negative supply

What would be your favorite way to turn +24 volts into -6 at a couple of amps? Isolation not required.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin
Loading thread data ...

if you favorite buck can handle 24+6V

formatting link

Reply to
Lasse Langwadt Christensen

Yeah, if I can live with -5, I could split up my loads and use a few of these:

formatting link

That one can do 24 to -5.

My fave switcher, TPS54302, can't quite handle the voltage.

We sometimes use a Cuk circuit, but that involves a lot of parts.

We stock LTM8023, which would work too. A bit expensive, but nice part.

Just asking.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

What about the ol' MC34063? It's not fancy or nothin'. It's extremely cheap. It'll handle the voltage requirement easily and can invert 24 to

-6 no prob in this configuration:

To meet the current requirement u can just add a PNP or depletion MOSFET (pull-down on the gate to the negative voltage to turn it OFF) or something with the base or gate connected to pin 1, emitter/drain to +24 and collector/source to point 2 to make a three-transistor "NPN" Sziklai-thing.

Reply to
bitrex

7350283

silly but, use two, first one to drop it Vin to say 12V

Reply to
Lasse Langwadt Christensen

I recently designed (haven't tested yet) a circuit with exactly these specs . I used two interleaved buck controllers, connected as buck-boost convert ers (as in the link posted above by Lasse). For the controllers I used LT8

643S. Like I said, haven't tested yet, so I don't know if the LT8643S has any warts but it looks like a nice part. I used an external clock to synch ronize them; maybe there is a similar part with some kind of sync/phase pin so no external clock needed.
Reply to
sea moss

I often use them in cascades, to reduce the duty cycle on low voltage supplies. Switching 24 to 1.2 isn't as efficient as switching to +5 first, especially when I need some +5 anyhow.

Maybe I could switch to +15, use that for opamps or something, then use one or two more to get a couple of -6 rails.

I haven't tried TPS54302 as an inverting buck, but it ought to work.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

LM2956 in inverting mode, probably. The downside is that it takes a biggish inductor and a fair amount of board space, but the upside is that its slow edges are pretty benign for EMI.

I'm having a whack at using the LMR23625 (that 650-ps edge part) in a Faraday cage with decent filtering on both ends. We'll see. If it can be tamed that easily, it will be an excellent candidate going forward. I'll post updates as they're available.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Lots of 7660's in parallel (might wants something to knock that +24volts down a bit too)! :D

--
Using Opera's mail client: http://www.opera.com/mail/
Reply to
David Eather

On a sunny day (Sun, 16 Sep 2018 11:29:13 -0700) it happened John Larkin wrote in :

Why are so many afraid of transformers?

Reply to
<698839253X6D445TD

what advantage does a transformer provide here when isolation isn't required?

the real WTF is that most suggestions posted so far have been parts like the LTM8023 - $20 in singles, like really you need a $20 part or two "interleaved" $10 LTC buck controllers to invert 24 volts to -6 at a couple of amps? reaalllly

I think PC mobos would cost five thousand dollars if these folks were designing 'em, maybe "favorite" meant "favorite" in the sense of "what's your favorite food?" okay yeah, caviar and champagne would surely rank highly on my list. Living large:

Or else there is some other specification which requires that performance which was not revealed in the original post.

Reply to
bitrex

Something to do with wire diameter, turns ratio, and ohmic losses.

I have some EI cores of various sizes, when I need some voltage I just wind something... It is not that hard.

In my PC mobo supplies there are transformers.

crap.

Maybe someone wants small, light, cheap, fully integrated, 99.99% efficiency, 100A out, spread spectrum, low EMI, foldback current protected, remote controled via internet, flashing blue LEDs, ... .. would be nice. I would buy that chip.

Reply to
<698839253X6D445TD

Post a solution and we'll discuss it.

MoBos are built by the million from incredibly cheap Chinese parts, and sell at minute margins. We sell in fairly small volume at high margins, where parts cost is not as important as getting a design done quickly at low hassle levels.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

One can usually find a stocked, multi-sourced inductor cheap in distribution. You have to get lucky to find the transformer that you need. Custom design and winding is a huge nuisance.

This is a nice little transformer for power apps:

formatting link

It's sole-source, but Coilcraft has assured me that it will be in production for at least 10 years more. We've used over 4000 so far.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

John Larkin wrote

OK, nice, but I don't see why one cannot wind a simple coil like that, and then pick the right diameter wire in the process, in just a few minutes.

When you make hundreds or thousands of things with some transformer, yes, have those made, there are companies here that will wind transformers for you. Else hire a guy... freaking 3 x 5 turns what is it?

You make small series expensive equipment was my understanding? Same reason I cannot understand your remark 'if I used trimmers my production department would kill me' I would fire them. Ever looked in a Tek scope? Did you see how many trim points there are? Some of the fast digital CRT ones were even 'trimmed' by bending the wires to the defection plates a bit.

Reply to
<698839253X6D445TD

Transformers are wonderfully diverse; there's options for high isolation (the heater winding for your microwave oven magnetron needs kilovolts), for high coupling, for triple windings (one secondary passes current, and the other only senses voltage, so has no IR drop), for high frequency with a gap for switchmode, or without a gap for coupling, for low frequency and high efficiency. Then there's exotics like slug-tuned IF transformers and Tesla coils...

And there's variants like current transformers and ferroresonant circuits, and three-phase. And of course many winding ratios and tap options. Some even come in surface-mount packages.

So, that's the problem right there. There's too many variables, only a six-dimensional table can hold them. With ten values for each variable, that's a million different transformers, and nobody can stock that range. I've wound my own, and was able to turn a pile of surplus cores into pretty good network adapters, but I knew that the design wasn't optimal (just good enough, the wasted space and materials weren't crucial).

A few makers (Coilcraft) have catalogs of designs, but you have to search those catalogs for at least three or four of the six parameters that makes your gizmo work. Then, maybe you can order from stock. Otherwise, it's months of leadtime before custom parts come in, and six parameters means a lot of ways the parts can prove unsatisfactory.

So, mainly your best experience with transformers is going to be with off-the-shelf units (perhaps intended for another purpose), which won't be optimal for YOUR purpose. And, it's unlikely that you can find a good second source. Also unlikely, is under-a-buck pricing. Beyond having a kit of cores and putting together a special in an hour or two at the workbench, there's just not a lot of good parts-for-the-project experiences associated with transformers.

Nowadays, even the build-it-yourself route is difficult: magnet wire in a dozen diameters costs... WHAT!!!!

Reply to
whit3rd

Do the math, get the core and the header and the wire, and wind/solder/test it in a few minutes? A month would be more reasonable.

I'm not a hobbyist; we design for production. If something goes into production, we need formally released specs and drawings. We need a qualified vendor, prototypes and purchase orders and such. Everything has to be controlled and reproducible, or we accumulate production nightmares. Custom magnetics are the last resort.

Hire a guy to wind transformers? No thanks.

They are awfully good, and I like them. We prefer to do closed-box fully automated cal and test.

Not "are", "were." Look inside a modern scope.

"Were" again. Labor used to be cheap in Oregon.

--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin

I dunno, I have always wound transformers, I remember in my technical school days, was sort of redesigning a tube TV for transistor TV, needed a horizontal output transformer, Now that IS the works, scan coils drive, EHT, inductance, resonance flyback, parabolic waveforms. Anyways after scribbling pages and pages full of formulae and that for different ferrites, airgaps, circuit types... I had the -insight- :what! .. those TV H output for BW in that time all had a DY97 HV 1.4V heater rectifier tube on a one turn loop for the heater....

formatting link
that made ALL those BW transformers 1.4V effective per turn..... From all the million variants that could exist... So wound my own, kept that in mind..... So many possibilities. It is usually simple PMT HV supply:
formatting link

Bit of HV to keep drone in the air via thin coax:

formatting link

formatting link
formatting link
formatting link
formatting link
formatting link
formatting link
formatting link
Heating a baco with a spiral air coil:
formatting link

transformers is easy.

Reply to
<698839253X6D445TD

John Larkin opiniated

See my other reply to John, no offense, it is the mathematician and the ball player. The mamathemagician will need airspeed, angles, Newtons, diameters, windspeed, temperature, distances. surface properties, air pressure, and a few hundred other things, among those TIME. and even a seal can catch a ball on its nose, just like that. Never Under Estimate the Neural Net, and it does not do any math.

Transformers is easy, that movie was a hit too.

Yes, nightmares so bad bad bad

Sell it as truly hand made to reach utmost precision.

Is till here, call an agency, tell them you need somebody to wind some wires.... show him how to do it, plenty of east Europeans available ....

Does the US not still have sweatshops there on the west coast ;-)

formatting link

Did you know we are even free here to visit anty place in Cuba?

How was the skiing?

Reply to
<698839253X6D445TD

That would extend the design time to several months. I could wind up paying $100K to wind a few transformers.

Last year it was great, but I tweaked my knee on my last run of the year at Squaw... the last 50 yards in fact. My friend M likes Squaw, but it's full of dangerous yahoos and horrible bottlenecks and I usually get hurt there.

I'll stick to Sugar Bowl this year. Squaw is fabulously expensive anyhow, $1000 for the season pass.

People are predicting a good snow season coming up, which is about as good as all such weather and climate predictions; worthless.

--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin

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.