Dual input (i.e. AC and DC) power supply design

Although the diode does act as a rectifier, it will only be "half wave" when presented with an AC input (see

I don't know the specification of your switching regulator, but I would guess that feeding it with this sort of DC (unless it's

*heavily* smoothed with a large cap) is outwith the recommended operating parameters.

Personally, I'd use a bridge rectifier arrangement followed by a smoothing capacitor. This has the added advantage that you can connect a DC supply any way and the circuit will still work (with a blocking diode, it will only work if the polarity is correct).

Hope this helps.

Problem I see is that depending on the tolerance of the 24VAC source, there might be an issue with the 36V upper limit of the regulator.

Tam

Bob

Good point - 24VAC would full wave rectify to approx 33.84VDC. As long as the AC input remains less than 25.5V it should be safe.

There are 35V transient suppressor available which may help to catch any spikes but it's still towards the upper limit of the vreg.

Cheers

--
Yup.

Another problem would be if one side of the input and one side of the
load were supposed to be common, in which case a bridge wouldn\'t work.


JF

Which Micrel IC are you using and does it use an external MOSFET? A

24VAC input implies line operation with a standard 15% tolerance, and that puts a 24VAC input at nearly 40 volts peak. You may have to clamp the V+ drive to the IC itself. Then there is the issue of the input filter capacitor which is probably not large enough for adequate filtering of 50-60Hz. There are simple ways to regulate the VAC but they will not be very efficient, and that is the whole point of using a switcher to begin with. There is insufficient information to answer your stupid question.

Just a quick question - how would this "common" work with the AC input the O/P was talking about?

This isn't a facetious question - I'm genuinely interested how it would (could?) work.

Thanks

Yeah, the simplest thing that comes to mind is a floating transformer secondary with a bridge rectifier and a series diode on the common ground DC output. That may not work, depending on where else the 24VAC goes. If it doesn't go anywhere else, it could be 18VAC.

Tam

Ignoring the fact that I've shown no feedback into the PWM stuff from the output, and assuming he's using a buck regulator with no isolation, then I think he's doing the DC input thing like this, and the diode in series with the DC input will protect the circuitry if the input polarity is reversed.

9-36VDC>--[DIODE>]-+-[switch]---+--[L]--+--->DC OUT | | | | | | |A |+ [PWM STUFF]-+ [DIODE] [C] | | | | | | GND>---------------+------------+-------+--->GND

For isolated AC in with a single diode and a ground-referred load he could do something like this:

24VRMS ~33VDC / / MAINS--+ +--[DIODE>]-+--------+--[switch]---+--[L]--+----+ | | | | | | | | P||S |+ | | |A |+ | R||E [BFC] [PWM STUFF]-+ [DIODE] [C] [LOAD] I||C | | | | | | | | | | | | MAINS--+ +-----------+--------+-------------+-------+----+ | GND

but if he tried it with a grounded 24VAC source and a full-wave bridge he'd have this:

24VAC>----+--[DIODE>]-+ | | GND>----+-|--[DIODE>]-+--------+--[switch]---+--[L]--+----+ | | | | | | | | | | |+ | | |A |+ | | | ) [BFC] [PWM STUFF]-+ [DIODE] [C] [LOAD] | | ( | | | | | | | ) | | | | | | +--[

Hi John

Thanks for taking the time to give a detailed ex plantation.

I hadn't considered the possibility of a grounded AC source - I've never encountered it in my limited experience. My idea relied on the following:

24VAC----+--[DIODE>]--+ | | 24VAC--+----[DIODE>]--+-----33.84VDC | | | | | |+ | | [CAP] | | | | | | | +--[

^^^^^^^^^^^^^ Sorry, victim of over-reliance on spill cheek...