Output tube for 600V B+, 5800 ohm primary PP, tertiary 25% UL taps?

[GordonW]

OK. As I mentioned in the "output transformer testing" thread, I just got a pair of Bogen MX-60A amps. Not gotten to bandwidth tests yet, but I did some more basic tests (ratios of windings, and have some interesting results, in any case.

The amps originally had 8417 output tubes, in pentode mode... 615V plates, 295V screens. BUT- these transformers have two "ancillary" windings... one of which was originally a 70V output, and the other was a feedback winding. However, I've determined that the windings are EXACTLY the same ratio... about 25% of the outputs, from a turns or voltage standpoint.

Sooo... I'm thinking... ultralinear? 25% is a bit low, but NOT out of the range for SOME tubes. What I'd like to find, is a tube that would work well with a 5.8K CT transformer, with a 25% ultralinear "tap". Mind you, that since these are SEPARATE windings... doing stuff like lower DC voltage on the ultralinear/screen windings is perfectly plausible... which opens up usage of stuff like sweep tubes and such. My current leaning is maybe something like a 6146B or maybe a 6JB6... nice thing, is that the 6146 would only be asked to deliver 60 watts or so, power output... about 60% of its maximum "CCS" rating (IIRC, that max is about 94 watts or so).

I'd guess, that with the high screen gain of a 6146, and that since the screen voltage would be MODULATED with the plate voltage (they should follow each other), that I should get decent results... the 25% ratio should give enough "feedback" to make it act similar to a 6L6 or 6550 with 40% taps, due to the higher screen gain, I would think. And, also, the modulation would stop any possibility of "arc-over" in the 6146... the screen and plate would never be in "opposite polarity" (i.e, the screen high and the plate low)...

Right now, the alignment that's in my head, is the 6146B, 615V plates, 175V screens, through the 25% tertiary windings. Also, I would use output plate-to-LTP-plate feedback (resistor between the two plates), to further reduce output impedance (and increase high-frequency bandwidth)...

Any thoughts on whether this might work? From what I see, it looks like I'd get TONS of gain from the 6146B (due to the low screen voltage), so I could use an LTP phase splitter and still have plenty of drive voltage, with the right tube in the LTP, even WITH plate-to-plate feedback...

Thanks!

Regards,
Gordon.

[Bandersnatch]

hey-Hey!!!,
Interesting discovery indeed!!!. I am now wishing I hadn't gotten rid of my pair of them. You could also run 25% cathode FB. With a fixed g2 supply, you get that same U-L effect. That would also bring you to about 9k load with the 25% cathode coils. 1.25-squared times 5k8 a-a is how it is determined. That will make you quite a driver requirement...but not as bad as the unity coupled McIntosh.
cheers,
Douglas

[GordonW]

I'd be a little concerned about putting the coils (especially the feedback coil) in the cathode circuit... might be a current-handling problem, potentially. However, with the screens never drawing more than maybe 20ma max, shouldn't be any problem, as an ultralinear winding.

Oh- further testing has revealed that the windings DO have different DC resistance... one is about 13 ohms, the other 5 ohms or so. But, the one with higher DC resistance has very slightly higher output (like 2% higher) in open-load, too. I'm guessing that the two effects will probably cancel out, as far as drive is concerned. Also, realistically, a 13 ohm load is pretty much still a "dead short" as far as the screen is concerned... 13 ohms with 20 milliamps through it is only a .26 volt drop (out of 175 volts)... or about .15% error. Don't expect any DC-voltage related imbalance problems there...

Regards,
Gordon.

[GordonW]

BTW: When I pulled the lid off the MX-60A. I just stood back and just looked at the output transformer...

My first reaction was "holy s***, that thing is FUGE!"

My second reaction is, now "I guess I probably won't have any low-frequency power problems..." I would think that it should have no problem making bass!

Regards,
Gordon.

[Bandersnatch]

hey-Hey!!!,
Well, your fine investigation has officially put me into the market for some MX60's. Looking forward to hearing of your results.
cheers,
Douglas

[cerrem]

Not sure of your exact voltage ratios..... Sometimes these output transformers had 500 ohm winding for driving cutting heads, remember records????? SOme outputs used 100V line outputs also.... So at 5.8K at 60W you get 590V AC across the primary...
25% of that would be 148 volts that would be roughly equivelent to a 362 ohm tap......
I am not sure what it is?????
How accurate was the voltage ratio test??? DId you account for excitation current....???

Chris

[Bandersnatch]

Originally posted by cerrem

How accurate was the voltage ratio test??? DId you account for excitation current....???

Chris

Chris,
Would you care to explain how excitatin current would throw off a voltage-based ratio-ing job?
cheers,
Douglas

[GordonW]

Originally posted by cerrem
Not sure of your exact voltage ratios..... Sometimes these output transformers had 500 ohm winding for driving cutting heads, remember records????? SOme outputs used 100V line outputs also.... So at 5.8K at 60W you get 590V AC across the primary...
25% of that would be 148 volts that would be roughly equivelent to a 362 ohm tap......
I am not sure what it is?????
How accurate was the voltage ratio test??? DId you account for excitation current....???

Chris

Yep, if you put the two tertiary windings in series, you would get about 140v... one of them was the 70v winding (with SEPARATE return/common lead, independent of the "normal "secondary common lead), so it makes sense that two in series would do 140V.

And yep, the original B+ on the amp was 615V... so 590V swing on the primaries is quite in line...

Excitation current should affect all windings the same, no? Why would it skew one in relation to another, in terms of changing voltage ratios?

If you meant "load current"... as I mentioned, the screens draw very small currents (less than 20ma)... there should be very little voltage drop there... and the primaries are properly sized for the power and current of the output tubes (about 225ma or so per tube, for the 6146, as opposed to max of 275ma for the original 8417s- pretty comparable)...

Regards,
Gordon.