On Friday, 5 September 2014 08:52:01 UTC+10, Maynard A. Philbrook Jr. wrot e:
tion notes AN45, AN49, AN51, AN55, AN61, and AN65, Baxandall's Class-D osci llator proved to be the most popular circuit that Jim Williams ever publici sed, and drove the cold-cathode back-lights on a whole generation of portab le computers.
rcuit you posted a while back of a power supply which was overly complicate d and based on the Baxandall
Do post your less complicated and more effective solution to whatever the p roblem was.
If you had done so at the time, I'd certainly have remembered it - it would have been extremely unexpected.
On Friday, 5 September 2014 09:50:02 UTC+10, Maynard A. Philbrook Jr. wrot e:
e circuit you posted a while back of a power supply which was overly compli cated and based on the Baxandall
he problem was.
ould have been extremely unexpected.
fixing it. Wave forms very bad looking, it's a wonder it even oscillate.
etter results.
w and then when I see it in my JUNK folder.
That was probably the thread "An imaginary variable-voltage photomultipler power supply" posted here on the 9th March 2014. You didn't comment at the time, and neither did Klaus Kragelund.
I pulled the LTSpice file and re-ran it - which took more than 10 minutes o n my computer. The switching waveforms do show ringing - which isn't uncomm on on switching regulators. The circuit doesn't show how the driving wavefo rms are going to be generated, though I did speculate about that when I pos ted the LTSpice file, and I did mention that "the simulated ripple is likel y to be worse than the real ripple, because both skin effect and the finite conductivity of the ferrite core are likely to kill the ripple quite a lot faster than it dies away in the simulation."
Nice to see that you were interested enough to keep a copy - no surprise th at you couldn't see the point of the circuit.
I can't recall seeing anything here that would justify "Must simpler circui ts have been posted here many times generating much better results."
If you want a switched mode variable voltage photomultiplier supply, mine l ooks pretty minimal. Which component would you take away? And why?
In modern construction, those 'concrete' columns would either enclose steel beams (concrete is for fire protection), or are 'ductile concrete', which are intended to crack, and bend, but NOT crumble. Grout the cracks, and it's good to hold the building up for another few decades.
real life. It's like the papers from IEEE, novel this, novel that. Almost all crap ;-)
It's entertaining that not only Klaus Kraglund but also Jamie have never co me across Linear Technology's application notes AN45, AN49, AN51, AN55, AN
61 and AN65, which are all about the Baxandall Class-D oscillator, not that Jim Williams ever seems to have found this out. Far from "No company would ever use that in real life" a whole bunch of companies used it it to drive the cold-cathode back lights in a bunch of different lap-top computers.
Admittedly, it takes specialised knowledge to be aware of this, rather than just common sense, so Jamie's failure to appreciate this isn't exactly sur prising.
in real life. It's like the papers from IEEE, novel this, novel that. Almos t all crap ;-)
come across Linear Technology's application notes AN45, AN49, AN51, AN55, AN61 and AN65, which are all about the Baxandall Class-D oscillator, not th at Jim Williams ever seems to have found this out. Far from "No company wou ld ever use that in real life" a whole bunch of companies used it it to dri ve the cold-cathode back lights in a bunch of different lap-top computers.
I have read the app notes and I also pointed out to you earlier in the thre ad that I was not commenting on the Williams design, but rarther the design you posted of an overly complicated multi inductor design that AFAIR could be solved with just one inductor
t in real life. It's like the papers from IEEE, novel this, novel that. Alm ost all crap ;-)
.
r come across Linear Technology's application notes AN45, AN49, AN51, AN55 , AN61 and AN65, which are all about the Baxandall Class-D oscillator, not that Jim Williams ever seems to have found this out. Far from "No company w ould ever use that in real life" a whole bunch of companies used it it to d rive the cold-cathode back lights in a bunch of different lap-top computers .
read that I was not commenting on the Williams design, but rather the desig n you posted of an overly complicated multi-inductor design that AFAIR coul d be solved with just one inductor.
What you posted was "The Baxandall is overly complicated". If you don't lik e the circuit I posted, identify it.
I posted possible candidates on the 8th February and the 9th March. I very much doubt that a single inductor circuit could do the job I had in mind, b ut I'd love to see your proposal - though I suspect you've merely failed to get your head around the problems that I was thinking about solving.
Baxandall's class-D oscillator does use two inductors. I've got a lower dis tortion - but rather less efficient - variant that replaces the feed induct or with a current mirror and a rather messy control loop, which did go into production in 1988, but in very small volume.
hat in real life. It's like the papers from IEEE, novel this, novel that. A lmost all crap ;-)
is.
ver come across Linear Technology's application notes AN45, AN49, AN51, AN
55, AN61 and AN65, which are all about the Baxandall Class-D oscillator, no t that Jim Williams ever seems to have found this out. Far from "No company would ever use that in real life" a whole bunch of companies used it it to drive the cold-cathode back lights in a bunch of different lap-top compute rs.
thread that I was not commenting on the Williams design, but rather the des ign you posted of an overly complicated multi-inductor design that AFAIR co uld be solved with just one inductor.
ike the circuit I posted, identify it.
y much doubt that a single inductor circuit could do the job I had in mind, but I'd love to see your proposal - though I suspect you've merely failed to get your head around the problems that I was thinking about solving.
istortion - but rather less efficient - variant that replaces the feed indu ctor with a current mirror and a rather messy control loop, which did go in to production in 1988, but in very small volume.
This one:
formatting link
sion
5 inductors and a transformer with 5 individual windings and +30 components to generate a single output voltage
that in real life. It's like the papers from IEEE, novel this, novel that. Almost all crap ;-)
t is.
never come across Linear Technology's application notes AN45, AN49, AN51, AN55, AN61 and AN65, which are all about the Baxandall Class-D oscillator, not that Jim Williams ever seems to have found this out. Far from "No compa ny would ever use that in real life" a whole bunch of companies used it it to drive the cold-cathode back lights in a bunch of different lap-top compu ters.
e thread that I was not commenting on the Williams design, but rather the d esign you posted of an overly complicated multi-inductor design that AFAIR could be solved with just one inductor.
like the circuit I posted, identify it.
ery much doubt that a single inductor circuit could do the job I had in min d, but I'd love to see your proposal - though I suspect you've merely faile d to get your head around the problems that I was thinking about solving.
distortion - but rather less efficient - variant that replaces the feed in ductor with a current mirror and a rather messy control loop, which did go into production in 1988, but in very small volume.
that in real life. It's like the papers from IEEE, novel this, novel that. Almost all crap ;-)
t is.
never come across Linear Technology's application notes AN45, AN49, AN51, AN55, AN61 and AN65, which are all about the Baxandall Class-D oscillator, not that Jim Williams ever seems to have found this out. Far from "No compa ny would ever use that in real life" a whole bunch of companies used it it to drive the cold-cathode back lights in a bunch of different lap-top compu ters.
e thread that I was not commenting on the Williams design, but rather the d esign you posted of an overly complicated multi-inductor design that AFAIR could be solved with just one inductor.
like the circuit I posted, identify it.
ery much doubt that a single inductor circuit could do the job I had in min d, but I'd love to see your proposal - though I suspect you've merely faile d to get your head around the problems that I was thinking about solving.
distortion - but rather less efficient - variant that replaces the feed in ductor with a current mirror and a rather messy control loop, which did go into production in 1988, but in very small volume.
ussion
ts to generate a single output voltage.
Two of those inductors - L6 and L7 - are ferrite beads. L8 is a off-the-she lf part for the output filter. L10 is the Baxandall feed inductor - it woul dn't have been a Baxandall circuit without it.
Similarly the five windings on the transformer are - in fact - two bifilar windings and the output winding, and absolutely standard. Every last one o f Jim Williams AN45, AN49, AN51, AN55, AN61 and AN65 circuits contain essen tially the same transformer and the feed inductor. Since that circuit was p opular driving for lap-top cold-cathode back-lights, it seems that you can' t appreciate a good solution when you see it.
Your fifth inductor - L9 - doesn't exist. It probably did at some stage whe n I was playing with the circuit, but now serves only to show that you didn 't look at the circuit diagram very carefully.
The base drive arrangements are a mess - they do produce a schematic that c an simulate. If I wanted to build a real circuit, I wouldn't build it that way.
As I posted at the time "In real life I'd use some sort of digital driving scheme to get non-overlapping drives for M1 and M2."
In other words, you weren't paying attention. We wait for your superior one
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.