Is my OT blown ? and associated question.

[hamfist]

I continue with my quest of finishing my restoration/modification of my 100W Selmer PA - converting it into a two channel Marshall 2203 plus extra gain stage type amp.
I've got the two separate pre-amp's both working a treat - these can be linked out of the FX loop send of the amp, to the FX loop return of my tube-rectified 2204 semi-clone. So I know the pre-amps and the FX loop of the Selmer are working OK.


However, if I play the Selmer's pre-amps into the PI and power section of the Selmer (which is now in 2203 spec), the volume is very low, as if it's a 1-2 Watt amp. I can easily play the amp at full volume in my front room. The volume does not vary, except with different MV settings.
This also happens if I play the 2204's pre-amp into the FX loop return of the Selmer.

I've tried everything I know. Checked all the wiring a million times. tried new PI tubes & new power tubes. I've checked the impedance across the OT's windings. The primary winding reads about 18 ohms between the centre tap and each outer wire. The secondary windings all read about 0.4-0.5 ohms between each tap and the common ground wire. I know my DMM is not very accurate at those low levels, but there's definately no direct shorts.
All leads (primary and secondary) have a high impedance to the amp's ground, when the transformer is disconnected.

THe only other thing I am not sure about is the grid stopper resistors, which are each 10K, instead of the standard Marshall's 5.6K, although I am sure all this does is dampen high end frequencies a little bit.

Oh, I've also either checked or replaced every component associated with the PI, from the FX loop return to the power section. No difference.

Here are the relevant voltages of the PI and power tubes (all power tubes have pretty much identical voltages).

PI

pin1 169V
pin2 0.045V
pin3 2V
pin6 167V
pin7 0.045V
pin8 2V


Power tubes

pin3 458V
pin4 454V
pin5 -33V

Any suggestions ? is there anything it seems that I might have missed, or is my OT toast ?

I don't feel confident testing my OT further, but was wondering if it would be safe to jumper the signal from the PI outputs (pins 1 & 6) of the Selmer (and ground, of couse), into the same point in my semi-2204 clone.
if I kept the power supply to the Selmer's PI intact, and put the output signal from pins 1 & 6 straight into the PI coupling caps of my other amp, that would keep the bias supply of the "receiving" amp intact. The PI coupling caps would keep all DC isolated between the amps.
Is this madness ? If it were safe and possible, it would at least identify if the problem lies in the PI circuitry or power tube circuitry onwards.

[tubeswell]

Have you got a schematic of what you've done? That might help to see things a bit

[hamfist]

Have you got a schematic of what you've done? That might help to see things a bit

Unfortunately no schematic.

But the from the PI and power tube circuitry is identical to a Marshall 2203 now, apart from the grid stopper resistors (mentioned earlier).

The power supply arrangements are also different. The main filter caps are two 100uf's in series, whch directly supply the OT (this is original to the Selmer).
From that point the supply goes through a choke, and then to a number of 32uf caps, which supply the pre-amp and PI. Directly after that choke is the supply to the power tube's screen grid supply.
the bias cap is a single 47uF cap.

Oh, BTW, all the electrolytics are brand new, put in by myself.

The OT is the standard Selmer one, which I believe is a Partridge. It is wired exactly as it was when I got it. Outer primary wires to Pin 3 on each tube pair. Centre tap connects straight to the + voltage point of the two 100uf caps in series.

Apart from potential differences tonally, is one 4 x El34 OT much the same as any other, in specs ?

[6267]

If it's playing undistorted, it's most likely an okay OT. Did you change the power section to match a marshall? The old amp may have needed a funky primary impedance that would excessively damp or otherwise limit output with a marshall config. Alternately, is the speaker on the right secondary tap? How much AC voltage are you getting at the PI and or between power tube grids? I suspect an impedance problem or weak output from preamp for some reason before I'd jump to conclude a dead OPT (I've only killed one OPT, and it was an AA5 radio OPT, and I've certainly tried to kill others accidentally...60w amps with no load + input signal even).

[hamfist]

If it's playing undistorted, it's most likely an okay OT. Did you change the power section to match a marshall? The old amp may have needed a funky primary impedance that would excessively damp or otherwise limit output with a marshall config. Alternately, is the speaker on the right secondary tap? How much AC voltage are you getting at the PI and or between power tube grids? I suspect an impedance problem or weak output from preamp for some reason before I'd jump to conclude a dead OPT (I've only killed one OPT, and it was an AA5 radio OPT, and I've certainly tried to kill others accidentally...60w amps with no load + input signal even).

The Selmer's PI/Power section/OT is definately also adding a lot of extra distortion, even though the output is quiet. I forgot to mention this in my original post.

THe speaker is definately on the correct secondary tap.

I currently have no means of measuring AC voltage. I'll have to look into it.

The thing that makes me think that I have a good output from the pre-amp, is that when I plug it into the PI/power section of my 2204, I get a very healthy volume.

[Steve Conner]

A long-tailed PI, such as both the Selmer and 2203 have, should have about 50, 90, 100, whatever, volts on the cathodes. Definitely not 2V, so go and check your work a million and one times. (Did you connect the 1M grid resistors to the right place? It's not ground)

And yes, jumping the PI outputs of one amp to the power tubes of another should work fine, provided you do it so that the power tubes actually in use have grid stopper resistors on their sockets. Otherwise the long wires could cause parasitics. There'll be no negative feedback, so it should be scary loud.

[hamfist]

A long-tailed PI, such as both the Selmer and 2203 have, should have about 50, 90, 100, whatever, volts on the cathodes. Definitely not 2V, so go and check your work a million and one times. (Did you connect the 1M grid resistors to the right place? It's not ground)

And yes, jumping the PI outputs of one amp to the power tubes of another should work fine, provided you do it so that the power tubes actually in use have grid stopper resistors on their sockets. Otherwise the long wires could cause parasitics. There'll be no negative feedback, so it should be scary loud.

Great. Thanks Steve. I really hope you're right. The last thing I need right now is to have to buy a new OT.
I'll be able to have my 1,000,001st check of the wiring tommorow some time.

I'll report back !

[stokes]

I would also suggest getting those 10k grid stoppers out of there,the 220k/5.6k form a voltage divider that attenuates the signal to your power tubes,the 10k you have there will attenuate it even more.

[hamfist]

I would also suggest getting those 10k grid stoppers out of there,the 220k/5.6k form a voltage divider that attenuates the signal to your power tubes,the 10k you have there will attenuate it even more.

THanks. Yes, I certainly plan to do that soon. I just don't have any 5.6K's at the moment.
My main focus is just to get the thing working to start with !!

[hamfist]

A long-tailed PI, such as both the Selmer and 2203 have, should have about 50, 90, 100, whatever, volts on the cathodes. Definitely not 2V, so go and check your work a million and one times. (Did you connect the 1M grid resistors to the right place? It's not ground)

And yes, jumping the PI outputs of one amp to the power tubes of another should work fine, provided you do it so that the power tubes actually in use have grid stopper resistors on their sockets. Otherwise the long wires could cause parasitics. There'll be no negative feedback, so it should be scary loud.

Just been looking at the marshall PI schematic again, and a huge realisation has just dawned on me. I'm not with the amp currently but I'll bet I know what stupid thing I've done.

I bet I've connected a 470K instead of a 470R to pin 8.

Oh, please, please, please let that be it !!!!!!!!!

[MWJB]

That would elevate voltages, not reduce them...47ohms instead of 470 would drop them.

What voltage is feeding the PI plate resistors, that connect to the caps at pins 1 & 6 of th PI?

[hamfist]

Well, like a fool I HAD put a 470K onto pin 8 of the PI, instead of a 470 ohm.

Steve - you're a star pointing out those low voltages. It was fairly straightforward to suss it out once I had little clue.

Anyway, the resistor is now changed, and the volume is loud, and appropriate !

However, as nothing is ever easy with this project, another problem has now surfaced, due to the 470R in the correct place ... I think that'll be another thread !

Thanks to all who commented.