Quote:
Originally Posted by trobbins
... does that indicate you will just use the 13A domestic feed CB as the over-current protection to your equipment...
|
Erm, no, it doesn't!! That would win a Darwin award.
But just to be clear, in case it wasn't before.
1)...of course, there'll be a plug dedicated to the power side, with a suitable fuse. I have no idea where the impression might have come from that this would be omitted!!!
2) If there is a fault condition (load drawing too much current instantaneously, which in effect means in less than two RF periods), then a current transformer will be the feedback mechanism to logic circuitry I've designed that will shut down the drive to the whole circuit. As I say, it will do this 20,000 times faster than the MCB. (0.1mS, for the sake of example/5uS)
In this event, until the fault clears, the total mains current will at that point be defined by the miniscule leakage current of the capacitors plus the tiny current flowing in the safety (bleeder) resistors in parallel with each. Not very much at all, in other words.
The overcurrent protection circuitry is, of course, a separate issue to the charging currents in the caps, but there has never been any confusion over this anywhere in the thread.
You repeat another point:
Quote:
Originally Posted by trobbins
If the ...mains effective series impedance was really low, and there was a significant load on the filter caps, then the mains charging current pulses would trend to being of very short duration and very high crest factor, as was generically described by Schade
|
I get your concern, but, in the nicest possible way, we're just going over exactly the same ground as before.
What you're stressing here is that even if (and it is a given)... even if I have inrush current limiting circuitry, then the mains charging current pulses in the diodes would have a very high (your are proposing too high) crest factor.
Yes, this is relevant...but... also please note that Hugo addressed this, along with a solution (charging reactor/inductor) earlier on. (See post no 32.) We also discussed the possible role of a zero crossing switch. This is a good idea, although one suitably rated costs around £100 here.
But...I not going to wait until I've built the whole power side before I investigate the nature of the charging pulses into the caps. I'll run a simulation, e.g in LT Spice, and then build and test that subassembly (maybe a single diode and a single cap - each with the properties of the ones in the final design - to start with) before I go on to the rest of it.
Schade, writing in 1943, would likely fall off his chair if he knew that 80 years or so later, there'd be silicon diodes and bridge rectifiers with surge current ratings of 500A, and for a few £!
Hope things are all clearer now?
