|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
66.220.116.223
In Reply to: RE: Speaking of Linear Power Supplies.... posted by AbeCollins on December 17, 2014 at 18:07:19
Quite correct. HF decoupling caps only work a little, but not good enough for these high power di/dt events.
Follow Ups:
Hi,
Curious... I cannot make any active regulation circuit to have a greater Delta I / Delta T ability than a correctly implemented passive decoupling combo. Nor have I seen any regulator IC or discrete circuit that could make such a claim. At least not directly at the IC Pin's, where an 0402 or 0603 cap usually fits well and where it matters.
The trick is not to make the regulator as fast as possible (it is ALWAYS going to be too slow), but to tune the whole supply rail so that for each element in the decoupling circuit any current "too fast" is taken over by an element closer to the IC Pin supplied, of a value sufficient to deal with any current steps.
In this kind of set-up there is no benefit making the regulator fast, but low noise is always good. On the contrary, often it is better slow the regulator down (most brutally using a choke in series) so it does not fight the decoupling caps. A network analyser can help here. There is no absolute "patent recipe", one must analyse the demands a given circuit node imposes on the power supply and then design accordingly.
In the end it is easy to see if you succeeded even without network analyser. Just play a low frequency, high level square-wave and look at the supply pins with a differential probe. If there is no appreciable noise, whatever is decoupling/regulating the pin has enough Delta I / Delta T. If you see a lot of noise, usually there are insufficiently damped resonances or insufficient decoupling.
Oh and watch ground bounce, because sadly common logic ties the negative supply and ground together, silly thing that. That is why looking at PSU decoupling differential probes are a must.
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
Thorston - you can convince yourself of anything you want. I know what I hear and the faster di/dt regulator sounds better. These are not trivial to build as they are discrete and the devices needed are really hard to find.
Why do I get the impression that you know what you are talking about?
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Because he does.
It simply makes sense. I thought in layman terms we want a "stiff" power supply. When you have low noise and a supply that is as immune as possible to effects of the load... What else do we need?
Maybe Steve's regulators can read the circuits mind and react before needed... Not that would be fast!
We need good transient response and wide bandwidth, with low noise extending into hf.
Some regulators actually become oscillators when tested for transient response.
Goto the Belleson website for info and look at the data sheet on low noise regulator chips such as the AD and TI ones. Scroll down and look at the transient scope shots.
"Maybe Steve's regulators can read the circuits mind and react before needed... Not that would be fast! "
I want about 30 seconds advance notice of the market tickers. :-)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
You don't need that much time!
I believe I know how to fix this problem permanently, i.e. how to produce a provably fair trading system. Unfortunately, this is against the interest of Wall Street and they would have to be forced to use it, since the masses appear smart enough to realize the benefits. Worse, there is little prospect that government will enforce such a solution, since the politicians are owned by Wall Street. More specifically, the government would outlaw such a situation, as they already regulate securities trading.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: