I largely agree with your assessment vis-a-vie black holes. I've had many a false start get foiled by sourcing magnetics. To be completely blunt, I stopped reading as you started going through the math, sorry :-) In the range of 48v in and 3ish amps, National Semiconductor (now, TI I believe) Simple switchers are quite useful. I'm particularly fond of the LM2576 HV variant that can handle 60 volts in and uses magnetics that are well defined and available off the shelf with second sources - all detailed with manufacturers and part numbers in the data sheet. Several output voltage versions exist, including an "adjustable" that you can set with a resistor divider. I'm sure competitors are out there, but I've been using these for near ten years now without significant issue. Here's hoping this advice proves to be twice as valuable as what it cost you :-) -Denny On Mon, Jun 5, 2017 at 11:13 AM, Byron Jeff wrote: > OK, so here's the story so far... > > Byron believes (with tongue firmly planted in cheek) that switching > magnetics are a black hole for all knowledge and the source of all that i= s > evil in electronics design. Russell has countered: > > "Do NOT be afraid of an inductor based circuit. With off the shelf and of= f > the net circuits and advice from here you should have no excessibely majo= r > [tm] problems with the converter." > > So I'm here to put this hypothesis to the test. For my electric riding > mower project I see two switching projects that would really be helpful. > Both are based on the 48V 47Ahr Chevy Volt module that I bought to run th= e > mower. > > > The first is pretty simple. I'm also aware that it's available as a black > box cheap enough that it isn't really worth the effort. Call it a test > case. The main electrical switch is a Tyco 500A contactor: > > http://www.mouser.com/ds/lrg/418/NG_CS_EV200_R_TBD_KILOVAC_ > EV200_Ser_Contactors_080-256777.jpg > > My model has the 9-36VDC coil. Because there are numerous USB car charger= s > available a buck regulator down to 12 VDC could easily serve the threefol= d > purpose of driving the contactor, powering USB, and providing power for > auxillary electronics. > > Since the contactor is the primary user the specs for the power supply of > 3.8A inrush peak current and 130 mA of holding current give the ranges of > current usage of the regulator. > > Now remember this is COTS design. And for me specifically, that means > designing the project around the inductor and the current requirements an= d > not start with a fixed switching frequency and designing to a specific > inductor. So the candidate is here (because I already have them in my par= ts > box): > > http://www.bgmicro.com/512uh-coil.aspx > > This part was choosen for a lot of reasons. The large inductance generall= y > means a lower switching frequency. Also larger inductances facilitate > continuous current operation at low current draw. Also both the core > material (#26) and the size (T-184) facilitate higher current designs.The > fact that they are available and relatively cheap doesn't hurt either. > > So one excellent example of the black hole of information is this TI Buck > Regulator Formula sheet: > > http://www.ti.com/lit/an/slva477b/slva477b.pdf > > Note like most others it uses the design parameters of inductor ripple > current and switching frequency to do inductor selection. However, since > the inductor is fixed. We need to rearrange the basic formula: > > I(ripple) =3D (Vin(max) - Vout) * (Vout/Vin) / (f * L) > > by swapping I(ripple) and f. Typically I(ripple) is computed to be betwee= n > 20 and 40 percent of maximum current. Just for some overhead let's put th= e > max current a 5A and I(ripple) at 30% of that: 1.5A. So given Vin (Max) o= f > 50V, Vout of 12V, L of 0.000512H, and the given I(ripple): > > f =3D (50V - 12V) * (12V/50V) / (1.5A * 0.000512) > > f =3D 11.85 Khz > > That's well within the spec of the inductor, so bumping it up to an even = 12 > Khz or so should be fine. > > But wait! In order for continuous current mode, the current draw must be = at > minimum half of I(ripple). So that means a continuous draw of 750 mA is > required. But there's no guarantee of that from the contactor. > > These are the types of problems that drive me nuts in trying to design su= ch > projects. And let not even get started with trying to identify a flyback > transformer that'll take a 170 VDC bus voltage and generate an 8A charger > for the battery. > > So any insight on how to design workable switching projects without > resorting to black magic sure would be welcome. > > BAJ > -- > http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist > > -- > Byron A. Jeff > Associate Professor: Department of Computer Science and Information > Technology > College of Information and Mathematical Sciences > Clayton State University > http://faculty.clayton.edu/bjeff > -- > http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist > --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .