Tube Amp Problem [SOLVED]

[isophase]

What may have happen, is that while (difficultly) plugging the tubes in, a pcb trace may have broke under one of the tube socket... If it's the case, I'm afraid you will have to disassemble the whole thing.

good point
zayance on est avec toi ;-)
this amp will soon be fixed

 

[zayance]

Thanks man, well will do my best to make it work, don't think it's a big deal, but i have to organise myself and make it clean so... , best way to learn, thanks a lot for you guys here in Prodigy, you're great!
Anyway tomorrow the amp is on my desk and i'll start from the beginning and make sure to spot the prob before going any further...

 

[zayance]

Ok here are the pictures of the Amp guts.
So i need to get back to the simple testings and discharge before wanting to do anything on it, actually i have a nice tool that i bought in Germany, i remember somebody here posting a link to what it looks like, anyway nothing fancy just a circuit breaker that you plug your gear into, with a test and reset button, gives me double safety  .

 

[johnR]

That RCD (Residual Current Device) won't protect you against shocks from the amplifier power supply. It will only trip when the mains live and neutral currents become unbalanced (eg. when someone receives a shock directly from mains live). You still need to be careful when working on the amplifier!

 

[zayance]

Yes of course, i will, but for now, just simple testings to try to spot the problem, with the advices i had from you guys, before going any further.

 

[mad.ax]

Hummm... Looks like probing with a Oscope with the tubes plugged in won't be easy...

We already know that the plates connections are good since you measured HV on those. Another test you could do is to Ohm out the cathodes when the amp is cold (unplugged and discharged). This would tell you if you had a broken connection on the cathode...

Axel

 

[zayance]

Thanks a lot for your help mad.ax

In order to confirm that V3 is indeed the driver, you can plug it out and see if the 'return loop jack trick' still sounds well... If it doesn't and plugging it back make the sound appears, then V3 is Ok and we can concentrate on V1 and V2 (and whatever else there may be between them and the send loop jack...)

No sound when V3 out.

Did you try  to check what comes out of the send jack?

No sound from Send out, but buzzes when jack inserted in and out, just for info.

Hummm... Looks like probing with a Oscope with the tubes plugged in won't be easy...

Yes  

Another test you could do is to Ohm out the cathodes when the amp is cold (unplugged and discharged). This would tell you if you had a broken connection on the cathode...

Ok so first thing to do, is power off, take all tubes out and with discharging tool go from Plate to plate, and with DMM constantly wired on that res., to discharge all HT, just writing down so too make sure i got it right, even if i got it  , correct me if i'm wrong.
Now when discharged, take my DMM, put it in Ohm mode and if there is high resistance then it's ok, but if there is no resistance then something wrong  ???, again correct me if i'm wrong.
Thanks a lot, waiting for your answer.

 

[mad.ax]

Cathode resistors are usually in the 1 or 2K range... We are not looking for the exact value, just to make sure that the cathode is connected. If your ohm-meter says : infinite resistance between ground and cathode, then the connection must be broken...
Go on..

Axel

 

[zayance]

Ok i think i've discharged correctly, i could see some mV lowering on each plate, bigger mV on the EL84, i stayed a little on each one when moving to the next.

Ok here are the resistance of each cathode:

     Cathode 1           Cathode 2
V1  2.7K                    3.3K
V2  2.7K                    1K
V3  48K                     48K   :

 

[mad.ax]

Ok. Plates are connected, cathodes are connected, and filaments are connected, or the tubes wouldn't glow...
That's leaves the grids to test... Even with the amp unpowered, I guess that if you inject a signal in the input jack you should see it on one of the grid of V1 with the scope...

If that's the case, plug V1 back, pull out V2, Switch the amp on, and probe  V2 grids for a signal.... You may have to play with volume control and footswitch to make a signal appears on both grid....

Axel

 

[zayance]

Allright life was on the way but i'm back to business.
So i injected a 1Khz signal on input an i have signal at Grid 1 V1 Pin7, nothing on Grid 2.
So if i power up with V1 installed, leaving V2 out, do i have to put back V3 and all the EL84?? I don't think so, but just asking.Thanks for the clarification.

 

[mad.ax]

No need to plug back V3 nor the EL84s, cause obviously those are working...

Axel

 

[zayance]

Ok, i have sine at Grid 2 of V2 but not at Grid 1 of V2, playing with gain 1 makes sine move, Gain 2 when footswitch engaged makes it move also, but only on Grid 2.

 

[mad.ax]

So, We know that V1a and V2b works. there's a reasonable chance that V1b is the second stage and works too.
We now need to be sure about V2a status.

Do you have a non-polarized cap, something like 470nf/400V that you could use to inject signal?

My idea is to plug back V3 and the EL84s, and to inject 1KHz sinewave at V2a plate, or cathode, to see if the sound comes out of the amp...
If you have a dummy load that's perfect. If not, turn down the volume on your signal generator before injecting, to avoid being blasted with a full overclipped 15W!

Axel

 

[zayance]

Well, i have 100nf 250V, voltage too small i believe....
I've injected unbalanced sine wav at -40dB for doing test btw, from my little signal generator, just for info>
And sorry but what do you mean by Dummy Load, like a crappy speaker  ???

Thanks a lot Man!! you rock.

PS: BTW i have acces to Grid 1 of V2 when tube in, the sockets are installed in different positions, and i just checked, it's possible to check with probe sine at Pin7, so maybe don't need to inject???

 

[mad.ax]

Since you measured 310V unloaded on the plates, 250V is indeed a bit short...

Unless you have 2 of them in serie...

A dummy load is just a high power 8 Ohms resistor that one hook in place of the speaker. It allows to test amps at their full rated power without starting WW3 with your neighbors...

What we want to know is if the pcb trace is indeed connected to pin7 of the socket. If you have access to it, then you can ohm out the connection. But before, plug back V2 and probe this point with the scope.

Axel

 

[zayance]

Oh ok.
Well there is signal on V2a, so....., and signal on V3b

PS: and there is signal on tip of the send jack, well seems i've checked when jack was not on input but on return loop  :.
But i don't get it, i even checked the amp again for sound before doing any other tests.....

 

[mad.ax]

Do you mean on grids only, or also on plates?

Axel

 

[zayance]

Signal on Grid1 of V2, and Grid2 of V3, sorry.

I could check this:
Signal on Plate 2 of V2, on Plate 1 of V2 also,.
Signal on Plate 2 of V1, can't reach Plate 1 of V1 tough, well i could actually, if i take out the metal support around the socket, the one to keep the tube in place, but i have to grab the standoff, so it doesn't get loose from down, for now, same on the other sockets, could make it easier to probe.

If i do i'll make sure to discharge again 

 

[mad.ax]

So if you have signal on V3b grid, that should mean that V1 & V2 are ok.
On the other hand we know that the phase splitter stage is OK too.
Maybe the phase splitter use only V3a, and the problem is between the two halves of this tube.

Try to inject the signal in the loop return jack, and scope V3b grid

Axel