building a micro amp using 5902 submini's & nuvistors-Which of these output stages would you use?

[SoulFetish]

Quick one here:

So I drew up a couple of different output stage configurations using the available data, and found a nice site which plots out some data curves.

The first example uses a plate voltage of 176V, with a 10k p-p load, and the screen voltage set at around 90V. Here is what the load line looks like in this operation.





This was my initial idea, because I have a couple of OTs with this primary impedance that I could use (although they are ultimately oversized for this application.)
But I did have to drop the screen voltage a bit to pass through the knee with a 10k load, and I'm wondering how this might affect the tone?

In the second example, I took a bit more traditional approach and chose a load and bias level according to the screen voltage (110V), rather than a part I had on hand.
For screen voltage of 110V, and a plate voltage of 160V, a p-p load of 6k6 and idle bias of 22mA looks nice to me.


I'm liking the look of the second example, but honestly, I'm making audio decisions with my eyes here. Any of you have any thoughts on which direction you would go here? How one might objectively sound vs other?
Is it much ado' about nothing?

[mhuss]

Just looking at the graphs, the first seems to be a little safer, staying outside the max line, while the second seems to meet or pass into the unsafe operating area?

[Helmholtz]

Loadline partly intersecting the plate dissipation curve is quite usual and not necessarily a problem with class AB operation, as it only happens at some medium signal level and only during one half-cycle.

[SoulFetish]

actually, come to think about it, I know there are a couple of OT models available with dual primary taps of 10k & 6k6. I have a bunch of high capacity multi-deck rotary switches I could use to switch between the two settings.
Unfortunately, all the ones I know of are built to handle 20-30 watts or so. Even so, might be fun find out what kind of differences in response and sound there might be between the two. If I had to have a custom wound transformer, I doubt there would be much difference in cost between having one wound with a single primary or taps for both.
I did a little tweaking on the theoretical operating voltages so I could use a single shared 300Ω cathode resistor for both settings to achieve the desired bias voltage (one less thing to have to switch), and here is what the circuit might look like (including the minor changes in load lines):

[Helmholtz]

You can't short parts of the primary.

[SoulFetish]

You can't short parts of the primary.

Of course not. That's like, "No Shorted Turns" 101. Just making sure everyone's paying attention

[SoulFetish]

Here is an update with some practical results:
I bought the transformers for the output and power supply and built a test setup. Voltages were pretty much as expected, although the B+ and heaters ran a bit high (as they weren't fully loaded). I may need to tweak the B+ to drop it a bit, but I'll know better when I get the preamp designed and built.
For the audio test I disconnected the PI output from my other amp and ran a twisted pair to the test rig. Sounds great, but even so it's louder than I would have expected. I should have measured the sine wave output to see what the actual output power was. I can do that later.
Here is the test rig:


For the actual build, I'm going to make a mini-head unit so I can use what I have as salvage. The chassis is pulled from some small Fender solid state practice unit... a something-15, I forget.
For the preamp, I've got 7 Nuvistors to play with. 3 high mu triodes (6CW4 and 2 x 7895s) and 4 Med mu/high gm triodes (4 x 7586s).
So far, the pre is wide open to ideas if anyone has any suggestions? I'm tentatively thinking of taking some of those tubes and incorporating a true Vibrato/Chorus.
Here is the chassis and the components:

[SoulFetish]

I was able to find optimum Output Transformer chassis position for the lowest induced PT noise running the power transformer unloaded, and connecting the 16 ohm secondary into an amplifier.

I’m just looking for some advice before I drill the mounting holes for the output tubes. I've taken 2 photos of potential mounting positions for the output tubes using black hole plugs as place holders.
The idea is to position them so any external common mode noise (ie from the transformers) might be shared equally and balanced between them.

[Helmholtz]

I've taken 2 photos of potential mounting positions for the output tubes using black hole plugs as place holders.
The idea is to position them so any external common mode noise (ie from the transformers) might be shared equally and balanced between them.

Which pictures do you mean?

Why "common mode"? What is your coupling model? Common mode plate signals don't produce an output. A PP-OT only transfers differential signals.

I don't think an OT can emit significant interference, especially when having endbells.

[SoulFetish]

Which pictures do you mean?
.

Can you see them in my last post now? For some reason they didn’t get saved last night.

[SoulFetish]

I don't think an OT can emit significant interference, especially when having endbells.

The radiated hum from the PT is my main concern.

Why "common mode"?

I want them positioned so that any stray flux from the PT (primarily) is shared equally between the output tubes because common mode noise is rejected in the output transformer.

[Helmholtz]

Can you see them in my last post now? For some reason they didn’t get saved last night.

Yes, showing now.

I like the positions in the lower pic better because it means almost equal distance to the PT.

Also power tubes should not be too close to each other to prevent mutual heating.
According to Tomer lower bulb temperature tends to increase tube life.

[SoulFetish]

update: wiring heaters sucks.... that is all.

[SoulFetish]

okay, a bit more of a substantive update:
I'm without my access to my ban saw and drill press at the moment, so I have no reliable way of cutting or drilling FR4. So, I'm using minimal terminals and going old school style out of necessity. Might look a little haphazard, but it should be very quiet if I've done my job.



Obviously, it's incomplete. The OT centertap will run against the chassis and 3 preamp stages haven't been built

[SoulFetish]

Made some progress on the preamp tonight. Probably 1 more short session to wrap up some loose ends and install and wire up the panel controls.