Help with a new Build ODS

[Richard1001]

On the voltage chart the problem seems to be between the MV (2,9V rms) and phase inverter input (60mv)

Looking at the phase inverter, what is the value of the grid resistor on pin 2. It looks like a 1k resistor, it should be 1M. The grid resistor on pin 7 is not visible, but should also be 1M.

[Wardy]

Many Thanks for looking at this.

My new signal generator arrived which is much better and I have taken a comprehensive fresh set of readings. I have established that Vin = Vout at the master Vol with the reverb circuit removed, but not when it is in place. I have taken a lot or readings with it in place but then I ran out of time. I want to redo all the readings side by side with the reverb circuit removed to compare.

I hope you are right about the grid resistor values, I'll check them tomorrow eve as thats the first chance I'll get. it would be great if you are right.

i'll get back hopefully in 24hrs

best

steve

[Richard1001]

Many Thanks for looking at this.

My new signal generator arrived which is much better and I have taken a comprehensive fresh set of readings. I have established that Vin = Vout at the master Vol with the reverb circuit removed, but not when it is in place. I have taken a lot or readings with it in place but then I ran out of time. I want to redo all the readings side by side with the reverb circuit removed to compare.

I hope you are right about the grid resistor values, I'll check them tomorrow eve as thats the first chance I'll get. it would be great if you are right.

i'll get back hopefully in 24hrs

best

steve

The colour bands of the resistors on V4-2 and V4-7 should be: brown/black/green.

Good luck!

[sluckey]

Looking at the phase inverter, what is the value of the grid resistor on pin 2. It looks like a 1k resistor, it should be 1M. The grid resistor on pin 7 is not visible, but should also be 1M.

I agree.

[Wardy]

Just to let you guys know that I'm partially color blind and I can't read resistor codes, I measure every single resistor just before fitting. It is very possible i made a mistake and I can't wait to get home and check this out, I'll be over the moon if this is it. i can't remember whether I rechecked these ones as I was focusing around V1 after the DC offset thing sent me in the wrong direction.

many thanks


Steve

[Wardy]

Well Chaps,

I owe you a massive thank you, you were exactly right about the phase inverter grid resistors. I probably saw 0.998 show up on my meter when measuring them and forgot to note that this was K not M!

Once again thanks for taking the time to look at this it was driving me crazy. I did waste a lot of time looking in the wrong place due to the DC offset thing. Lesson learned!

The amp sounds like I would expect now. plenty of power in both channels and all behaving as it should. Then I played around with setting the PI Trimmer. The build guide supplied by Tubetown recommends balancing by measuring the pk-pk on both sides with a scope and balancing. On this forum it seems to favour the Ampeg method of measuring the cathode voltage and setting to 0. I tried both but they did not concur. When setting using the Ampeg method I read an imbalance on the scope and visa versa. When set using the Ampeg method, I then measured the total output power with the scope to see if I was getting 50w. I got close to 50W but the wave was clipping on one side a lot more than the other. I'm going to try resetting using the scope method and try this again but ran out of time this eve.

best

Steve

[martin manning]

Ampeg method recommends balancing at ~half power.

[dbharris]

Glad to see you got this issue sorted out!

-Dan

[Wardy]

Hi again,

thanks for all your help. I'm not sure if there is still something wrong here. i have tried to measure the power output to see if I'm getting 50w and although the amp sounds pretty loud now I am unable to get a clean 50w power output when using the scope. I'm now getting maybe 35w. The bottom of the wave goes out of shape and then clips quite a long time before the top. Is this normal?

best

Steve

[Wardy]

So I had another good look at this today.

I checked the PI trim pot and it made no difference on when the wave starts to clip. It was also similar at any of the impedance setting. The bottom of the wave starts to clip at 7.4vrms and the top at 13vrms when on 4ohm load/ so power about 13 clean watts and rising to 42w for the top. Quite a big difference.

I have tried different power valves and currently have in some Tung Sol 7581a that can go to 35w each. Before that I had Nos Philips 6l6wgb's.

I measured the bias and its steady at the grid pins and the -50v supply but rises with increased power at the cathodes/bias check points.

On the NFB circuit I checked all the resistor values and they are right. I removed the wire from the impedance selector and the asymmetric clipping stopped. Can get 13vrms from both sides of the wave. Still a bit under power though. The transformer ratings are 300mv for the power transformer and the paperwork that came with the output transformer says max 75w so not sure why I can’t get more or why the NFB circuit would cause this.

I’d previously swapped the GNFB wire from the 4ohm to the 8 ohm tap as this was recommended by ampgarage as the existing nfb circuit is apparently designed for 100w not 50 on a 102 ODS. It was said that doing this makes it behave the same as the 100w and is correct for 50w. However I swapped it back to the 4 ohm tap and the asymmetric clipping improved a lot . Its still there but now the bottom of the wave starts to clip at 10 vrms rather than 7.4. the top still clips at 13.

I’m not sure whether this is “just the way this amp is” It sounds and plays OK. I tried it with the gnfb disconnected and it had a lot more gain- however the overdrive channel became very difficult to set up as then the volume on V1 had to be kept very low. I had added a mod on a switch as there was a spare hole in the back for some reason. The designer of this amp left out the local NFB on V1b which should be there so I put it in on a switch. I think the switch would serve better if it were changed to a three way resistor setting for GNFB and hardwire the other circuit as it makes little noticeable difference whereas the GNFB seems to make a lot.


Any thoughts on the above? Do you think the amp still has a problem or am I seeing things that aren’t there?

best

Steve

[martin manning]

Is the idle current on the power stage well balanced?
Do you know if the power tube gm’s are matched?
Do you know if the triodes in your PI tube are well matched?

[Wardy]

Thanks Martin,

I have implemented dual bias pots so they are exactly balanced. Both sets of power tubes I have tried were matched sets. I have tried several PI tubes with similar results. The originally supplied tube was marked as "balanced"

The amp actually sounds very good and I experimented with some different tubes this eve and ended up with nos mullards in V1&2. a 12dw7 JJ in the single tube reverb and a long plate sovtek in the PI. The amp sounds great just isn't making 50 clean watts no matter what I do. As I said if I take out the GFB circuit, everything becomes balanced, with zero asymmetric clipping, but still can only get to about 40w clean.

don't know whether to call it done or whether I should keep looking?

[martin manning]

Truth be told, you're not likely to get a clean 50W out of it. Maybe double check your component values in the feedback loop just to be sure.

[Wardy]

Thanks Martin,

I've had some comms with this particular clones designer in Germany and he says he thinks 40w is in the right ball park. I was also considering whether its possible that the limiting factor is the phase inverter tube and not the power tubes as its got a pre PI master Vol. I wonder whether it would be able to develop more output with a post PI master Vol? Am I correct here or barking up the wrong tree? Not that I'm too worried now it sounds great. I do think a useful mod would be to add switchable settings for GNFB though.

many thanks for sticking with me on this I appreciate the support.

Steve

[martin manning]

If designed properly, the MV should allow full power. If you can drive the power tube grids to 0V, that's all you are going to get. The NFB level will affect that, of course.