Re: current measurement - hardware

["Mark Davis" <maadavis@nyx.net>]

We have an in-house design we use extensively for a 2-channel Beam Current 
Monitor (BCM) that has 8 ranges from +/- 100 picoamps to +/- 1 milliamp 
(those are full-scale values - 12-bit ADC's are used, so divide by 4096 for 
a rough idea of sensitivity).
The accuracy at the most sensitive ranges is not great, but at least it 
gives you a good idea how much change has occured.

The units contain a ZWorld/Rabbit BL2105 Single Board Computer with a 10 
Mbps ethernet.  The interface to EPICS is via an Open ModBus/TCP connection.

Kelly Davidson (davidson@nscl.msu.edu) is the engineer who has done most the 
design work on these if you are interested in more details (I did the 
programming).

Not sure what the university or the lab's general stance is on our in-house 
work or how much we can share without major hassles (or if they might be 
interested in selling the device to others).

Some contacts if you are interested:

John Vincent (vincent@nscl.msu.edu) - my boss
Kelly Davidson (davidson@nscl.msu.edu) - The engineer responsible for most 
of the design
Mark Davis (me) (davis@nscl.msu.edu) - I wrote the software and the latest 
requirements/specs document

Mark Davis
Control Systems Software Engineeer
National Superconducting Cyclotron Laboratory
Michigan State University

----- Original Message ----- 
From: "David Dudley" <DavidDu@cctexas.com>
To: <tech-talk@aps.anl.gov>
Sent: Monday, November 26, 2007 12:40 PM
Subject: Re: current measurement - hardware


> Might look into something like a Measurement Computing PCI2517 card,
> which has 8 differential analog inputs.
>
> Looking at the spec sheet, it says it has a 40pA bias current, and
> 20Meg input impedance, while measuring a 100mv signal.  With proper load
> resistors (if you even need them at these low currents), you could
> probably measure currents in the Ua range with no problem.
>
> David Dudley
>
> > > > Heinrich du Toit <heinrichdt@tlabs.ac.za> 11/26/2007 8:25 AM >>>
> Hi guys
>
> I'm looking for some information on how you solve this problem...
>
> We need to measure currents.. from harps and slits and faraday cups.
>
> All of this gives of currents that is important.
> Some of the currents we measure with keithleys but it is way to
> expensive to buy keithleys for harps as you will need 100's of them :)
>
> We need to measure currents between nA and uA range for harps and
> slits.
> For faraday cups this sometimes goes to 100+ uA but our main problem is
>
> harps/slits.
>
> I hope that we are not the first to have this problem, is there some
> cost effective method to solve this problem? That can give descent
> results.
>
> The problem can basically be given as:
> How do I measure 50+ current inputs (harp = 96, 4 per slit-set) across
> a
> wide range (2nA fullscale = 50pA sensitive up to 100uA full scale)
> still
> have resonable accuracy especially between the lines and have a descent
>
> update rate and not 2 high noise levels.
> (I guess noise can to some extend be solved with a software filter if
> updates are high enough)
>
> Thanks for your support :)
> -Heinrich