24 Vac -> 5 Vdc isolated?

I'm planning a project that has 24 VAC available to it, and needs about an amp of 5v. I've designed a circuit with a bridge rectifier, a couple of 1000uF caps, and one of these:

formatting link

Small footprint, wide input range (I calculate about 33vdc out of the caps), electrical isolation, but kinda pricey. All total, the PS takes up about 2 SQ in of board space (4"x0.5" - yeah, I don't have a lot of space), half of which is the caps.

Can anyone suggest a simpler, smaller, or less expensive solution? No, 120VAC isn't much of an option - this board will go inside my furnace[*], so running additional 120VAC lines is nontrivial. (and the 24VAC is needed on the board *anyway*).

[*]
formatting link
Reply to
DJ Delorie
Loading thread data ...

Interesting. We are custom ordering something like that, which is a switching regulating module. Except that we needed 2 resistors to set the voltage between 1.2V, 1.5V,1.8V, 2.5V, 3.3V and 5V. Maximum of 3A. We plan on vertical mounting only, with the Vin and Gnd on top. So that multiple Vin and Gnd can be tied together, for support and cooling.

The manufacturer claims to be able to do it for $2 to $3 each, qty

1000.

Reply to
linnix

Interesting. We are custom ordering something like that, which is a switching regulating module. Except that we needed 2 resistors to set the voltage between 1.2V, 1.5V,1.8V, 2.5V, 3.3V and 5V. Maximum of 3A. We plan on vertical mounting only, with the Vin and Gnd on top. So that multiple Vin and Gnd can be tied together, for support and cooling.

The manufacturer claims to be able to do it for $2 to $3 each, qty

1000.

Reply to
linnix

Well, I need qty one. That makes *everything* expensive :-P

Reply to
DJ Delorie

Might be able to arrange it. Just drop me a note at

formatting link

Reply to
linnix

How did you decide how much capacitor you need? It seems to me that you could get by with much less capacitor on the rectified side of your bridge.

Unless I'm missing something, you might start by reducing your

2,000 uF to about 470 uF. That would shrink the size quite a bit.

There must be loads of DC-DC converter modules out there, although I imagine the isolated ones are more pricey than the non-isolated ones.

Best of luck! Thanks for all your work on DJGPP and other projects.

--Mac

Reply to
Mac

I figured it based on the max output of the module (2.6A), the voltage ratio, and converter efficiency, giving about 0.5A max supply current. Then, add in the max ripple the module allows, plug it all in to a power supply simulator I downloaded (PSU Designer II), and play with the cap numbers. I was looking at the Panasonic M-series; a 1000uF

50v is only 12mm diameter and inexpensive. Yeah, I could probably go with less, but as it's inside my furnace[*], a little overkill is a safe bet.

Besides, as far as space is concerned, because of layout of the the isolation gap in the PCB traces between the AC and logic sides, as long as it fits into the top 1/2" of the PCB I can use the whole width of the board.

I just simulated an ideal 470u cap with a 140mA load (average load with my current amp budget) and it offered a 5v ripple. The module I was looking at wanted 2v or less.

Yeah, the isolation is the thing. But since I have no idea what those

24vac wires might come in contact with, I think I need it.

That and fitting it into a tiny space. The ACON CX module is a vertical-mount one, so the footprint on the board is small. Most others are flat-mount, which would eat up space really fast.

You're welcome :-)

[*] Technically, it's beside my furnace, but physically attached to it inside a small enclosure. Still a nasty place for electronics. See the photos.
Reply to
DJ Delorie

Oh, I used 1 Amp output current, and plugged in 5 V of ripple to the ripple equation, and I got 400 some-odd uF. I didn't realize there was an input ripple spec. I though you just had to keep the input Voltage above the minimum listed.

The ripple equation: I = C * dV / dT C = I * dT / dV

I used 250 mA for I, 1/120 Hz for dT, and 5V for dV.

That gives something like 417 uF.

Sounds like a lot of ripple for 140 mA. If you drain a 470 uF cap at

140 mA for 8 ms, you should only get about 2.5V of drop...

I did a quick search on digikey, and at vicor's website. I didn't find anything better or cheaper than what you found already. I'm not that up on this sort of thing.

If you want to design your own, I'm sure you could get rid of quite a few uF. You should be able to let the input droop all the way down to 10 V and still easily stay in regulation at the output. But that is just another subproject to stretch things out even longer. I would guess that the module you have found might be the best way to go, all things considered.

All the ones I found on digikey were flat-mount.

Good luck!

--Mac

Reply to
Mac

Hmmm... I just re-ran the simulations and got numbers consistent with yours. Sigh. I could probably go with half the uF of my original plan, which means four smaller caps (thinner strip) or two of the same caps (might leave space on one side).

Thanks!

Reply to
DJ Delorie

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.