From: Winfield Hill
Subject: Re: An interesting little prob.
Date: 30 Oct 2002 07:00:52 -0800
Organization: Rowland Institute
References: <email@example.com> <firstname.lastname@example.org> <email@example.com>
X-Newsreader: Direct Read News 4.00
> Winfield Hill wrote:
>> . _____ -> 25mA
>> . V+ --|__|__|-----x--|>]--,
>> . | |
>> . ,--[150k]--' x---x--[<|--x--->
>> . | | | |
>> . | | \ PNP V | |
>> . VDAC--- | --|+ \ |/ | |
>> . 0-5V | | ---------| | |
>> . x---|- / |\ | |
>> . | | / \ | |
>> . | | | |
>> . [49k9] 0V | | | \
>> . | | '--|+ \
>> . 0V | | --- high for
>> . '----------|- / CV mode
>> . | /
> I'm leaning towards an inverting opamp there Win
> since this would allow for some roll-off capacitance
> and (more importantly) a diode that limits the opamp
> Vout to -0.7V. (Ref the 2v pedestal in the post that
> replied to Speff.)
Yes I can see the appeal of that. But a non-inverting emitter-
follower approach avoids the painful issue of compensating the
common-emitter stage, whose gain is g_m times the load, where
g_m is a function of clamping current and the load is who knows
what? The resulting compensation forces a slowing of the clamp.
In contrast, the emitter-follower clamp is very fast (as fast
as the EF transistor, the opamp doesn't have to move) and it's
naturally compensated already, doesn't need more. Also, the
opamp/BJT is always on, so an opamp diode limiter isn't needed.
> That 25mA is larger than I was going to use, but I can
> see the sense in that. Because of the 0-20V range it
> will probably have to be a const-I source.
The BJT has to be a serious heat-sinked honker type, right?
To handle fill current at full voltage.