🚀 go-pugleaf

Hi Charles
I've been contemplating a project for a while that would need analogue
input - specifically for home/science projects like monitoring discharge
voltage of some NiCad battery packs, to assess capacity and whether my
older ones are ready to scrap, or whether they can be rejuvenated with
some controlled charge/discharge cycles - but that's another story(:-}.

I was thinking of designing a scheme using the parallel printer port
LPT1 as an I/O method, and use 7 bits of it feeding a resistor network
(or maybe a custom A/D ladder network chip) and a simple comparator
(LM339?). The remaining 8'th bit would be used as a sensing input for
the comparator output. So it would work something like: a BASIC routine
outputs binary 0000001 to the resistor network, the comparator senses
whether the reference voltage is high/low compared to the input measured
voltage and causes the BASIC program to either increment the binary OUT
value, recompares etc, or logs it in an array and recycles round a loop.
Refinement of this would use the STROBE/READY capability of the printer
port, rather than the 8'th bit, thus freeing bit 8, to give an extra
degree of resolution.

The accuracy of this scheme probaby would'nt suffice for +/- 1 mV,
without some attention to an accurate reference voltage. Depends whether
you 'really' need accurate absolute voltage measurement, or just chnges
up/down. In it's simplest form, the idea above would just use whatever
the LPT1 outputs as binary 'high' which could be 4.3 volts, say, and
maybe not too stable. A refinement would be to use the LPT1 outputs just
as 'switches' to gate fractions of a really accurate reference voltage
to the resistor network. BASIC might also be a tad slow for 10
samples/sec (although perhaps not ALL that much slower). If you come
across any other ideas which DON'T require addition of a card to the PC
innards, but utilise software and basic homebuild electronic hardware, I
would be interested to follow up. (Access to specialised A/D converter
card gizmos is a bit difficult/expensive for the average home hobbyist
like myself in the UK!).
Regards
                        Jim

In article <31CB1B7E.205A@worldnet.att.net>, Eugene Abel
<EUGENEA1@worldnet.att.net> writes
>Charles Mulks wrote:
>> 
>> Not sure any of these groups (comp.sys.ibm.pc.hardware.misc,
>> biz.comp.hardware, alt.comp.hardware.pc-homebuilt) are the
>> right place to post, but gotta start somewhere.
>> 
>> I'm looking for an A/D card for an ibm-compatible.  Has to
>> handle sampling rate of 10/sec & slower (shouldn't be hard)
>> and input voltage range of +- 1 milliVolt max (might be
>> trickier).
>> 
>> Also need appropriate software, preferably for Win95.
>> 
>> Any suggestions?
>> 
>> Thanks,
>> Charlie Mulks
>> Mich. St. Univ. Microbiology
>
>One source you might try is Omega.  http://www.omega.com
>They sell a lot of Labratory quality interface cards for PC/MAC.
>They'll probably even send you a free catalog if you call. Huge hardcover 
>catalog!
>
>What are you measuring? Why do you need it to be so sensitive?
>Another alternative - RS232 multimeter. There are a number of accurate 
>multimeters which have 
>RS232 interfaces and come with datalogging / measurement software.
>
>I've been using one my company bought from Jameco http://www.jameco.com
>I'm using it to take temperature measurements of a component we are considering 
>using in our 
>products. It only cost about $120!!!  
>It uses a type K temp probe which can measure up to 1200 deg. C!  I haven't seen 
>many RS232 
>meters that can use these type of probes...
>
>Also, you might want to check out some electronics magazines.  There are always 
>tons of adds for 
>stuff like this...
>
>EAA

-- 
Jim Ball, Stony Stratford, Milton Keynes, UK. Fax: (44)1908-284248.