question about PI in Magnatone 280

[dinkotom]

I'm probably totally off-target here, but can anyone explain why did they use two different tubes (half of 6CG7 (V5-B) and a half of 12AX7 (V6-B), but with same resistor values, for the PI's in this amp?
A friend told me (if I understood him correctly) that those two pairs of output tubes will be out of balance if PI's have different amplification factor (6CG7 has 20, 12AX7 has 100).
http://www.schematicheaven.com/bargainbin/magnatone_280a.pdf

[sluckey]

It's hard to guess what the Magnatone engineers were thinking. Looks like they were not thinking at all! Seems more logical to me to have used the same tube for the voltage amp and PI in each channel. That would have simplified wiring considerably. But then it might not look like a Maggie under the hood!   

It looks like they designed the amp from left to right and when they got to the PA, said we may as well use that spare triode from the Vibrato oscillator.

The voltage amps just before the PIs are the same type tubes and will provide the same voltage gain. I wouldn't be concerned about the different tubes in the actual split load PIs. A split load PI operates with a voltage gain slightly less than one (similar to a cathode follower), so the actual gain factor for the tube is not an issue.

Are you still considering building this amp?

[dinkotom]

Hi Steve!
Thanks for explanation.
And, I'm happy to say, I already built this amp, see here: http://www.el34world.com/Forum/index.php?topic=7854.msg69851#msg69851
but, I'm still tinkering with it...

[sluckey]

Ah yes, I remember now. I remembered the 260 but had forgotten you also did a 280.

[jjasilli]

Yes, in the split load PI the output of the two reversed phased signals comes from the plate & cathode of the second triode.  The second triode must be balanced for unity gain, so that the plate puts outs out the same signal strength as the cathode.  

The second triode needs to be driven by a prior gain stage, which could be anything.  E.g., some hi-fi amps of that era used a pentode-triode tube in one bottle.  Different types of tubes have different tonal characteristics, so the use of the 5CG7 as the PI driver may contribute to the tone of that amp.

[PRR]

As an approximation, the gain of a cathode follower can be estimated from the gain of that tube plate-loaded in similar load.

Call gain "A", plate-loaded gain "Ap", cathode-loaded gain "Ak".

Ak ~= Ap/(1+Ap)

12AX7 with load gives gain Ap of about 50. So Ak is about 50/51 or 0.98.

6CG7 with load gives gain Ap of about 15. So Ak is about 15/16 or 0.94.

The difference is less than 5%. Compare this "error" against 10% resistors and three +/-20% tubes, it is quite insignificant.

In split-load cathodyne, the difference is a bit more, but I doubt it is worth computing.

Anyway, in tremolo mode, some difference may enhance the effect.

I do think that it is more an accidental convenience than a design decision.

[billcreller]

 Everything I build is an accidental convenience 

[dinkotom]

PRR, thanks for explanation!

There is another thing I probably should have done differently regarding the output tubes - instead of 6CZ5 I used 6V6, and found this yesterday:
''The cathodyne operates under 100% internal feedback, like a cathode follower, so it is extremely linear before clipping. Therefore, when choosing a load we are really only concerned with output signal swing. If the power valves are sensitive types like EL84s or 6V6s, then we don't need huge amounts of swing and a total load around 47k to 100k would probably do. If we need to overdrive bigger valves like EL34s then 200k is probably in order.''
(from Valve Wizard page about cathodyne PI):
http://www.freewebs.com/valvewizard/cathodyne.html

My amp has 100K plate and cathode resistors as on schematic, maybe I should change to 47K?

[PRR]

> maybe I should change to 47K?

Why? Is the amp working wrongly?

Rather than wave generalities, you can use the amplifier tables published by the tube makers. KNOW how much grid swing you need. Generally a bit more than the static bias. For the published plan, 24V. For 6V6 at 10W idle, 375V on plate (25mA target), 312V on screen, about 27V bias. Add 10%-40% for overdrive, 30V-40V peak per grid. Two bottles, so 60V-80V total cathodyne output.

A very important point: what power grid resistors will you use? Self-bias can often be 500K or more. Fix-bias is usually required to be 100K (or less). The Maggy plan is fix-bias and shows legal 100K resistors. The cathodyne sees them in "series", so the effective load is 200K.

What supply voltage is available to the driver? Maggy plan shows point B as 290V.

The G.E. 1961 sheet for 12AX7 page 2 shows good amplifiers. We have two suggestions for Rs=0.24Meg (close-enuff to our 200K load). Under 300V we learn we can expect 40V or 37V RMS, 56V or 52V peak total, 28V or 26V peak per grid. We slightly prefer the 56V, which was gotten with Rp = 0.1Meg = 100K. In cathodyne terms, this is two 50K. However the performance with 0.24Meg (two 120K) is not a lot less.

Interesting that we want 60V peak or more, and Maggy has given us 52V peak. This barely works with their tubes and bias, seems to be a hair short with your tubes. Going 47K is "better", 56V peak, but not a lot.

As a gain-stage, THD will approach 5%. As cathodyne this is divided by the gain reduction, about 25, so THD will be negligible at these levels. We might get a little more, but it will run into clipping pretty soon.

The G.E. 1956 sheet for 6CG7 page 3 shows 6CG7 operation. With the same parameters we can get 65 or 53V RMS, 92V or 75V peak two grids, 46V or 37V peak per grid. So the 6CG7 side is ample.

That's the cathodyne alone. But it is direct-coupled to a gain stage, which is biased for good cathodyne DC level. V3-A V5A can't swing more than 45V.

Some of this may be "good enough" hasty-work design; or it could be months of ear-testing. But amps are never "perfect", and guitar often loves "imperfect" amps. I would not mess with it.

I don't think there is any large difference between most 6V6 and the 6CZ5. The CZ is clearly traced over a 6V6's role, but in a smaller bottle with the improved materials used in TV-tube production. Bias and drive will be within a few percent either tube.