Composite amps

Compund, not composite. Thanks.

What is this "predistortion" discussed in the article but itself a form of "regular" negative feedback?

It says "In essence, 'inverse distortion' is introduced into the input of the amplifier, thereby cancelling any non-linearity the amplifier might have." Yes, that's what negative feedback is.

Except that predistortion is not negative feedback; all the signals are going in the same direction. It's hard to do feedback on a GHz power amp.

I broke the feedback loop with an RC lowpass. (int-amp reading the R, there was at least one gain adjustment in each leg.) But I don't think that's what Phil is asking about.

George H.

Right, I've never done that, but as long as one can be tuned down to match the slowest (time-wise).

George H.

ch

folks

ed

f!

the

e. Have you peaked at the in between opamp signal? GH

ch

folks

ed

f!

the

e.

Oh you've got some time delay front to back, maybe feed back around the 1st opamp... an RC in series? at some intermediate freq./time. George H. Inside cooling off, I'm grilling more meat later. Hey a shoutout to any vet's on Memorial Day. Thanks! (I should give some buddies a call.)

GH

I don't have time (*) to join the discussion, but read this...

(*) Pursuing a new trick for behavioral modeling ;-) ...Jim Thompson

I have that one, thanks. Lots of the circuits are summing-junction snoopers, i.e. circuits wih a slow-but-accurate amp looking at the time-averaged input error of a fast-but-cruder amp, and nulling it out. That's a useful trick sometimes, and are examples of "putting a bandaid on the fast circuit", which I was talking about upthread.

There are other sorts of bandaids, e.g. the White cathode follower and many sorts of local feedback. Often the key is to figure out a way that the bandaid can be much slower than the main amplifier, as in the snooper circuits.

I often use op amps to force JFETs to run at exactly I_DSS, for instance--the problem then is to keep the low-frequency noise from going nuts.

Cheers

Phil Hobbs

It's generally well behaved until the output amp rails, the reason being that the input amp is the slow one.

I've been playing with a CPH3910/ADA4899/THS3091 combination that is starting to look reasonable in the spherical-cow universe. It's a little gain-of-50 amp with about 120 MHz bandwidth, Zin of 10M //

Overcast here--has been all weekend. :(

Cheers

Phil Hobbs

I had a look at the article, and it does have some points of interest, thanks. The author distinguishes between his technique and feedforward, but I'd argue that for circuit purposes it's a species of feedforward, whose basic principle is to sense just the error of an amplifier and apply the appropriate correction to the _output_ and not the _input_.

I didn't know that FF was invented by Harold Black.

Cheers

Phil Hobbs

Ah, helpfully the Wiki seems to never mention that it's a feed-forward and not feedback topology.

Have you tried designing a composite-amp oscillator instead? If my theory is correct it should make a wonderful amplifier.

I'm taking today to remember my Dad, who passed about two and a half weeks ago just shy of 92. S/Sgt 10th Mountain, served in Italy late 1944 to 1946.

Thanks, Dad!

And also sort through all his stuff, he was of the generation that never really liked to throw anything away that looked like it would come in handy someday. Sigh...

th

much

u folks

ave

ited

elf!

nd the

at

of

onse.

I'm sorry for your loss. My dad was part of the 'greatest' generation too, but he passed away in his 70's. Years ago. He served in the South Pacific .

George H.

I'm the youngest of four kids, I got more time with him than most guys whose fathers were in their early 50s when they were born. He remained pretty much independent until he died but required a few hours a week of help from me and a nurse the past few years that I was glad to assist with, we were still going out for lunch about once a week until last month.

About a week before he died he'd got some test results back that didn't look too good. He was rarely a pessimist, about the closest he came was that day when he looked very tired and said that he missed his friends, all the people who he used to know who by now were all gone.

I could understand that, one son who visits regularly is nice but can hardly make up for all the rest. He became very ill the next week and passed away about 48 hours after arrival at the hospital, just enough time for the rest of the family to make it to be there with us.

It went about as "well" as it could and at that age and I'd known for quite a while the day would come sooner or later, but I'm still thinking it's him when the phone rings. It's all still very strange. :)

And thank you! Dad was kind of hit-or-miss as a father (like most fathers I suppose) but it was remarkable hearing the stories about the time he came from which was both so different and so much the same as th time I grew up in, I am happy to say he was a man I was proud to have had a chance to know.

ps

with

s much

you folks

ent

DC

ne.

ng

an

have

ses

th,

efited

yself!

uch

and the

e at

e of

sponse.

er

too,

ific.

Sure, my dad died one night, massive heart attack, he was set to go hunting the next day with buddies. It was totally unexpected from my point of view, but we learned after the fact (from his girlfriend) that he wasn't taking his high blood pressure meds. Some type of advanced warning is mostly a good thing, let's you sorta prepare... emotionally.

George H.