If compact size is not extremelly important, you can try what I use for
"mobile" datalogging:

A custom made PIC small board that captures whatever you want on A/D or
DI/O, and connects RS232 to a simple PDA, recording everything into a CF
4GB. You can post-process that to achieve required data output on the very
same PDA or later on a PC, and it is cheap if you buy an used PDA on eBay.

HTH

2007/9/19, Sean Breheny <shb7@cornell.edu>:
>
> Thanks for the response, Eoin.
>
> I need something which is stand-alone since it has to go on a mobile
> robot platform. It'd say it should be able to fit in about a 6 inch
> cube and should have its own battery or be able to take a DC supply in
> from the robot.
>
> Sean
>
>
> On 9/19/07, Eoin Ross <eross@chemstation.com> wrote:
> > Without having mobility/size/wattage constraints....
> >
> > http://www.measurementcomputing.com/
> > http://www.measurementcomputing.com/dasylab.html
> >
> > Measurement computing does USB modules reasonably cheaply, and they have
> a labview style development system for logging and display - of course this
> requires a PC though.
> >
> >
> >
> > >>> shb7@cornell.edu 19 Sep 07 14:59:13 >>>
> > Hi all,
> >
> > I often run into situations where a device behaves fine "on the bench"
> > but not when installed (in this case, in a mobile robot). If I had a
> > device which could monitor and record several analog and digital
> > signals, store this data, and was battery powered, I could install it
> > in the robot and gather debugging data under actual circumstances.
> >
> > I've looked at various dataloggers (the only name I know for something
> > like this) and none had the specs I need. Most had sampling rates
> > which were too low. Others had too little memory. Also, ideally, I'd
> > want something with trigger capabilities (for example, start recording
> > when a line goes high, start recording on a rising edge of an analog
> > signal, etc.) and maybe the ability to record a synchronous or async
> > serial signal. I haven't found such a device.
> >
> > Here's a description of something which would fit the bill:
> > 2 or more channels of analog input with an adjustable sample rate
> > (with a max of at least 1 megasample per second) and a memory depth of
> > at least 1 million samples. 8 bit resolution OK.
> >
> > 5 or more channels of adjustable threshold digital inputs (switchable
> > between 3.3V and 5V logic at least). Similar sample rate and memory
> > depth to analog channels.
> >
> > Several triggering options, including at least triggering on level and
> > edge of digital inputs, and edge of analog input. Should have some
> > ability to start and stop recording and tag those times with a real
> > timestamp.
> >
> > Has anyone found something like this? Even if it runs into significant
> > $, I wouldn't mind because of the time it would save.
> >
> > Thanks,
> >
> > Sean
> > --
> > 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
> >
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>



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