good cheese

Most LDOs have an inherently high output impedance (PNP collector or PFET drain) and some fixed internal compensation. So the capacitance and ESR of the load is entangled in the loop compensation.

I don't know why nobody seems to use the LM8261-type "c-load" compensation. It has one dominant pole, and that's hung on the output pin, Miller-style, so adding load C doesn't add a separate pole to the loop. Any C with any ESR is stable.

But it's stable and works just fine. And if there's smething wrong, why don't you just say what, instead of hinting and dodging and clucking?

At your age, you've never thought about this before?

That's what most people can buy. You seem to only design linear ICs.

Cool. I'll check back every six months or so and see if there's any progress. Sounds like we have a turtle race going on here with Sloman's oscillator.

Perfection is shipping rev A on schedule, with no board spins, and getting paid for it.

Stop clucking and do some electronics. You'll feel better.

Except at low delta-V, when the output devices go from current sources to ohmic, which entirely changes the dynamics.

Come on, Jim, quit clucking and discuss electronics.

Done. Look for it. ...Jim Thompson

That *is* perfection. Perfection++, considering how vague the customer was about what they needed. We had to architect stuff we knew that they would need, even though they didn't know. They even objected to stuff that it turned out that they really needed.

And the first etched boards *are* rev A. And they all work.

I don't know why you are always whining about the electronics that I design. It's actually pretty good, and it works, and people buy it.

They hired us to do the electronics because we're better at electronics than they are. We're working with different groups, and our box pulls all the parts together. So we see some things, and anticipate some things, that they don't. Why are you so prissy about this? Big systems have complexities. This ain't a couple of 555s.

The first released files are rev A. The first bare and built boards are rev A. If we ECO a board, it's still rev A. We've had that debate in s.e.d. before.

Manipulation? Like, getting simple things like multipliers and power plugs right?

I don't care much about respect. I care about getting electronics right. Good electronics carries its own message.

Offer respect? By pretending that you are right, when you go out of your way to be wrong? By admitting that my LDO oscillates, when it doesn't?

You mean, like, folks should be nice to each other? What a concept.

Pretty novel around here, if at all sincere. :(

Cheers

Phil Hobbs

I'm occasionally wrong, but not often, because I avoid making definitive statements that I'm not sure of, and because I look things up if I'm not 100% sure. Being careful becomes second nature to people who design complex systems. The more complex, the less margin there is for error in any part.

We do tend to be less careful about things that can be fixed with firmware, or with parts value changes. We sometimes include parts/jumper options just in case. Probably one reason JT is so compulsive about parts values is that he can't change his; we can.

This is a public forum. Don't expect to have your feelings coddled. You yourself said that anything is fair game, when you were defending JT's gross insults and lies about my wife. You said it would be fair game if I gloated over his son's death... which I chose not to do. Now you whine because I'm not being nice to you.

Try to get stuff right if you don't like being corrected.

I'm most always friendly and helpful to people who are sincere about discussing electronics, at any level. But wrong is wrong, and it tends to show up when posted in public. You and JT enter objective technical threads with personal insults. And you are both so personally involved that it distorts your ability to think clearly. That's bad for engineering.

Falsifying data should be a capital offense... should it not? ...Jim Thompson

-- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

LOL :)

Jamie

Not except in cases like nuclear weapons and climate science, where a lot is at stake. ;)

Cheers

Phil Hobbs

Or "The lot is at the stake" ? :)

Jamie

Isn't there enough offensive data falsified at the capital?

The usual pattern is,

You say something that's factually wrong

I correct it "without rancor"

You escalate

You're still wrong

Sometimes you admit it, especially if someone else makes the correct argument.

Electronics, especially the theory part, isn't your skill set, and you often say stuff that's wrong, and you don't seem to like having it pointed out. You could avoid that easily.

Nobody is "in charge" of truth.

And my LDO does not oscillate.