Invisus ultra super regulator

[chalky]

I think they may be talking about something like this http://www.wenzel.com/documents/finesse.html or similar

[Terry Demol]

Originally posted by chalky
I think they may be talking about something like this http://www.wenzel.com/documents/finesse.html or similar

I don't think it is a copy of Charlie Wenzels finnesse reg,
the OP Z is too low.

But why do I keep getting the feeling it is based on a
negative impedance generator, sort of like the
Hawksford OP error correction scheme.

Very low OP Z, very good BW, very good stability unlike
opamp based designs that use lots of OLG to get low OP
Z.

I can feel a new reg coming on.

Cheers,

Terry

[peranders]

The pcb is a multilayer board with gilt (thin gold layer) pads, AD797, IRL3502, ADR425(?). It's a mystery why they have chosen a 20 volt mosfet
OS-CON as input and output caps. Black gate as LP-filter for the voltage reference and for the feedback. Very small melf resistors and 0402. 42 x 54 mm, very cute.

The topology seems to be a conventional Sulzer regulator.

[carlosfm]

Originally posted by janneman
You smell another bussiness opportunity, Per? Best I keep my copy to myself...

Jan Didden

Zing!

[peranders]

Originally posted by carlosfm


Zing!

PPR1 (positive) and NPR1* (negative) utilise a high speed and ultra low output impedance FET and guarantees peak current demands are instantly and effortlessly met. This coupled with our propriety self noise cancelling system give Invisus DC regulators outstanding resilience, reliability and performance.

It's always interesting what's behind the marketing words.

They have mixed Rdson with output impedance, for instance. Rdson is only 7 mohms but only interesting if it is a switch application.

"propriety self noise cancelling system" is a lowpass filter.... R+C

[peranders]

I change my mind, it's not a Sulzer regulator, it's a Jung Super regulator but not with the enhanced emitter follower. The design is more or less identical with the article Jung wrote
http://home.comcast.net/~walt-jung/w...g_Circuits.pdf

but since they use a mosfet the driving is different. Resistor between drain and gate and from gate a 6.8 V zener down to the opamp output.

Voltage reference is ADR421 (2.5 volts), which seems to be a good choice according to datasheet.

All and all I think it looks good but I'll doubt it is as good as Walt Jung's enhanced model with class A emitter follower and fast bipolar transistors.

A fat OS-CON 330 uF is at the output and I think we come to the conclusion that big and good elelctrolytics didn't do any "more" good, more like harm in some cases.

I wonder if a Black Gate as LP-filter for DC is for making the BG crowd pleased?

[peranders]

Pretty nice looking, don't you think?

[peranders]

Backside or front, depending how you see it.

[jackinnj]

where's the "additional" compensation for the AD797 (per Analog Devices datasheet) ? -- as we know this is a tricky opamp to work with. In issue 4/2000 of Audio Amateur Walt Jung recommended the AD825 or AD817 insead of the AD797.

I recall that in the original Audio Amateur articles there was a sidebar conversation by Galo: "So It Oscillates, Now What ?" (4/95)

I do hope, for the sake of general knowledge, that the people at Old Colony Sound make a CD-ROM for the 1980's and 1990's -- the ARRL folks do this for QST and QEX, as well as Circuit Cellar

[peranders]

The regulator oscillates at 4 MHz, 20 mV peak, if you attach 100 nF right at the output pins and load it with some current 750 mA in my case.

The thing is that the oscillation goes away with on 10+10 mm wire so this is hardly an issue.

The AD797 seems to run without any special treatment. Don't forget that the pcb is 4-layer board so you will have pretty good ground.

AD825 seems to be a better choice and the noise will get slightly higher. I have measured around 5 uV with AD825 which I think is pretty good and by that you'll will have a more "pure" performance, phasewise.

BTW: The Black Gate is 100 nF which seems to be a mystery to choose instead of someting good, non-electrolythic.