Subject: Re: A protection problem
Date: Wed, 8 Jan 2003 15:26:21 +0000 (UTC)
NNTP-Posting-Date: Wed, 8 Jan 2003 15:26:21 +0000 (UTC)
User-Agent: tin/1.5.14-20020926 ("Soil") (UNIX) (Linux/2.4.20 (i686))
Winfield Hill wrote:
> Ian Stirling wrote...
>> Ken Smith wrote:
>>> I'm currently working on a voltage regulator with a major protection
>>> problem. The specs are: Input 23 to 28V. Output voltage DAC
>>> controled 0 to 25V. Output current limited at 100mA, isn't too hard.
>>> Some moron may hook a (+) or (-) 600V supply to the output. This
>>> 600V supply can supply a zillion amps so MOVs aren't the answer.
>>> PTC devices don't seem to be able to handle this case either.
>> Hmm. Let's say the pass element is a P channel MOSFET with a
>> 800V 500ma rating in the positive rail. This should cope with a brief
>> connection to a -600V supply, though it will explode into lots of
>> little bits if the control circuitry doesn't shut it off rapidly.
>> So, the remaining problem is +600V. Can you put up with a diode drop?
> Ian, I think you've got a solution. There aren't any 800V p-channel
> mosfets, but there are some hefty 500V parts that'll likely withstand
> -600 volts, such as Fairchild's FQPF3P50 (otherwise, two series FETs
> can be used). The required 100mA current limit can be of the foldback
> variety, keeping power dissipation under say 25mA (15 watts) for 600V
> fault conditions.
Hmm, if the load doesn't need to be referenced to ground, a N channel
FET can be used in the ground lline, and those certainly are available
at that voltage.
The problem with the diode drop is that as it's not a boosted converter,
it lowers the maximum voltage, but I can't answer if that's OK.
http://inquisitor.i.am/ | mailto:firstname.lastname@example.org | Ian Stirling.
What a wonderfull world it is that has girls in it! -- Robert A Heinlein.