I'm designing another 3-phase PM alternator simulator, and I need a roughly +-60 volt floating power supply. It will feed three PWM half-bridges to make the three output phases.
Here's my first pass at the dc/dc thing.
Maybe an appropriately-selected ferrite bead for the gates but perhaps not a gate resistor or anything elaborate for the LT4444, looks like they figure their adaptive shoot-thru protection can prevent any misbehavior & that kind of fancy is what you pay the LTC device price for I suppose.
Looks like just a fat trace is what they recommend.
Oh. So you don't want to buy one or need help. You just feel compelled to tell us about it, right?
It's an electronic design discussion group.
Design something. Discuss it. Or whine. Whatever you're best at.
The LTC4444 is about $2.80. That's in the noise floor for this board.
The three half-bridges are more voltage, so need the UCC27712. It's more like $1 so I may as well use it everywhere.
Gate resistors might be prudent, stuffed as 0 ohms initially. Could be a resistor or bead if needed.
I might also put schottky diodes across the fets, in case the substrate diodes get weird. Excess caution in both cases.
John S snipped-for-privacy@invalid.org wrote in news:t6p8ag$iuu$ snipped-for-privacy@dont-email.me:
Pre fashioned, vertically mounted POL devices have been massaged out pretty good. No need to try to design your own power source if the pros already did the hard works for you and offer a far less expensive solution than spending your own time trying to fashion your own device. And most folks doing that examine and copy features from the big boys as well. A bit of a cheat.
snipped-for-privacy@highlandsniptechnology.com wrote in news: snipped-for-privacy@4ax.com:
For the primary, build a trough on the PCB where you then add a nice Silver Plated Copper wire, or get a custom wire formed and stamped to lay in the trough. Done. Your primary is of a large capacity wire now, instead of a circuit trace. Then make a nice 0.032 secondary PCB to sit under your core element. The is for the planar transformer.
I think we used Transzorbs and beads on the FET leads.
On a sunny day (Thu, 26 May 2022 17:16:56 -0700) it happened John Larkin <jlarkin@highland_atwork_technology.com> wrote in snipped-for-privacy@4ax.com:
Interesting circuit. In the long ago past I used zeners between source and gate to protect against drain-gate spike injection. Dunno much about this driver chip.. The zeners however caused oscillations to occur at many MHz, a few nF caps between drain and source killed that. Scope the thing! Maybe some gate resistors too?
Neat transformer! How much do they weigh? I'd be wary of relying just on surface mount pads on anything over a couple grams in case the unit gets dropped one day.
piglet
Take care. Chances are, core loss estimates are for a regulated system with '48V' input. For straight DC-DC unregulated, your flux swings will be double the typical application for 48V 'primary', and triple for a 60V 'primary' - core losses scale with a positive exponent greater than one.
You might make this up due to lower current and crest factor in the copper, but . . .
RL
Take a look at the UCC27712 data sheet, fig 44. It's absolutely decorated with parts.
With 150 ns of anti-shoot delay, why all that gate junk?
The LTC works fine without all that stuff. 98 more parts if I use the TI as shown on my board.
The Coilcraft is stock and is pick-and-place. That's sure easy.
The transformer certainly looks very interesting, thanks for posting.
I don't check on that sort of products frequently enough (not being in high quantities and being battle hardened by winding in house our
5kV transformers, recently there was a hiccup running out of 0.07mm wire... one of the attempts to get new supply ended me up with about 100-200 grams of.... 130 Ohm/meter 0.07 wire, noticed it after I got home :). I'll explore these now though.
Where would I buy a 48v to +-60 volt isolated 200 watt dc/dc converter? What would that cost?
My circuit will cost maybe $20. Besides, designing things is what I do.
I've used the LTC4444 before and it works great with no gate resistors. But in my output stages I need more voltage than it can handle, hence the UCC27712 part.
Maybe I'll hack a proto board with the TI, to avoid surprises.
22 grams. It has 9 giant nail-head pins. Should be fine.
John Larkin <jlarkin@highland_atwork_technology.com> wrote in news: snipped-for-privacy@4ax.com:
I doubt there is much market for that in a POL scenario, which is what I spoke on. A small PCB. Maybe I did not read the original thread, because I would not have answered with POL had I seen the numbers you gave.
48V to +-60V @ 200W is a little big for that form factor.
DC to DC converters are things which power engineers spend years massaging.
At a mere $20 outlay, I doubt you will get the DC power you desire. PARD and other elements make it hard to get right the first time. If noise is an issue with the circuit, you may have a hard time providing that DC power cleanly in a one off first time doing it design. And then there is the cubic inches of space you pack it all into, and the operating efficiency numbers.
Lots of reasons to use the boys that have been doing it for decades, especially if the input power requirements of the circuits require it to be clean and free of noise throughout it load range. Good luck.
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.