The circuit on your website bears a lot of resemblance to the one in Figure 2 of G3VA's Technical Topics column in Radio Communications magazine from May 2002. He refers to it as the "Mesny VHF power oscillator of the 1920s" though of course then it used triodes not MOSFETs, and needed a separate winding for the gates to get the DC bias right. There is a picture of it under the name Mesny here:
formatting link
I think some people call them Kalitrons too:
formatting link
It is one of the most popular topologies for on-chip LC oscillators in the synthesisers of cellphone radios - though usually with fancy amplitude control etc. and sometimes without the inductor in the supply, or with the supply inductor moved to the tail of the diff pair. Whatever it is called, it was a very old topology in 1959 and yet billions of them have been made in recent years.
The converters John suggested are isolated, you can stack them in series. Phil only suggested using one non-isolated, then a charge pump. So, they are not suggesting to stack the non-isolated converters.
A $1.50 boost and a $1.50 buck connected as an inverter gives the effect of stacking two of the fancier ones, using the isolation of the USB brick. Of course it may not work if the OP needs to use the 5V in the same circuit.
If I turn this hobby product into an actual product, space is also at a premium. The current requirements of the -5 volt rail are modest - a while back someone pointed out a circuit that used a MAX232 to generate a negative rail for an entire board. How would the "stack-up isolated" work?
Yes, have John ('s suggestion) sitting on top of Phip ('s suggestion). John would be totally isolated from the rest (Sorry, John, it's just a joke). A Max232 can also generate the negative rail (probably -7 to -10) then regulate it to -5V.
ctor - would work well. MOS-FET transistors do a bit better than bipolar tr ansistors, but Baxandall's paper rather pre-dates them.
illator1.htm
web-site "The circuit is probably best known from Jim Williams' series of a pplication notes for Linear Technology, on high frequency inverters for dri ving cold cathode back-lights used in laptop computers (application notes A N45, AN49, AN51, AN55, AN61, AN65)." These apparently were Linear Technolog y's most popular application notes for quite a while.
hout the Baxandall label.
rrent mirror rather than the feed inductor, but that's not the low distorti on oscillator I'm working on at the moment.
re an idiot, but you really don't have to do it quite as often as you do.
Jim William's version has the inductor in the tail of the differential pair .
No inductor gives you a rather different beast, though I've used a current mirror to do much the same job to get a rather less distorted sine wave.
"The circuit is probably best known from Jim Williams' series of applicati on notes for Linear Technology, on high frequency inverters for driving col d cathode back-lights used in laptop computers (application notes AN45, AN4
9, AN51, AN55, AN61, AN65)." These apparently were Linear Technology's most popular application notes for quite a while, so it's not surprising that t he circuit is now well-known.
Nobody has ever called it a Mesny power oscillator before anywhere I've loo ked. Peter Baxandall seems to have invented the circuit to drive high-turns ration transformers to step up +12V supplies to a kilovolt or so to drive photomulitpliers.
That's not remotely a VHF circuit, though VHF circuits are presumably rathe r more sensitive to the rather lower inter-winding capacitances seen in mor e conventional transformers.
The Proceedings of the British Institute of Electrical Engineers is a peer- reviewed journal so presumably the reviewers didn't see the similarity to t he Mesny power amplifier either.
web-site "The circuit is probably best known from Jim Williams' series of a pplication notes for Linear Technology, on high frequency inverters for dri ving cold cathode back-lights used in laptop computers (application notes A N45, AN49, AN51, AN55, AN61, AN65)." These apparently were Linear Technolog y's most popular application notes for quite a while.
hout the Baxandall label.
rrent mirror rather than the feed inductor, but that's not the low distorti on oscillator I'm working on at the moment.
re an idiot, but you really don't have to do it quite as often as you do.
Sure - you don't like being shown up as an idiot. Sadly, the response ident ified you as an idiot just as clearly as all the other rubbish you post.
tor - would work well. MOS-FET transistors do a bit better than bipolar tra nsistors, but Baxandall's paper rather pre-dates them.
llator1.htm
John Larkin is too dim to notice the difference between a low distortion os cillator and the Baxandall/Jim Williams oscillator.
"The circuit is probably best known from Jim Williams' series of applicati on notes for Linear Technology, on high frequency inverters for driving col d cathode back-lights used in laptop computers (application notes AN45, AN4
9, AN51, AN55, AN61, AN65)." These apparently were Linear Technology's most popular application notes for quite a while.
Nobody has had any trouble building them in jig time. Except perhaps John L arkin, who can't cope with transformers that he can't buy off the shelf.
uctor - would work well. MOS-FET transistors do a bit better than bipolar t ransistors, but Baxandall's paper rather pre-dates them.
cillator1.htm
oscillator and the Baxandall/Jim Williams oscillator.
Though it can hardly come as a surprise that you get teased about it since you have been talking about for years and still haven't gotten around to bu ilding it, as if a bit of table space for a soldering iron and time, for so meone who doesn't have to go to work everyday, is impossible to find
tion notes for Linear Technology, on high frequency inverters for driving c old cathode back-lights used in laptop computers (application notes AN45, A N49, AN51, AN55, AN61, AN65)." These apparently were Linear Technology's mo st popular application notes for quite a while.
Larkin, who can't cope with transformers that he can't buy off the shelf.
Aiming to build stuff from things you can get cheap and fast off the shelf is just common sense
nductor - would work well. MOS-FET transistors do a bit better than bipolar transistors, but Baxandall's paper rather pre-dates them.
oscillator1.htm
n oscillator and the Baxandall/Jim Williams oscillator.
e
r
I'd gotten most of the way to building an example - bought all the parts, i ncluding getting the specially wound transformers - a few years ago, just b efore I moved back to Australia, which gave me enough time to think up a be tter variation, which I may get built real soon now, if I'm lucky.
It's not got much to do with Baxandall's class-D oscillator (though my vari ant of that is what got me started) but Jamie and John are too dim to make that kind of fine distinction.
cation notes for Linear Technology, on high frequency inverters for driving cold cathode back-lights used in laptop computers (application notes AN45, AN49, AN51, AN55, AN61, AN65)." These apparently were Linear Technology's most popular application notes for quite a while.
hn Larkin, who can't cope with transformers that he can't buy off the shelf .
f is just common sense
Restricting yourself to components which you can buy cheaply off the shelf rather narrows your range, particularly when it comes to transformers.
with the outputs in series would work. They are fairly quiet, and you
Oh God, You just couldn't keep that failed OSC out of th subject!
Oh, I don't know about that. A big PSU or one with special requirements might well need custom magnetics, and it probably pays to micro-optimize high volume designs. For annual quantities of tens to hundreds, though, (where John lives and I aspire to) ;) the performance advantage probably doesn't pay for the increased pain in procurement.
SMPSes can be fun once in awhile, but doing vanilla ones as a steady diet must be booooorrrinnnggg.
Custom magnetics is a real pain. Winding them yourself is crazy; they will wind up costing ten or more times what a commercial surface-mount part would cost. Just a bobbin will cost more than an entire finished inductor. Ordering them from a magnetics house takes weeks of waiting and a lot of engineering work documenting the part.
It can take a little ingenuity to make a circuit from standard parts, but there are tons of standard inductors available too.
There's seldom a justification for building an off-line switcher. Things like wall-warts and MeanWell bricks are absurdly cheap and have all the UL/CE/FCC stickers you would ever want.
There are lots of interesting packaged switchers around these days, like 7805 drop-in replacements, hybrid bucks with magnetics and caps, tiny dc/dc converters; they tend to cluster in the $4 range. It's seldom worth even the parts cost of making your own standard switcher; well, some small synchronous bucks aren't too bad, like to go from +5 to +1.2 or something.
Time to move up the abstraction stack.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
Then you probably need to learn a bit more about the ranges of cores and fo rmers available off the shelf.
Obviously. But neither you nor John do much of that.
Transformers and inductors offer rather a lot of variables, so getting one right for a particular application isn't "micro-optimisation".
John Larkin wanted a tune-able inductor which he could have had wound onto an off-the-shelf former for an off-the-shelf core that came with a tuning s lug (I gave him links to both), but he found that too hard.
Transformers for low volume production aren't mass-produced, and there are plenty of small winding shop that will make them for you pretty cheaply, or at least wind the wire onto formers. They'll probably clip the cores aroun d the wound formers for very little more.
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.