Re: clamper

On Wed, 6 Nov 2024 21:35:38 +1100, Chris Jones
<lugnut808@spam.yahoo.com> wrote:

On 6/11/2024 1:23 am, john larkin wrote:

On Wed, 6 Nov 2024 00:32:48 +1100, Chris Jones
<lugnut808@spam.yahoo.com> wrote:
 

On 5/11/2024 11:02 am, john larkin wrote:

I want a small cheap voltage clamper device. I'd like under 2 volts
drop at 1 mA, but minimal current at a few hundred millivolts.
>
I tried two small diodes in series, but that's terrible. Better is a
logic-level mosfet with gate connected to drain. It clamps nicely at
1.5 volts or so but conducts picoamps at a few hundred mV, over 1000:1
better than the diodes.
>
This will go between the force and sense leads of a 4-wire temperature
sensor thing so it automatically works in 2-wire or 4-wire mode.
>
The alternative is to use two SSRs and let the user explicitely
declare 2-w or 4-w mode.
>
>

>
>
Use two diodes in series, with the junction between the diodes actively
driven by an op-amp follower to be the same voltage as the low-leakage
node. The op-amp can be a low bias current type, like LMC662 etc.

 
https://www.dropbox.com/scl/fi/pj7fd43f787urqav1dewc/B450_Front_1.jpg?rlkey=69oh858vsod4efnoes1ya5uql&dl=0
 
Seems to me that poking any current into the junction of the upper
diode pair must make a current source error.
 
Maybe I'll just use two isolated switches and let the user select 2w
or 4w measurement mode. One ACPL227 should work, and I can use that
somewhere else too.
 
 

>
>
You didn't tell me that *both* ends of your clamp needed to have low
leakage.

I did explain the function, automatic 2-wire or 4-wire measurements.