current and cathode bias SE

[Wilder Amplification]

Ideal class A amps average current applies more so.

No...average current applies anytime you're dealing with AC. In a perfect ideal text book symmetrically biased Class A amplifier, average current does not change.

In a non-symmetrically biased Class A amp average current will change a little but not much. For example, say we bias our amp to 70mA instead of 75 and peak current is 150mA. Average current will only increase by a factor of 5mA when under signal.

[tubeswell]

I have a related question which is probably useful to ask in this thread. (Thanks lowell for starting it)

Since 'everyone' seems to be in agreement that Class A SE output stages idle at about 1/2 the peak current, then, for the purpose of determining sufficiency of supply choke current rating, is it generally a 'safe enough' rule of thumb to assume that an output tube should theoretically hit no more than 2 x idle current under signal conditions?

The reason I ask is I have a 6L6 loaded with 4k Pr Z idling at 332V on the plate, g2 = 288, Rk (236R) = 20.6, (i.e.: tube current ~43mA) and the whole amp is supplied by a CLC filter with a 100mA 3.5H choke. But when I draw a load line with these conditions I get 75mA idle current, (which suggests 150mA peak current, however this doesn't seem to be borne out by the actual measured tube current under idle conditions? - granted that I'm not hitting the plate with a very high B+).

Or does the charging-discharging cycle of the input cap and the smoothing cap in a CLC filter reduce the need to worry about peak current through the choke?

[MWJB]

I very much doubt that your 6L6 is idling at 43mA with a 236R cathode resistor. It's not what your own figures suggest, 20.6v on the cathode of a 236ohm resistor = an estimated 87mA at idle.

[guitarmike2107]

My guess is that the 100mA is a DC rating, not a peak rating which is what you are worrying about.
Anyway depending on what you have in your pre amp you appear to be operating near that.

[tubeswell]

I very much doubt that your 6L6 is idling at 43mA with a 236R cathode resistor. It's not what your own figures suggest, 20.6v on the cathode of a 236ohm resistor = an estimated 87mA at idle.

Duh yeah sorry my typo - should be 20.6 V on 472R resistor - it is definitely idling at 43mA. So is the choke safe or ?

[tubeswell]

Yep it is the DC rating.

[MWJB]

Yes, I reckon so, your idle current will rise some under signal, but won't be averaging anything like 100mA, you could measure current accross a 1ohm resistor whilst you play if concerned? 43mA at idle, allow half as much again at the outside, plus 10mA for the screen half a dozen mA for the preamp tube...and you're still well under 100mA.

[tubeswell]

Thanks for that assurance MWJB.

[Wombaticus]

No...average current applies anytime you're dealing with AC. In a perfect ideal text book symmetrically biased Class A amplifier, average current does not change.

Average current may not change in a symmetrical SE amp, but rms current does. At full pre-clipping signal, Irms = Iq*sqrt(3/2) ~1.22 which is probably the more significant consideration from a design point of view, and from the point of view of any iron expected to deliver that current.

This result is not mine, it was derived by Mike Knauf awhile back, but seems to be confirmed by both sims and real world measurement. You can check Mike's math for yourself here:

ClassA_Current_Draw.pdf

[Wilder Amplification]

Average current may not change in a symmetrical SE amp, but rms current does. At full pre-clipping signal, Irms = Iq*sqrt(3/2) ~1.22 which is probably the more significant consideration from a design point of view, and from the point of view of any iron expected to deliver that current.

This result is not mine, it was derived by Mike Knauf awhile back, but seems to be confirmed by both sims and real world measurement. You can check Mike's math for yourself here:

ClassA_Current_Draw.pdf

Correct. However, the OP happened to read something that stated that "current draw does not change in Class A amplifiers", and either they left out the fact that it was "average current" that they were specifically referring to that doesn't change and/or the OP just failed to catch that part of what he read not knowing what "average current" was.

[Wombaticus]

Correct. However, the OP happened to read something that stated that "current draw does not change in Class A amplifiers", and either they left out the fact that it was "average current" that they were specifically referring to that doesn't change and/or the OP just failed to catch that part of what he read not knowing what "average current" was.

OK - none of which changes the fact that in the truly important way of thinking about it (the RMS current), the current draw DOES change under load, and this should be recognized for the purposes of deciding how much iron you need and how much heat will be dissipated, etc. It's worthwhile to point out that this difference in RMS current becomes even more dramatic when you are outputing a rail-to-rail square wave (yes, I realize a real pentode is not going to manage this). In this case, Irms will approach Iq*sqrt(2) ~ 1.41, a forty percent increase in current relative to quiescence, even as the average remains the same.

[tubeswell]

So are chokes really rated for rms AC current or for DC current (even tho' they may say "DC current")?

When I asked the people at Mojo about the rating for their 777 choke, the answer I got was:

"We can not provide much more information than is already posted on the Mojo website. We publish the mounting dimensions (X Mount on 2 3/4 centers) and the electrical specifications (3.5HY, 100mA, 100ohms)."

(Don't hold your breath I guess.)

[Steve Conner]

For the usual tube amp applications, the ripple in the choke is so small that the RMS and average currents are the same.

In a choke input filter, the same 1.22 derating factor would presumably hold. There may also be significant iron losses from the AC flux component, that would make it run even hotter and need further derating.

But you can't buy a current production choke designed to work in a choke input filter. They're all meant for the usual CLC arrangement. A good choke input filter really needs a swinging choke optimized for the particular application, anyway.

I have no idea whether designers of SE OPTs take the RMS-to-average ratio or iron losses into account when rating them. From looking at the average OPT in a classic Champ-style amp, it would seem they were mostly concerned with wasting enough bass energy to make the speaker last beyond the warranty period.

Screen current in a SE amp most certainly does change under load. Any non-linearity in the tube curves causes the average current to change under drive (RDH4 calls this "rectification") and the curve of screen current vs. drive is very non-linear, because the operating point goes into the region of high screen current on the negative excursions of plate voltage.

[MWJB]

Tubeswell wrote: "So are chokes really rated for rms AC current or for DC current (even tho' they may say "DC current")? When I asked the people at Mojo about the rating for their 777 choke, the answer I got was: (3.5HY, 100mA, 100ohms)."

I would assume dc mA rating unless otherwise stated, either way, I very much doubt that you have an issue 100mA dc is good, 100mA AC even better. What is the mA rating of your PT, for 2x6V6 or 1x 6L6 it may not be much over 100mA, if any over?

[tubeswell]

Tubeswell wrote: "So are chokes really rated for rms AC current or for DC current (even tho' they may say "DC current")? When I asked the people at Mojo about the rating for their 777 choke, the answer I got was: (3.5HY, 100mA, 100ohms)."

I would assume dc mA rating unless otherwise stated, either way, I very much doubt that you have an issue 100mA dc is good, 100mA AC even better. What is the mA rating of your PT, for 2x6V6 or 1x 6L6 it may not be much over 100mA, if any over?

The PT is rated at 150mA on the HT winding. (350-0-350VAC @150mA with 280-0-280 taps), which is theoretically good for 2 x 6L6s (Heater is rated at 5A) so no problems there