UTC CG-52AX - use as plate chokes?

[rongon]

I have a question to throw out there.

I have a battered-looking pair of UTC CG-52AX plate to push-pull grids interstage xfmrs I'd like to use for something. I haven't been able to find a data sheet for these.

From Googling, I found they're only rated for a freq range of 40 to 10k Hz, so I don't think I'd use them for their intended purpose. But... What about using them as plate chokes for a line stage or an input/voltage amplifier stage?

I measured their B to P inductance (primary) and it came out to ~ 22H. I forget the DCR, but I can add that when I get home.

The inductance from G to G on the secondaries measures about 12H. But I imagine the primary (plate) winding is a lot beefier than the secondary windings (P-P grids). I think I'd just use the plate (primary) winding as a plate load.

So, with only 22H of L, how would I figure out what tubes would be appropriate to use these with? Put another way, how does one figure the correct L for a given tube at a given set of operating points?

(I looked in M. Jones' "Valve Amplifiers" and the RDH4, but found only very short passages on this subject. It's likely I didn't understand what was there, but I wasn't able to glean an answer from those sources.)

I'm hoping I can use these CG-51AX with 6DJ8 or 5687, as I have boxes and boxes of those tubes. A 5687 with about 15mA and plate chokes might make a fun linestage. That should put the 5687 in a spot where its rp would be ~2.5k Ohm or so.

Any advice? Maybe a link to a proper tutorial? Or a book I should look at?

Thanks...
--

[kavermei]

The reactance X_L of the primary inductance should be considerably higher than the tube's rp (dynamic plate resistance) at all frequencies that you want to amplify. This reactance is equal to X_L = 2*pi*f * L, this increases linearly with frequency so basically it sets a lower frequency limit. At the frequency where X_L = rp, only half the signal voltage will stand over the primary winding, so that will be your -3dB point.

5687 @180V has an rp of 2000 ohms; with 22H the -3dB point will be 13 Hz

[rongon]

Thank you kavermei!

-3dB at 13Hz is just barely low enough to justify giving this a try. It wouldn't make really low bass, but it should be good with bookshelf speakers, etc. Maybe a paralleled 5687 or 6N6P?

reactance is equal to X_L = 2*pi*f * L

Let me see if I'm doing this right.

If rp = 1.4k Ohm for a given tube used with CG-52AX as a plate choke, then I should see this:

6.2832 * 10 * 22 = 1382.3

Does that mean ~10 Hz is the -3dB point if the tube's rp = ~1400 Ohms?

If 6.2832 * 14.5 * 22 = 2004.34, does that mean the -3dB down point for rp = 2000 Ohms is ~14.5 Hz?

The DCR of the primary winding is 1k Ohm, while the L = 22H. Is there any way to make a ballpark guess as to its current capacity? How to test this?

--

[kavermei]

Originally Posted by rongon

Thank you kavermei!

-3dB at 13Hz is just barely low enough to justify giving this a try. It wouldn't make really low bass, but it should be good with bookshelf speakers, etc. Maybe a paralleled 5687 or 6N6P?

Yes, parallelling will cut rp in half thus also cut the -3dB point in half. Check out datasheet of 5687, the rp for this tube is quite different at different plate supply voltages, that's also an opportunity to lower rp.

Originally Posted by rongon

Let me see if I'm doing this right.

If rp = 1.4k Ohm for a given tube used with CG-52AX as a plate choke, then I should see this:

6.2832 * 10 * 22 = 1382.3

Does that mean ~10 Hz is the -3dB point if the tube's rp = ~1400 Ohms?

If 6.2832 * 14.5 * 22 = 2004.34, does that mean the -3dB down point for rp = 2000 Ohms is ~14.5 Hz?

Yes and yes

Originally Posted by rongon

The DCR of the primary winding is 1k Ohm, while the L = 22H.

Actually, when used as a plate choke you are in luck and the voltage that falls over the DCR contributes to the output signal which you take from the plate.
So to account for the DCR you can actually add 1kohms to the left part of the equations above, which will lower the -3dB point!! (didn't think of that in my previous reply...)

(To see why this is, consider that the AC signal sees a voltage divider made up of 3 resistances: X_L, DCR, and rp. Since the output signal is taken from the plate (at the junction of DCR and rp, we can lump X_L and DCR together into a single resistance, so, we add them up. An interstage transformer has no such luck, since the useful voltage at the secondary is dependent only on the voltage over X_L, there we have to add DCR to the rp, and the -3dB point is increased).

So accounting for DCR we actually have a -3dB point for rp=2000ohms of 7.23 Hz (6.2832 * 7.23 * 22 + 1000 = 1999)

Originally Posted by rongon

Is there any way to make a ballpark guess as to its current capacity? How to test this?

Note: could someone else double-check this?

--

That depends on the width of the air gap, which was chosen at design time. If this was an interstage used to drive power tubes, it will most probably work also for a line stage!

It can actually be ballpark tested, but that gets pretty involved...