New forum member - just finished first build and have weird power supply issue.

[Pistolboots]

First off - this is a great forum and I have been learning good stuff here for a few months. I use Altium software for a client of mine and started going down a rabbit hole of tube amplifier design, finally building my first amp and getting it completely wired up as of yesterday. My design is pulling heavily from JCM800land but adding a few extras to make it adjustable enough to hopefully be able to swap 6V6, 6L6 and EL84 tubes.

Before wasting anyone's time I still need to do my diligence in retracing everything (for a fourth time), measure voltages everywhere, etc., etc. However, the problem itself might be something you have all encountered already and some immediate suggestions might save me a bunch of time.

Good news: after lots of rechecking, the initial power up through a Variac and a light-bulb-in-series box went well. The transformer I am using is from an old oscilloscope and not ideal but I am currently getting over 500 volts at the first filter cap using a 4 diode bridge rectifier circuit. After making sure nothing was shorted, blowing fuses or sparks, I loaded up some crappy tubes to see if I could get audio and it passes sound without any squealing, hissing, hums, or alien noises. Gain, master vol, treble, mid, bass all seem to do what they do. Not sure yet about bias, adjustable neg feedback, or presence but all seems well.

Weird news: after power up I get sound for a few seconds and then the amp fades to silent. When I turn the power back off I get full sound again for a few seconds and then the amp fades to silent.

Again - I realize I have not even started my diligence in measuring, testing, rechecking. But, I can't be the first nut to build their own box and have the same weird result of losing power after a few seconds and getting power back for a few seconds after turning the amp off. I am hoping one of the pros here recognizes the symptom and can help me look in the right places so I don't have to do too much measuring with the amp on and possible find the mistake by setting something on fire. Or someone will recognize an issue in my schematic. Anyone else have this weird effect??? I am assuming I have an obvious issue with my power supply but there are some other potentials for mystery: I am using the 90v secondary for the reverse bias voltage. I did not have current between the 0v tap and amp ground so I have them tied together to complete the circuit for the 90v tap - or I do not get 90v. I had already used an ohmeter to make sure the 90/0 tap was not part of the B+ secondaries. I floated the Earth connection that was part of the B+ secondaries since I am using a full bridge rectifier to avoid shorting the transformer. But full disclosure - I do not have a valid schematic for the PT so I am guessing/hoping/measuring and trying to avoid buying a new PT since I am broke. Since I am at over 500v I am already thinking about going to a half bridge and dealing with having less than optimal voltage for prototyping vs probably having too much. I am also using 6L6's to power up and have not adjusted or measured bias voltage. Basically - after some initial testing I thought it was safe enough to power up slowly with some scrap tubes and see what happens. The weird fade out after a few seconds and the weird second wind after turning off the power switch makes me think I goofed on the power supply somewhere. Any help is appreciated. I also understand I will likely need to do a bunch of testing to have enough good info to get some educated help but I am hoping somebody here has had the same issue before and knows right where to look.

I am posting the schematic and will post some pics soon. I have the entire thing wired up point-to-point in the case from an old Leader Oscilloscope and am using the scopes PT for B+ and the heaters for the preamp tubes. The tubes light up behind the screen. I thought the case would be cool for prototyping test designs and with the cover off I can get to almost everything without having to dig. Super excited my first amp made actual noise on first turn up. Hopefully I can get it to actually work next.

Thanks for any responses. Great forum!!!

[DummyLoad]

hopefully the CT of the 198-0-198 winding you're feeding a full wave bridge is not connected to earth as stated in the schematic - i presume that's a typo, because if it was, there'd be smoke possibly with sparks.

you do state that you have over 500V at the first filter so assuming that the CT is floating - that's good and bad. the good, the FWB is wired correctly. the bad, the calculated B+ with 400VAC feed is about 565VDC, which correlates with the what you report for B+ at the 1st filter.

do you know the current ratings of the windings? it may be possible to strap the 90V secondary in series with 1/2 of the 396VCT secondary - that should supply about 290VAC which should yield around 400VDC+/-. the unused 20VAC winding could be wired as delon full wave doubler for the bias supply.


--pete 

[Pistolboots]

Thanks Pete,

Yes, the CT for the main B+ is floating and labeled as “Earth” - and labeled as E on the transformer. Testing with an ohmmeter showed what is labeled as “0” is on the same secondary coil as the 90v tap. I made sure there was zero current between the 0 tap and the main amplifier ground so I wasn’t shorting the transformer but without the 0 tap tied to ground at all I don’t get 90v for the bias voltage.

I do not know what the current/voltage rating is for the PT. I am using the secondary PT for heaters to reduce current and also keep the volume down. Main purpose is to use the case for prototyping. First design that I fall in love with likely go on turret board and into a head with an appropriate PT.

If I can safely run the separate secondary coils in series to get enough voltage I could switch back to half wave and pull the bias voltage off one side of the half wave bridge. Or maybe I should be happy with 400v? Would I still float the CT of the main B+ secondary if I strap the secondaries together?

FYI - there is also a 900v tap on that same PT for the oscilloscope screen.

[Pistolboots]

Hey folks,

I dropped the B+ down to about 200vdc and the weird fade to silent went away. The crazy fade after turn on and resurge after killing the power has disappeared. Everything else seems to be working aside from the B+ being too low for any magic to happen.

Thanks again for the help Pete. I may try running the 90v tap with half the main secondary like you suggested to get 300vdc. At least now I can start tweaking. Super excited to get my first amp working on Christmas.

[shooter]

the weird fade to silent went away.

sounds like a tube doesn't like the other volts
swap tubes, verify parts are right

[DummyLoad]

Hey folks,


Thanks again for the help Pete. I may try running the 90v tap with half the main secondary like you suggested to get 300vdc. At least now I can start tweaking. Super excited to get my first amp working on Christmas.

be careful, i asked for the current ratings for good reason: if that 90V winding is not rated at least at or near (preferably greater) to the 296VCT winding, you may fry it. i suspect it was used for a negative PS and those are typically low current.

if you don't know what the current ratings are on any of the windings of that PT, you're rolling the dice. losing at playing PT roulette can make some spectacular smoke!   


--pete