Yes, I know. That was suppose to be the "test source" that would supply the 18v @300A to his power supply. I hit "send message" before I was done scribbling the rest. Sorry(tm).
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
How stable? How accurate? Over what load range? Over what input range? Minimum efficiency at various loads? Maximum ripple? Over what temperature range? Humidity? Any other environmental specs? Any limits to size, weight, materials, i/o connectors, acoustic noise, shock, vibration, etc? Short circuit protection, shutdown, or survival? Fold back protection? Does it need to comply with any EMI/RFI conducted/radiated specs? Duz it need UL, CE, EMC, FCC, CSA, municipal certifications, and/or other alphabet soup approvals? ISO
9000/9001 documentation? Kosher certification?In case it's not obvious, you need to nail down some target specifications. It's difficult to know that you've arrived if you don't know where you're going. Many of the aforementioned specs have big ticket dollars attached to them. If you're not sure if a series resistor will or will not work, then you don't have a clue as to what you're designing or building. I suspect that you may be attempting something more difficult that your current skill level allows. If you insist on building this DC to DC converter, I suggest you find an experienced mentor to help you along and keep you from killing yourself. Good luck.
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
John Larkin really doesn't like thinking about getting inductors wound for him.
But he probably does have a point here - while ferrite cores are available in a range of sizes, what you have in mind would call for a rather bulky core if you restricted yourself to a single wound component ( or a single set of wound components).
-- Bill Sloman, Sydney
There's a big difference between a power supply and a battery charger. This should help:
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
On Fri, 18 Sep 2015 23:32:01 -0400, krw Gave us:
Fair-rite will custom wind rods, toroids, etc. for the customer spec..
[OT] My most recent polyphase buck converter had 4 phases with single turn (150nH) coupled inductors. It was only 50A total though.
This was my first coupled-inductor DC/DC converter. I was quite impressed by what I could achieve compared to the uncoupled (i.e. 4 independent inductors) 4-phase case.
I'd known about the technique for a while, but I only properly understood it recently. The coupled inductors are like a transformer with a lot of leakage inductance, e.g. Ll is 20-50% of Lm. The result is a big reduction in peak inductor current (which helps efficiency) and a reduction in output ripple current. The latter can be used to reduce the amount of output capacitance, or (for a given amount of output capacitance) can allow the use of smaller inductances, leading to wider bandwidth and faster transient response.
The 50A load is an FPGA. The total voltage deviation (including tolerance, IR drop, transient response, etc.) is less than +/- 30mV at the load. The current can switch from a few amps to 50A (or back) in less than 1us.
+/-50A switching + +/-30mV tolerance = fun.I'm also thoroughly impressed by the second generation of Al-poly caps. These have lower ESR and ESL than the first generation, and are in smaller packages.
Allan
I'm guessing he has a "resuce cart" with three large 6V batteries in series, and he wants to use that so charge depleted 12V batteries in situ.
So the input could be up around 22V especially at low currents.
voltage regultion is less important when charging batteries, the battery will absorb rapid fluctuations, but you still need to control the average output voltage
-- \_(?)_
If you want to use HF switching I'd use (a lot of) MosFets, if not you could use bjt. You didn't mention whether or not you need galvanic insulation. Sepic is a up/down-converter. If your input is really 18 V and you don't need insulation between in- and output and can even tolerate a common ground, I'd advice a simple down-chopper (no, not the motorcycle :). With those voltages that would be quite efficient.
joe
It's that time of year again.
What is the load? It's often more practical to point-regulate at the load/loads.
For a first project, a simple buck regulato, or phase staggered buck regulators in parallel makes the most sense.
If you breadboard the control circuit at lower power, then beef it up or multiply the number of phases as you gain more familiarity with the hardware.
Solutions that might make more sense in a commercial design are likely inappropriate or impracyical as learning tools.
A working circuit can be beefed up with the later introduction of synchronous switching, fancy new semiconductor switches, external digital modulator functions etc etc etc.
RL
Sorry for going off-topic, but this reminds me to an idea that popped up in my mind last week when my car battery (24 V) was empty. Is there anything against kicking its ass with DC inverter TIG-welder? Of course my intuition says it can be done, but I'm afraid I'm overlooking that one cute little detail which will make it blow up in my face when I try it.
joe
How about Powerex 600A modules?
joe
I think the difference is mainly in the controller, not even the driver. I mean power circuit and driver can be the same for the charger as well as the power supply. The 14.5 V sounds like a charger.
joe
On 19 Sep 2015 13:08:58 GMT, joe hey Gave us:
If it is, then it does not need to be as accurate as he describes.
Noise would also no longer be a significant factor.
TIG welding uses a constant current power source. You don't want to use that to charge a Pb battery.
Now there are constant voltage welding power sources used for MIG wire welding. And 14 volts is within the realm of those units.
I have a voltmeter, I don't need a 'constant voltage source' which actually isn't a 'constant voltage' source rather than a voltage controlled power supply which, when the battery is big enough and sufficiently discharged, will not be able to start with a 'constant voltage' anyway.
I'm quite able to reduce the current of the welder when the voltage reaches 28.8 for a wet cell- or 27.4 V for a 'maintenance free' battery.
My main concern is whether one can convince me that I won't blow up my welder doing this. Of course I'll disengage the HF current pulse of the welder for starting the arch.
Ah, didn't know that, thanks.
joe
50 amps into an FPGA is terrifying. What does that chip cost? I'd sure hate to have some transient or something blow up a few of those.
I independently invented the polyphase switcher, a long time ago. It was a 3-phase, 120 amp, 6 volt supply for a CAMAC crate, using bipolar switches at 50 KHz. The triangular ripples canceled nicely into the filter caps. The only problem was that I kept burning the paint off the MicroMetals powdered-iron toroids, which was fixed by using the brand-new KoolMu cores.
I have a customer that wants a 250 amp laser driver, so I may have to do it again. I really don't enjoy designing high power stuff.
Yeah, the polymer caps are amazing. Unlike wet caps, they do fail instantly if overvoltaged.
That knocks out the carbon pile approach then.
NT
On Sat, 19 Sep 2015 11:06:15 -0700 (PDT), snipped-for-privacy@gmail.com Gave us:
Makes me think of scenarios like when kids used to go out and siphon gas from cars back in the seventies. Go out and hook into someone's electric car battery and steal 'lectrons. Gazillions of 'em!
Linear mosfets would work, dissipating a kilowatt or two.
This is 4500 W DC output! I really don't think you can do this with a SEPIC. Generally, stuff like this is done with a 2 transistor push-pull switch and a center-tapped transformer, when run from mains voltage.
But, you will be pulling in 250 A (plus losses) from the 18V source! If you don't need isolation, that opens up some more topologies. If just a simple buck converter, you'd use some MONSTER low-voltage FETs in parallel, and also need some monster Schottky diodes. Litz wire for 300A is going to be huge, but you will likely need it. God knows what you'd use for a core for such a thing.
Jon
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.