Viewing ohms change on oscope???

Build or buy a precision current source to source 1mA, then use a multimeter. My cheapo sears multimeter can resolve 1 mV, and it only cost $30.

A fairly reasonable current source can be made with a TL431. Use a 1%

• V(cathode) - 2500.

You can make a window comparator that puts out a logic 1 if the input voltage is within a certain range. The resistor under test can be supplied by a precision current source to produce a voltage signal. Or, if the comparator reference voltage divider and your resistor are driven from the same voltage source and precision resistors are used, a supply with reasonably stable voltage regulation will suffice.

The comparator output can drive a simple timer circuit, or an LED if stopwatch precision will suffice.

It

the

need

resistance

to

as

seconds.

to

The device may have some kind of current limit, but 1 mA probably wont be a problem.

If you are afraid of higher currents, you'll probably have to amplify the resulting voltage. Use the following or something similar so as not to disturb the voltage across the device too badly:

A 100x amplifier will give you good results with 10uA. You could also use a sample and hold circuit to match the voltage before the test, and do a comparative measurement between the two during the test. That way, you can amplify 1000x or even 10000x, since you will only be amplifying the change, which should be less than 200uV at 10uA.

There is an autonulling dc lab amplifier schematic in Art of Electronics, second edition that might be perfect (perhaps a more than you really need...) for your application. The schematic is Figure 7.1 on page 393. Set the gain, null it out, and start the test.

The quick way to do this is to use a 9-volt battery or two and make a

Refinements: a zero-null pot, and a switchable 10-ohm cal check resistor.

John

You can almost make this into a Radio Shack project using the circuit below- the only matching required is on the two 1% 12.4K ohms resistors. As shown, the circuit will produce slightly less than 1mV per ohm change in your 1355 RUT to within a few percentage points, and you can gain this up as required. Adjust Rpot for Vout=0V while RUT is in quiescent state. View in a fixed-width font such as Courier.

. . +----------+--------+ . | + | | . | | --- . | Vref | /// V+ . | [12.4k] | . _|_ - | U1:B | . / \\---------+---------------------|+\\ . --- LM385 | | >--+-> Vout . | Q2 e +---|-/ | . | |/ | | | . | Q1 e---| 2x PN4250A | | | . | |/ |\\ | V- | . +--| c | | . | |\\ | +-[12.4k]-+ . | c | | . | | | U1:A Q3 c . | +-----+--------|+\\ |/ . | | | >---| Vref . [100k] || +--|-/ |\\ Vout= ---- x d(RUT) . | 100uA|| | e RUTo . | v| | | . | | +-----------+ _ . | | | /| . | [RUT] [Rpot] . | | /| . | (1355)| | . | | | . +----------+-----------------+ . | . --- . V- . .

That would be a dual single-supply OA running off +/-5 or so, low offset type preferable, Q3 is high gain.

Here is how you can gain that sensor ohm change by x10 with T-feedback around the output difference amplifier. It may not look like it, but this is a bridge. The "cal" switch is either an NC momentary or a jumper strap, open it and record the change in Vout to get the exact gain constant. Do this after you zero the output with Rpot. Rpot will be adjusted to be equal R.U.T. within 1-2% for zero, so you figure out the range you need. View in a fixed-width font such as Courier.

. . +----------+--------+ . | + | | . | | --- . | Vref | /// V+ . | [12.4k] | . _|_ - | U1:B | Vout . / \\---------+-------[12.4k]-------|+\\ . --- LM385 | | >--+-> 10mV/OHM . | Q2 e +---|-/ | . | |/ | | | . | Q1 e---| 2x PN4250A | | [68.1k] . | |/ |\\ | V- | . +--| c | | . | |\\ | +-[6.19k]-+ . | c | | | . | | | U1:A Q3 c [6.81k] . | +-----+--------|+\\ |/ | . | | | | >---| 2N5962 --- . |cal o| | +--|-/ |\\ /// . | --+ [10] | e . | o| | | | . | | | | | . | +-----+ +-----------+ . [100k] | | . | || | U1=LT1013 . | 100uA|| | _ . | v| | /| . | | [Rpot] . | [RUT] /| . | | | . | (1355)| | . | | | . +----------+-----------------+ . | . --- . V- .

Not sure if I'm missing something here - what's wrong with a 'meter with IEEE-488 bus and a PC so-equipped. Takes a few minutes to write the program with HPIV or VB and you can graphically display it. Alternative is to use a higher-end Fluke or similar with RS-232 bus. The rate at which you can read and precision of the reading is then a function of how much you spend on the meter.

Ken