Subject: Re: 2.5kW-3kW-SMPS-(cheap)-suggestions?
Date: Wed, 18 Dec 2002 17:59:13 +0000 (UTC)
NNTP-Posting-Date: Wed, 18 Dec 2002 17:59:13 +0000 (UTC)
User-Agent: tin/1.5.14-20020926 ("Soil") (UNIX) (Linux/2.4.19 (i686))
John Jardine wrote:
> wrote in message
>> John Jardine wrote:
>> > wrote in message
>> > news:firstname.lastname@example.org...
>> >> John Jardine wrote:
>> >> > Help!
>> > [clip]
>> >> Take a few uf of motor run capacitors (well, low ESR anyway), attatch
>> > across
>> >> the plug.
>> >> Now attatch a bridge to give you DC.
>> >> Now put a biggish FET between that DC and the load, and PWM it so that
>> >> the overall power factor comes back to a resistive load.
>> >> Yes, the waveform across the resistive load isn't DC, but do you care?
>> >> No magnetics or stability problems, ...
>> For example, 50A load current RMS at 40V, and a peak voltage of 211V,
>> you need to PWM the load at about 6.2%, with a peak current of ~200A.
>> Assuming 1us on-time, to keep a 10V ripple on the cap, that's 20uF.
>> You will of course need a choke between this and the generator output.
>> Oh. EMC.
> Yes. I was thinking of a series connection :-(. This arrangement though
> seems to have no electrial isolation. Which is a need. The idea's
> interesting though and It's worth spicing the setup to get a handle on the
> The cheap alternators I've come across in the past can have a lot of non
> linear series inductance. Over one alternator revolution I've seen it range
> from 10mH to 60mH. This may be of fortuitous benefit but I've got a feeling
> that higher peak currents may cause a problem here, as the alternator AVR
> setup may somehow notice the change in power 'take' and combined with the
> loss reactance the average o/p voltage may drop off substantially.
> Anyway, (After I've figured a frequency variable inductance), let the spice
> show me the true path :-).
There is no load reactance.
It's all resistive.
The capacitor across the line is fairly small, but simply modifying the
PWM controller a bit so that it has an overall slight inductance (or
putting the current feedback before the capacitor) it's cancelled.
The choke in the feed line isn't to cancel the input capacitance, but
to stop the 1Mhz RF getting back to the generator.
Mr Edison was the first to find that that was a bad idea.
> Ian ... it's a blast from the past, mantra ... "50Amp, 40V, 150Vac" ... I'm
> quizzical here, mmmm!. That's the old Bgas ECE spec'. Have you had
> involvement with the boxes? ;-)
http://inquisitor.i.am/ | mailto:email@example.com | Ian Stirling.
He who lives in a glass house should not invite he who is without sin.