sound pretty flat as well.
Can I just bypass the third gain stage and go straight to my PI from the plate of V1B via the .1uf coupling cap?
Gonna try making the plate load resistor on the first half of the PI to hopefully balance the gain there, seems out of line with other LTP layouts
I've really learned a lot in this thread and with this project in whole but it's driving me CRAZY
Hmm...changing that plate load resistor on the upper (first) half of my PI seemed to stabilize that unpleasant distortion to a useable degree.
Maybe my output section was just being driven with unbalanced signals, causing the unpleasant distortion? Not sure if this would be considered crossover distortion...
Still have the cathode bypass caps lifted as well as the tail resistor on the PI at 50k and I re-installed that 470k resistor between V1b and V2a that we had previously removed. I didn't check voltages much other than biasing it, only listened.
I'm gonna close it up and take it to my practice space tomorrow and see how it performs at this point.
I was gonna recommend a 1.5k partially bypassed (4.7-10uF) gain stage 1, then a 10k cathode resistor (no bypass) on gain stage 2, then a 820 ohm cathode resistor on gain stage 3 (no bypass).
You can run a 400Hz 100ma sine wave into the amp and measure the AC signal at the power tube grids using a multimeter to see how balanced the phase inverter is.