[Maxx134]

Ok, enough speculations of "DC offset" class 101 .
Take enineering questions /concerns to PMs so as not to confuse the masses..

Nice read, but the opamp don't care about the resistors or the tubes.
They only monitoring/regulating the plus and minus output levels of tubes...

That's why YOU had two different power tubes and didn't blow your amp.

Now suddenly your all worried about their parameters.

Oh well, semantics here,
You can't please everyone.

The main point is that all will be fine on next mod.

 

[coinmaster]

You're missing the point. The opamps provide the bias voltage to the top triode always keeping it's impedance in line with the bottom triode.
Therefore the current's through both tubes should always be equal no matter what, assuming ra and rk are the same value. The only thing I see changing resistor values would do in a servoed design is change the load line operating point.
 
I could understand the resistor optimization if this was a capacitor or transformer output WCF because there would be no dynamic bias adjustment but in our case the currents are always kept in line.

 

[Maxx134]

The "flat" IC in the center of the amp is the thing that cuts the output ...

We went over this in PMs long ago about that skinny chip...
You later confirmed it's function in PMs,
So you know Ridge78 knows all this.
What you don't know, is our testing on the timing cap that is next to it.

And of course the designer Sword-Yang once again made the best choice. .

In fact, the circuit board is a thing of beauty. ..
All plate resistors on component side,
All Cathode resistors on other side...
Grid caps,
WCF caps,
All is extremely well laid out.
The board design is a thing of beauty,
Once you can see it...

 

[Maxx134]

I don't see what you see.
I saw the opamp only recieving from output, and sending back a control to the * grid.


You will have to take up your conclusions with the designer.

In any case, I will have to label your theories incorrect.
Why?
Because the DC offset part of this design has nothing to do with what we are doing.

The DC offset keeps on doing what it always does.
You are mixing two different topics.

Edit:
Side note:
I been running my amp daily for months at a calculated setting of 31.5ohm,
and verified predicted sonic changes of both low impedance Ether, and high impedance HD800.

They behaved predicably according to an impedance change.

If you can't wrap your head around that, sorry.

But don't come in hear speculating and confusing others.

As stated from beginning of this thread,
All mods are done, tested and confirmed before being suggested here.

It is not a design thread.

You can make a design thread and we can go back and forth all day there.

 

[coinmaster]

All is extremely well laid out.
The board design is a thing of beauty.

Tell that to someone who reverse engineered it. The thing is an incoherent mess when you follow the traces 

, both channels are totally different in their trace layout as well.
 

Because the DC offset part of this design has nothing to do with what we are doing.

The DC servo has everything to do with the triodes operating point and current matching, without the dynamic grid voltage the tube would not remain at 75v/75ma when the load impedance changes or when the tube ages. This isn't speculation, this is fact. I've tested and confirmed it in simulations and on the bench.
 

I been running my amp daily for months at a calculated setting of 31.5ohm,
and verified predicted sonic changes of both low impedance Ether, and high impedance HD800.

You should have said that a long time ago, I've asked you at least 4 times if you've actually verified any sonic differences so we are not chasing our tails here and you kept ignoring me.
 

 

[Maxx134]

Yes, yes I know each opamp directly control each tube output levels,
But the parameters of the tubes are independent of the opamp control, as you already shown with using two different tubes. .
Although the Plate/Anodes where the same, yet even if they were off a bit,(affecting plate current AND circuit impedance) as in stock cheapo units,

I believe that is what the trim pot is for...

Side note:
I have had loose pin connection on one of the opamps, and the result is a dip or loss in volume on whichever tube it corresponds to..

So anyone having that symptom would know its most likely the board pin socket problem..

Edit:
Sorry my phone went crazy in my pocket! Lol

 

[coinmaster]

But the parameters of the tubes are independent of the opamp control, as you already shown with using two different tubes. .
Although the Plate/Anodes where the same, yet even if they were off a bit,(affecting plate current AND circuit impedance) as in stock cheapo units,

I believe that is what the trim pot is for...

Well the opamp provides the bias for the top triode, so the operating point of the top triode is directly based upon ra and the bias voltage provided by the opamp.
The bottom triode is always at a fixed operating point because the bias and rk do not change. Hence the opamp is always struggling to keep the top triode in line with the bottom triode in order to maintain the 0v output because the non inverting input of the opamp is attached to ground so the opamp will do everything it can so that the inverting input is also at ground potential.
The trim pot shunts the supply voltage to the opamp allowing different output voltages based upon the input voltage to the opamp. You want to make sure there's less than 100mv on the output of the amp, preferable below 50mv using the trimpot to adjust.

 

[Maxx134]

I bought 4 opamps to replace and will be looking into trimpot adjusting once I get my oscilloscope in.
Do you have oscilloscope?
I just got a kit it was cheaper than buying one.

Edit:
The reason for opamp purchase was a mistake as I didn't realize the chip contacts into the socket was loose.

The issue remains even though I cleaned and slightly bent the pins for optimal contact,
Because with the heat of the amp, the one chip in question, somehow still manages to get loose.


So I will be replacing it.

And since I get carried away, I bought 4 opamps chips.

 

[coinmaster]

I bought a https://www.picotech.com/oscilloscope/2200/picoscope-2200-specifications since picoscope is well renowned and it beats paying 600 bucks. Has worked well so far.

 

[Maxx134]

Yes that is an excellent idea as it has software for computer interface and can do more things easily.

I purchased a much simpler unit.
A small kit with 4 inch led screen.
Once it arrives I will see how good it is.

 

[baronbeehive]

  Make sure the mounting hole next to the driver stage power supply is connected to the chassis. It is the only ground exit for the board and should be the only direct connection between ground on the board and ground on the transformers/wall.

 
So I can't test it with the board out after all, this was what I was worried about. Lucky I haven't switched it on yet.

 

[coinmaster]

Just use a clip and a wire to connect it to ground. 

 

[baronbeehive]

  Just use a clip and a wire to connect it to ground.

 
Oh you idiot....... me, not you

. I had a feeling there might be a ground in there somewhere, luckily I was going over some past posts looking for solder locations and came across your post.

 

[coinmaster]

It's entirely possible that nothing bad will happen if you don't connect that part to ground. The black wire that leads from the transformers to the capacitors in the power supply is also connected to ground through some sort of tap in the transformer, but this tap is a mystery to me. It's certainly not a normal ground point as I've blown a few fuses while using it as one.
The grounding hole that I spoke of near the input stage put supply is required for the rectifier operation and bad things can happen if it isn't grounded. I wouldn't risk it.

 

[SonicTrance]

This should not be an issue for you @baronbeehive as you're not removing the board completely, only tilting it down leaving all wires attached, correct?