At 14.31 2008.05.28, you wrote:
>Query: How to protect uPower 5V rail at u leakage currents 
>for the protector.
>
>
>Starters for 10 points:
>
>
>A FET may do very well.
>Speed not up to tranzorb speed but may well be fast enough.
>
>A FET has a Vth gate voltage below which it is formally off 
>and you get channel leakage only.
>While Vth is a bit iof a variable feast you should be able 
>to produce something that triggers at less than a say 2:1 
>voltage increase and draws about zilch when off.
>
>You could add a gate divider with ultra low current gain at 
>the expense of slowing it down.
>A speedup cap may get you enough speed with a tolerably low 
>divider.
>
>If you run a low Iq LDO at just above dropout you can place 
>a transistor across it such that the transistor is off 
>usually but on when Vin rises above !~= 0.5V or a bit less. 
>(LM2936 may work). Olin mentioned something similar recently 
>and I have used it successfully for another uPower 
>application.
>
>A smps with low enough Iq can supply a half isolated supply 
>(common ground) that cannot fry if Vin rises high. The smps 
>per se has to fail to allow high Vout. I've used a CD40106 
>as a SMPS with minuscle Iq when no load current required.
>
>A 1N4148 is specd at 0.375V at 10 uA. Put 13 in series (!!! 
>:-) ) and you get a 10 uA leakage at 5V diode that conducts 
>100 uA at 6.5 V and 1  mA at 8V. (Temperature an issue). 
>Feed appropriate clamp trigger.
>
>An LM385 can be operated at as low as 10 uA - but again 
>speed may be an issue.

Or a comparator + a SCR. Speed should be high enough, if not
it can be a good starting point for improvements.

Greets,
Fabio


>
>
>        Russell
>
>
>-- 
>http://www.piclist.com PIC/SX FAQ & list archive
>View/change your membership options at
>http://mailman.mit.edu/mailman/listinfo/piclist

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist