Unusual power supply, please explain

[nashvillebill]

I recently acquired a couple of pieces from an ancient Ampex broadcast video tape recorder, including the tape motor power supply section. It used a quartet of 6550's (alas one of them was broken, but the other 3 Sylvanias were good...but I digress). The power supply is massive, must weigh 75 lbs. There's a very large PT (700VAC, CT on secondaries A4 and A8 if I am reading the rusty printing on the transformer right, 575 mA). Feeds the bridge rectifier off A4 and A8. B+ goes into a choke L001A--but notice they have B- going into another section of the choke, which is sharing the same core. Yes it's a heavy choke. Seems unusual to have B- going into the other half of the same choke.

What's also weird for me is that the CT (A6 on the PT) appears to be tied to the screen supply. Also the way they've got the caps after the chokes, that does not make sense to me.

Any explanations would be welcome. I'm not going to use the power supply as it was, I've already pulled the PT and chokes and OT's off and will either use them for a bench power supply or maybe eventually a massive amp. But the way they did the choke, screen supply and the way they divided the smoothing caps? Not immediately obvious. Thanks!

[bobshbob1]

Looks kind of like the power supply wiring in this JCM800 schematic; the center tap is connected between the first series filter caps. I'm not sure what the reasoning behind this would be vs. leaving it unconnected, maybe someone else will shed light on that, but it gives the same result voltage-wise. In your schematic the screens are connected to that midpoint too, so the plates are at full B+, and the screens and everything down the line connects to about half of B+.

Marshall JCM800 2203.pdf

Not sure about the two-part choke, maybe it just stiffens up the power supply that much more...?

[exclamationmark]

The two part choke is known as a common mode choke, which dampens any common mode noise which may be riding on either voltage rail. Line filters or power conditioners usually have them built in, though usually on the AC side (and more for HF noise). I suspect this kills birds with one stone, performing the role of common mode noise reduction plus PI filter duties.

[Gingertube]

This power supply configuration is not that unusual.
It is an easy way to get a main B+ supply and a 1/2 B+ supply for the output tube screens.

Here is an example from downunder:
http://www.ozvalveamps.org/goldenton...wbm1963rdw.gif

ASIDE: I have a 60W Bassmaster Goldentone. The 6DQ6 Output tubes with screens at 1/2 the main B+ compress beautifully. I really haven't heard that degree of power amp compression in any other amp.

Cheers,
Ian

[nashvillebill]

Thanks!

[Dave H]

Looks kind of like the power supply wiring in this JCM800 schematic; the center tap is connected between the first series filter caps. I'm not sure what the reasoning behind this would be vs. leaving it unconnected,...

Connecting the centre tap to the junction of the two caps ensures that it will be at half voltage for all load conditions. If the centre tap is not connected the voltage at the cap junction will depend on the loading across the two caps. i.e. It will only be half voltage when the loading is equal.

[Tooboob]

It seems to me that common mode chokes could be very useful in a hand-wired amp. If you connect a supply and it's ground through a common mode choke, the AC part of the current provided by the supply is returned through the choke (only opposite and equal currents see a low impedance through the choke) rather than taking a path through some other sensitive circuit. It would let you steer your return currents back to the supply cap instead of out of some signal connector ground on the way to another piece of equipment's chassis ground. I've used them for this on wall-wart supply inputs to high-frequency equipment to reduce system emissions and reduce spikes from the supply too, along with the usual solutions. Works.

It's kind of like splitting ground planes in a PC board. It can work magic, unless you do it wrong, which makes things worse. You may need to do some figuring. Since the AC load is a comparatively low impedance cap (or should be), and the flux cancels in the core for balanced current, you don't need a big core or lots of winds - just wire thick enough to handle the current.

That coulomb of electrons that flows through a 1A circuit every second would be hard to count, but as long as as many leave as enter, and both paths have the same winding structure, the choke is low impedance, with little flux in the core.

Of course, you need better insulation at high voltage, and trying to block audio frequency common mode noise make the required part a bit or substantially bigger.