The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: email@example.com (Bill Allison)
Subject: Re: Help - Power mosfets - difficult load
Date: 23 Oct 2002 03:44:12 -0700
References: <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <3DB5EEF9.BF30CE32@rica.net>
NNTP-Posting-Date: 23 Oct 2002 10:44:12 GMT
John Popelish wrote in message news:<3DB5EEF9.BF30CE32@rica.net>...
> Bill Allison wrote:
> > I am encouraged by the fact that Futaba Corporation have made millions
> > selling servos driven at 50Hz. I am further encouraged by their choice
> > of a frequency as low as 120Hz for their newest servos in which (they
> > are microcontroller-based) they could have chosen virtually any
> > frequency.
> > We may to an extent be at cross purposes here. I am concerned with fet
> > switching losses, because I know from the graphs that if I am not
> > careful in this area I risk straying outside SOA. The less time the
> > fet spends in the switching region the better from the point of view
> > of heat generation in it. When fully on the power dissipated will be
> > I^2 * R where R = 0.0033 ohm for the pair of FB180SA10 I will use i.e.
> > 33 watt at 100 amp and of course when fully off the power dissipated
> > in the fet will be zero. But when switching, the power dissipated,
> > ignoring inductive lag, will be highest at the midpoint when 6 volts *
> > 50 amps = 300w. So I want to be in that switching zone as infrequently
> > as practicable. I am not concerned with motor or overall efficiency,
> > because the motor has power aplenty for the job and battery capacity
> > is not an issue either.
> > I do accept that the *system* will operate more efficiently at higher
> > frequencies (although the motor vibration from 50 Hz excitation is
> > undoubtedly useful to overcome brush stiction and improve low speed
> > starting). But as the control input available to me is a 50 Hz pulse
> > train it is simplest and makes sense to stay with comparing that with
> > a similar train derived from towline tension and expand the difference
> > to (via a suitable drive IC) gate my fets.
> > So I'm going with 50 Hz. It's the practical solution and will work.
> The point you miss is that during a long on time, the current climbs
> much higher than the average current, so the switch off losses are
> quite a bit higher per pulse than if the ripple current were kept
> lower. This is the reason for the sweet spot around a kilohertz or
> two for line voltage motors. The lowest loss point may be lower for a
> big 12 volt motor but I am willing to bet a dollar that it is higher
> than 50 Hz.
> But you can find this out by experiment after you have a functional
I'm indebted to you for your clear, unemotional responses to my
ill-articulated attempts to get at and understand the information I
Re your latest - I understand that the low frequency will give time
for the current to overshoot the average, but will the losses due to
that be dissipated in the fet or in the schottky? My concern, because
of their cost, is to minimise the risk of blowing the fets. I'm not
that concerned about overall efficiency. And as I've said, 50Hz is
conveniently to hand and I suspect won't be that terrible with my
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup