New build debug...

[Paul-in-KC]

I found the schematic. You can increase the negative bias voltage by increasing the 27K resistor on the bias pot to 33K or 39K. Or, you can decrease the 3.3K bias range resistor. I wouldn't go below about 470Ω. Either of these changes should increase the bias voltage enough for 6L6s. It's OK to change both resistors but I doubt that will be necessary.

Okay - I had to go all the way to 100K to get -59.75. But that's probably getting most of it since the AC is only 53 VAC (RMS). I think that would be roughly 75 VAC peak. But absolute max would be something below that.

As you probably already knew - that eliminated the buzz from the OT. Yay!

What about those high plate voltages? Should I do something to get those closer to the values listed in the schematic?

Here is a PDF of my schematic. This is mostly accurate to what I built - except the FET module which I did with an NTE452 and the eyelet board documented by MP Manning 2016.11.13. And you can see I haven't updated the B+ voltages with what I am actually getting. OT and speaker out aren't exactly right (but close enough to get the right idea). And there may be one or two other items that I am not thinking of at the moment.

Essentially I tried to build the 3rd Gen circuit - adding Master Volume and Presence - to match the chassis that I bought. No PAB switch or relay. OD switchable manually or relay (at least that's the plan).

ODS50W-3rd-4th-hybrid.pdf

[sluckey]

Okay - I had to go all the way to 100K to get -59.75. But that's probably getting most of it since the AC is only 53 VAC (RMS).

I would not like having such a large resistor on the pot. I suggest put in a 33K or even go back to the 27K. Then replace the 3.3K with a 470Ω (larger if needed). 53VAC should easily make -70Vdc.

[Paul-in-KC]

I would not like having such a large resistor on the pot. I suggest put in a 33K or even go back to the 27K. Then replace the 3.3K with a 470Ω (larger if needed). 53VAC should easily make -70Vdc.

Can do.

I can see where lowering the 3.3K would deliver more voltage to work with at the pot (at least that's how I would intuit it). Which helps since I needed more. But I'm not sure why you prefer a smaller resistor on the pot. I'm sure there is an obvious reason, it's just not obvious to me.

Is it that the first resistor (3.3K) should be scaled to get you to a certain starting point based on your bias tap output and the desired center value of the adjustment? Something like that?

BTW - everything seems to be working quite nicely. I haven't tested the foots witch/relay yet - as I need to wire the cable.
Will bias after I get that circuit in the shape it needs to be.

Any thoughts on those B+ voltages? Try them first? Change them first?

Again - so many thanks for taking the time to share your knowledge!

-Paul

[sluckey]

But I'm not sure why you prefer a smaller resistor on the pot. I'm sure there is an obvious reason, it's just not obvious to me.

Just trying to keep the 6L6 grid resistance about the same. When you put a 100K on the bottom of the bias pot you almost double the 6L6 grid leak resistance. Trace the path from pin 5 to ground and you'll see what I mean. This ain't a big deal to me because millions of Fender amps use 220K grid bias resistor. However, RCA tube manuals specify 100K as the max. But if you wanna keep it true to the dumble…

Any thoughts on those B+ voltages? Try them first? Change them first?

Those B+ voltages will drop some when you bias the 6L6s hotter. Wait until you see where they end up before you start trying to drop them. 6L6s will work just fine with the voltages you have now.

[martin manning]

There are two reasons to keep the resistor going from the pot to ground small. One is it’s part of the grid leak for the power tubes. That’s not a big deal here since the grid resistors are only 100k, and there are only two power tubes. The other reason is the available bias adjustment range will be reduced.

I would keep the dropping resistors as specified, and adjust only the first one after the choke if you want to raise or lower the V1 plates to get in the 180-190V range. Get the power tube bias adjusted before you worry about that, though.

[Paul-in-KC]

But I'm not sure why you prefer a smaller resistor on the pot. I'm sure there is an obvious reason, it's just not obvious to me.

Just trying to keep the 6L6 grid resistance about the same. When you put a 100K on the bottom of the bias pot you almost double the 6L6 grid leak resistance. Trace the path from pin 5 to ground and you'll see what I mean. This ain't a big deal to me because millions of Fender amps use 220K grid bias resistor. However, RCA tube manuals specify 100K as the max. But if you wanna keep it true to the dumble…

Any thoughts on those B+ voltages? Try them first? Change them first?

Those B+ voltages will drop some when you bias the 6L6s hotter. Wait until you see where they end up before you start trying to drop them. 6L6s will work just fine with the voltages you have now.

See - I knew there was perfectly good and obvious (not to me) reason(s). I'll have to study that (grid leak resistance) so that I understand what's going on.

I went back to 27K on the POT to ground - and a 470 in place of the 3.3K. That nets me -64.25 as the max negative. I assume this is good and those values can be permanently adopted?

Roger that on the B+ (waiting until biased).

You guys are awesome.

-Paul

[Paul-in-KC]

There are two reasons to keep the resistor going from the pot to ground small. One is it’s part of the grid leak for the power tubes. That’s not a big deal here since the grid resistors are only 100k, and there are only two power tubes. The other reason is the available bias adjustment range will be reduced.

I would keep the dropping resistors as specified, and adjust only the first one after the choke if you want to raise or lower the V1 plates to get in the 180-190V range. Get the power tube bias adjusted before you worry about that, though.

Thank you, thank you, thank you!

Like I said, you guys are the best!

-Paul

[Paul-in-KC]

...On further investigation - I don't think I can keep the 470 in place of the 3.3K (looks like I'm going to need to bump this up).

I don't think I can get the tubes hot enough, but I'll have to put some 1 Ohm resistors on there to measure the total current. Which I can't get until tomorrow.

But that's okay, I still need to wire my foot switch cable anyway. So, I'll do that - test it and wait for the 1 Ohm resistors to get here tomorrow.

-Paul

[sluckey]

...On further investigation - I don't think I can keep the 470 in place of the 3.3K (looks like I'm going to need to bump this up).

Please explain why you say this. How can you know at this point?

[Paul-in-KC]

...On further investigation - I don't think I can keep the 470 in place of the 3.3K (looks like I'm going to need to bump this up).

Please explain why you say this. How can you know at this point?

Well - as I say, I don't know for sure - I'll have to wait until I can measure cathode current. But looking at the current across the 470 screen resistor - it seems very small. Having said that I don't know what percentage this current is of the total cathode current.

But here's what I see across the 470 screen resistor...

[ with the 27K (pot) and 470 "pre-diode" resistors ]

maximum voltage across screen resistor is 0.56 VDC; this is at the most positive bias pot adjustment - which is -47.6 VDC. So, according to Ohms law (if I have it right) that's about 1.19 mA.

I think my target total should be around 23.5 mA
Which is 70% of 33.6 mA
Which is what I figure to be (roughly) the maximum plate dissipation with a plate voltage of 444 VDC (which is what it is measuring now at that bias voltage)

So, if the screen current is a small percentage of the total then I may not need to change.

NOTE: with those two values (27K & 470) my minimum screen voltage is 0.013 VDC; at the most negative bias voltage which is -63.5 VDC

Thinking on the run here - but looks like screen should be about 12% of total (3 out of 23.5 total watts). If this is correct - that puts me at just under 10 mA total. So, I think I'll need to go less negative, no?

-Paul

[sluckey]

Quit guessing. Put some 1Ω resistors on the cathodes and start knowing.

[Paul-in-KC]

Quit guessing. Put some 1Ω resistors on the cathodes and start knowing.

Right - agreed. Was just responding to your request "Please explain why you say this".

I will put them in as soon as I get them (should be tomorrow) - and will post the values that I end up on.

Thanks!

-Paul

[martin manning]

You can always use the DC resistance of the OT primary to get idle current: Just measure Ra-ct on both sides (resistance from CT to plate) with power off, and then measure the voltage drop across the same two points on each side with power on. Current is (Va-Vct)/(Ra-ct).

[Paul-in-KC]

You can always use the DC resistance of the OT primary to get idle current: Just measure Ra-ct on both sides (resistance from CT to plate) with power off, and then measure the voltage drop across the same two points on each side with power on. Current is (Va-Vct)/(Ra-ct).

Of Course! Brilliant! - sounds like a good exercise. That would measure plate dissipation only - yes?

I'll try that for my present resistor pair (470 & 27K) and see what's happening.

Thank you (again) sir!

Will post results after breakfast

-Paul

[martin manning]

Yes plate current only. It’s more of a fuss than a 1R at the cathode, which is a direct measurement at low voltage, but it works fine.