Super Regulator, collecting the facts

[jackinnj]

Well, here's a much soldered over proto-type -- this one omits the clamp diodes on the AD825's -- I did get a wonky LM329DZ which wouldn't allow the SMT version to start-up -- replaced it and now works fine. The positive regulator has the LM317 pre-regulator, while I took the negative pre out for testing purposes. In the final version I think that I can save about 33% of the real estate by scrunching things up.

This one uses R1206 1K resistors for the error sense -- in the next go around they will be R2010.

Three are no bypass caps in either of these, no oscillation problems at all. I do use 499R to decouple the inverting input of the opamp.

Oh, and you can see where there was an error on the PCB -- had to patch up with some wire-wrap wire.

Edit -- the high current traces pass along the edges of the board. The grounds all intersect in the center.

[jackinnj]

Here's a comparison of the straight AD797AN SuperRegulator with a Pre-Regulated Version -- using an LM337 -- I guess it's pretty convincing -- I still want to do this with the HP3403C for 10Hz to 100kHz. FWIW when I repeated the AD797 tests the values still come up the same so there isn't some environmental wonkiness going on.

[peranders]

A tube based sort of super regulator, an article by Mr. Bob Pease.

http://www.elecdesign.com/Articles/I...ArticleID=1509

I think it look like an super regulator if the basic concept is considered as an voltage reference powered from the output and an error amp also powered from the output.

[Khron]

Mr jackinnj , would you mind if I asked you to post/attach the schematic / PCB / silkscreen for the through-hole version of the super-reg?

(For DIY purposes - powering some active crossovers)

[peranders]

Jack, some time has gone now, any conclusions? Does the board work alright?

My all SMD regulator works pretty good. The difference is that I used a 1 A SOT223 transistor.

[jackinnj]

Originally posted by Khron
Mr jackinnj , would you mind if I asked you to post/attach the schematic / PCB / silkscreen for the through-hole version of the super-reg?

(For DIY purposes - powering some active crossovers)

The boards are available from AudioXpress' webstore -- Old Colony Sound -- for $24.95. Here's a link: http://www.audioxpress.com/bksprods/PCBCTB.htm
For the noise testing I used the Old Colony boards and through-hole parts.

PA -- looks good -- I found that the R2010 resistors were best for the error sense circuit -- I had used R0805 originally. Why not solder in an AD797AR and see how it compares with the AD825 -- and in spite of admonitions against, I found that the LT11115 worked quite well on the Old Colony board.

My smt board works very well -- it incorporates the LM317 pre-regulation -- I have it working in the modded Crown IC150a preamp.

[Khron]

Well, with all due respect, unless you have some sort of financial aggreement with AudioXpress, i'd much rather etch and drill the pcb's myself

Plus, that's quite pricey for a few sq.in of circuit board, at least for my pocket...

PS: regarding the opamps used, apart from the GBW, low noise, low offset, what other minimum requirements would there be? (like slew rate etc)

[jackinnj]

P-A -- how much heat is your pass transistor dissipating?

[peranders]

You have lot's of options when it comes to parts but the faster you want the regulator to be, the harder to make it work properly.

One important thing is minimum supply voltage of the opamp and also common mode properites, important at startup.

[peranders]

Originally posted by jackinnj
P-A -- how much heat is your pass transistor dissipating?

Somewhere between 500-1000 mW. Since I have a groundplane I'll get cooling also. I haven't tested the absolute limit.