Amp Build From a Vintage Radio

[blown240]

I want you to make a few resistance checks with amp turned off. Clip one meter lead on chassis. Clip the other meter lead directly on the following tube pins. V2-8 and then V2-7.
I want one number for V2-8 and I want two numbers for V2-7 (one with treble pot max CW and another with treble pot max CCW.

Also, how is that ground buss connected to chassis? Is that chassis painted on the inside? If so, be sure to scrape off the paint at any point where you make a connection to the chassis. Don't try to solder to your chassis. Use screws, nuts, and star washers (or kep nuts) to make chassis connections.

And one final thing. Connect a jumper wire to chassis and connect the other end directly to V2-7. Be careful to get the right pin and only the right pin. Turn the amp on. Does V2-7 read zero volts now?

v7-p2  treble down = 995;  treble up = 1501

v2-p8 is 002 either way

Ground wire it bolted to the chassis at both ends, one end is a ring terminal, the other side is thru the ground block that is bolted to the chassis, no star washers. 

Chassis is bare clean metal at both places.

Jumper wire from p2-7 to chassis makes between 0 and .1 volts.  It makes the tubes arc though, and none of the knobs do anything.

[sluckey]

v7-p2  treble down = 995;  treble up = 1501

v2-p8 is 002 either way

I'll assume all three of those numbers need to be multiplied by 1000.

Jumper wire from p2-7 to chassis makes between 0 and .1 volts.  It makes the tubes arc though, and none of the knobs do anything.

Which tubes arc? All of them?

You have at least one more wiring error. If you can't find it, then consider ripping out all the wiring and start fresh. Or take it to someone who can find it. Your odd layout and schematic combined with all white wiring will make it a huge challenge for any of us to spot the problem via pictures. Someone needs to have the amp in hand to unravel this mystery.

Think about this... The first time you posted voltages, V2-7 was fine. Now it's not. What changed? Also, your bias supply is the only source of negative voltage in that amp. And V2 ain't supposed to be connected to the bias supply.

I've held off saying anything about your layout and wiring job because I didn't want to sound condescending or discourage you. But it's time to say it. Your wiring sucks and it's probably the root of most of your problems. This criticism is meant to help, not to hurt your feelings.

[blown240]

You didnt/wont hurt my feelings.  If I didnt want constructive criticism/help I wouldnt be here.  I have already considered starting fresh with my wiring. 

The thing that changed was when I found there was a jumper missing from pin3 to pin8 of the 12ax7 in the power stage.  Before that I had good voltages, but no output.

only the 2 power tubes arc.  I have considered pulling the 3rd 12ax7 and power tubes to see what my voltages are, then put them back in 1 by 1 to see when the problem surfaces.

Could a bad tube socket do this?  they are all used...

[John]

The thing that changed was when I found there was a jumper missing from pin3 to pin8 of the 12ax7 in the power stage.

When you connected 3 and 8 of the phase inverter (the 3rd 12ax7), perhaps that's where something went wrong? If you have a little strand of solder touching something it's not supposed to that can really mess you up. Someone else on here -I forget who- mentioned that a long time ago. And it can be only a hair width, really hard to see.

If you do re-do your wiring, try using at least 3 colors. You'd be surprised (at least, I was  ) at how easy it is to be dead sure you're tracing a certain wire and you're not.

[sluckey]

The thing that changed was when I found there was a jumper missing from pin3 to pin8 of the 12ax7 in the power stage.  Before that I had good voltages, but no output.

Actually, shortly after finding the missing jumper, you also found a missing 4.7K resistor in the cathode circuit of the PI. The PI voltages got happy after you put that resistor in, but about that same time V2-7 went to negative 37V.

A bad socket will not put a negative voltage on V2-7. Bad wiring will.

There cannot be negative 37 volts on V2-7 unless it's wired wrong. I can't see that wiring error without having my hands on the amp, but it's there. You just have to keep looking.

In one of your pics I see a three lug grounding block (like you'd see in a circuit breaker distribution panel). Looks like it is soldered to the chassis? If so, take it out and put those wires on a ring lug and bolt to the chassis. A small Weller iron is not capable of reliably soldering to the chassis.

One more experiment... dealing with the arcing output tubes. Set your meter to read voltage and connect your leads between pin 5 of one of the output tubes and chassis. You should read approx -40v. Leave your meter connected and monitoring the negative bias on pin 5. Now use that jumper clip lead to connect V2-7 to chassis. What happens to the meter reading?

[blown240]

Pn 5 on the output tube get me -34.  When I turn the volume and treble to get the tubes really screaming, pin 5 gets to about -37 to -38.    Doing the jumper from v2-p7 to ground keeps the pin5 at -34.

The3 lug grounding block wasnt soldered to the chassis.  It was bolted, but I took it out and put a ring terminal in its place.

[PRR]

> There cannot be negative 37 volts on V2-7 unless it's wired wrong

If the previous section is oscillating violently (and we have evidence of that), grid rectification will drive a grid negative.

I have not grokked the layout. High-gain amps (all guitar amps) are layout critical. Even a 2-toob Champ can oscillate; slight change of treble pot wiring can eliminate that. 

Find the scream.


Or remove/short preamp path to kill the scream while diagnosing power-stage voltages and currents (so the tubes don't die), then find the scream.

[blown240]

I may have just found it.  There was a broken pin inside the socket on v2-p6.  I am swapping in a new socket now and will report what happens...

[sluckey]

If the previous section is oscillating violently (and we have evidence of that), grid rectification will drive a grid negative.

What evidence? I'm not seeing that. The previous stage voltages look typical. In fact, the plate and cathode of this tube look normal even with -37Vdc on the grid. I would have thought the tube would be cut off with full B+ on the plate and zero on the cathode. I'd love to stand over this amp for a few minutes.

[blown240]

WE HAVE SOME PROGRESS!!!!!

The internally broken pin at V2-P6 was what was causing the -37 on pin 7.  I swapped in a new socket, and the voltage at pin 7 is now 0.

Tubes are still screaming at me though...

And just for the sake of testing, i plugged in a guitar and speaker.  The amp just screams feedback no matter where any of the pots are.

[PRR]

>> If the previous section is oscillating
> What evidence?

"Tubes are still screaming at me though... The amp just screams feedback no matter where any of the pots are."

[sluckey]

Guess the screaming had nothing to do with -37v on the grid. That negative voltage was the gremlin I was chasing. I expect there are still others about.

Pull the 2 preamps tubes leaving only the PI, power tubes, and rectifier. Does it still scream?

[blown240]

Pull the 2 preamps tubes leaving only the PI, power tubes, and rectifier. Does it still scream?

Ya it still does.  When I pull the PI tube the screaming all but goes away.  You have to really listen to hear it and that may be normal.

Now I did notice that one of the power tubes has alot more blue coloration in it.  This also may be normal, but I dont know, so I took some pics.

The tube closest to the camera has more blue in it, and was the one I really noticed was screaming. Would it hurt anything to pull one of the power tubes to try to isolate some problem, or is that a bad idea?   Or maybe I can swap the 2 power tubes to see if its a tube or amp problem.

[sluckey]

Do you still have the NFB loop disconnected? If not, disconnect it now. Does it still scream?

[blown240]

I do have it hooked up.  I will unhook it and report back in a couple days when I get home.  I had to go out of town for work.

Thanks again.

[sluckey]

Disconnecting the NFB loop is a method to check if the feedback phase is correct.

If the screaming stops when you disconnect the NFB loop the feedback phase is incorrect. The fix is to swap the OT plate leads going to pin 3 of each output tube. Then reconnect the NFB loop.

[blown240]

Ok cool.

[sluckey]

Any progress?

[blown240]

Not yet.  As soon as I got back from work. I left on vacation.  Won't be home till next Saturday.   But I will report progress a.s.a.p.

[blown240]

OK, Back from Vacation and time for an update:

Disconnecting the NFB helped alot with the squeeling.  So I swapped Pin 3 on the 2 power tubes.

Now with the speaker and guitar plugged in, and all the controls at full, you can get a halfway decent, fuzzy, tone.  Lots of hum though and no where near a good sound.  BUT if you turn the volume down a little it screams and makes all sorts of terrible sounds.  With the volume down, it sort of has a robotic sound, if that makes sense.  And if you hook up the NFB it screams again.

what do you guys think?

[blown240]

I did a little experiment this morning after tracing all the connections in the PI and power stage.  Everything was correct.

First I pulled the preamp tubes. Then I took my phone and clipped it in where the preamp section and the power stage come together.  This is with the NFB disconnected.





Here is the result:

http://youtu.be/38xrKoFLW00

So other than what sounds like a heater hum, I would say the power stage is fine.  Does anyone disagree?


Since I have a test method that seems to work, I came up with some other test points that I will try:

1.  Already tried.  Seems to point to the Preamp being the issue
2.  With the first preamp tube out, tie in right before the tone stack.
3.  With the second preamp tube out, jump from the first tube to the powerstage.  This will bypass the tone circuit and the 2nd tube.
4.  With the first preamp tube out, tie in right before or after the volume.



Thoughts?

[blown240]

Well thanks to all your help, this thing is FINALLY done!

I wanted to share a few pics of how it turned out.



Here is a shot of the wood with a flash, it catches the color pretty good:



The upper knob is on a piece of 1/4 inch dowel.  It comes out and the input jack is behind it.  Also I cut a piece of the original dial to put behind the original dial opening:





Here is a pic of the back:



The Speaker:



The Chassis:



Again, thanks to all your help and advice.  For my first build it sounds pretty good!

[John]

That is cool stuff, glad you stuck with it!

[blown240]

Thanks.  I am pretty stubborn with stuff like this when input my mind to it.  I do think I has a bad power tube.  I was playing it today and if I hit the strings hard I get a bad sound. It almost sounds me its clipping. I will try to post a sound clip tomorrow. 

[blown240]

Here is a sound clip.  You can hear it start to sound bad when I turn the volume up and hit the strings hard, that at about 15 seconds into the video.  The camera didnt really pick it up too good, but it get REAL loud when it starts to sound bad.

(pardon the crappy playing, I usually play bass)
http://youtu.be/ctrIhRG8HTg

I am thinking that maybe I need new power tubes, or that it is clipping, but before I go spend the $$$ for tubes I wanted to get your opinions.

[worth]

Saw your clip.. it just sounds like you have a bad solder joint ,( maybe where the heater wires are connected to the sockets). When you turn up the volume, the speaker vibration reveals the problem.

[blown240]

I looked at my solder joints and I dont see any bad ones right off.   

I was thinking it is bad power tubes because it get ALOT worse if I swap the 12AU7 PI tube for a 12AX7.  I was thinking the higher gain of the 12AX7 is pushing the tubes too hard, and the AU is just pushing them a little hard.

[blown240]

Here is a better sound clip.  Any ideas whats wrong here?

http://youtu.be/oEvYsksBF-o

[HotBluePlates]

Two questions:
1. Have you tried to reproduce the noise with a guitar?

2. Does the amp makes the same noise when a different speaker cabinet is plugged in (not using the speaker in the Philco cab)?

It seems from the pics that the speaker is a much bigger diameter than the cutout in the baffleboard. With a bass, you might get the cone pumping with loud, low peaks. I'm thinking maybe the speaker surround is buzzing against the baffleboard during those loud peaks.

I had some concerns about a similar potential problem on my current amp project, but experience from others says it doesn't happen much (if at all) with guitar signals.

If you can confirm the problem as being the items above (speaker buzz against the baffle due to large cone excursions on bass peaks), then you might want to install a shim between the speaker and baffle to get extra clearance. I'm thinking you could cut a wood ring a bit wider than the speaker gasket, drill holes for the mounting screws, and install between the baffle and the speaker. That would probably keep the speaker surround from buzzing against the baffle.

[blown240]

It does the exact same thing with guitar.  In fact it does it worse with the higher strings. 

Its a good thought, but there is already a 1/4 inch ring spacing the speaker from the grille cloth.

[blown240]

I noticed today that my treble knob is a little scratchy.  could there be DC on this pot that is causing all my problem?  or would that not give me sounds like this?