MagneQuest/Peerless Forum

RE: Suggested OPT for GM70?

76.22.120.252

In Reply to: RE: Suggested OPT for GM70? posted by arend-jan on June 11, 2008 at 07:44:00

I worked with Mike when he was developing of the FS-100PF, so I still have an interest in it. (At the time, we - i.e. Bottlehead R&D - thought we were going to try a serious assault on the 845 state of the art. Directions changed, but it was still a technically rewarding project for me!)
The transformer was derived from Mike's FS-100 series feed design. Presumably that one is still in the archives, though you might have to twist Mike's arm pretty hard to get some made. :^) It's quite a good design in its own right. If you want to do series feed without capacitors, all you need is a good shunt regulator at the end of the power supply - see the late John Camille's articles in Sound Practices for more on this. Combined with fixed bias, the signal current loop includes only the shunt regulator, tube, and load - no capacitors. (Note that the same analysis applies to a transformer-coupled driver.) The shunt regulator must carry at least as much current as the tube it is regulating, so it can create at least twice the heat. This inefficiency, along with the large transformer size required to carry both AC and DC, are the arguments against series feed - as Mike says, if it is done right it can sound as good as anything else done right.
Parallel feed does two things. First, it isolates the power supply from the signal current loop. This isolation is through the plate choke inductance, so it is most effective at mid and high frequencies; we have found that a shunt regulated power supply still improves the sonics, but the regulator does not need to carry the full current, and even with no regulator the sound has most of the benefits of a shunt regulated series feed arrangement.
Second, parallel feed allows the transformer to be smaller since it does not carry DC. The FS-100PF has a shorter stack even though its power handling capability is substantially larger than the FS-100 series feed. The shorter stack reduces parasitic capacitance and leakage inductance; between us Mike and I came up with a revised winding schedule that took full advantage of this. This included some design principles and strategies we acquired by studying Mike's Peerless archives. I learned a huge amount from this exercise, and I will be forever grateful to Mike for giving me the opportunity to participate in this adventure.
Parallel feed does have a capacitor in the signal current loop. For a top-line setup with a 10K load, I would hope for 100 henries of plate choke, calling for about a 2uF capacitor, rated for something like 1.5 times the plate voltage. At GM-70 voltages, that's quite a large capacitor, but it is much smaller than a power supply capacitor and hence one might be able to get a better performing unit.
I know that was a pretty long and complex post, I hope it is helpful. I'll be out of town for the next week or two, but happy to continue this discussion when I get back if anyone is interested.

Edits: 06/12/08

Follow Ups:

Post a Followup:

FAQ

Post a Message!

Forgot Password?

Moniker (Username):
Password (Optional):
	Remember my Moniker & Password (What's this?) Eat Me
E-Mail (Optional):
Subject:

Message:	(Posts are subject to Content Rules)

Optional Link URL:
Optional Link Title:
Optional Image URL:
Upload Image:
E-mail Replies:	Automagically notify you when someone responds.

To view your new posting or follow-up, click on the RELOAD or REFRESH button on your browser.