overvoltage protection

[Gogo-1962]

I need to measure small currents (0 - 10 mA) with PCI 6220 M series board. However there are also short cirquit currents of 0.5 A. I am using a shunt resistor of 20 Ohms but sensitivity is not enough. Going to a higher resistor value will require overvoltage protection or overcurrent protection. The smallest fast fuses are rated at 32 mA and at 120 mA will blow for a max time of 100 ms. Will e.g. 100 V applied for 100 ms at the differential input damage the board? Any alternatives for overvoltage protection which will not jeopardize accuracy of measurement? Please advise me for scheme resources as I am not a hardware guy. 

[AnalogKid2DigitalMan]

100V @ 100msec = POOF!

Manual states Max working voltage of +/-11V from AIGND, +/-25V overvoltage protection with DAQ powered on +/-15V powered off

What resolution do you require? This is a 16 bit card (65536 steps of discrete resolution). Worst case resolution would be using +/-10V input range. 20V/65536steps = 0.000305V/step.

0.000305v/20ohms = 0.000015A resolution (15uA). Not too bad, probably well below your noise floor. Things only improve if you use a lower voltage input range (+/-5, 2, and 1V available).

So with 0.5 short across 20 ohms sense resistor yields 10V drop, within spec of the DAQ card inputs. Can your setup allow a 10V drop across the resistor without affecting other things?

From the info you posted, I think you are OK. What do you think?

[Gogo-1962]

I need to measure with 2 uA accuracy. This is possible with 5 Hz filtering at home, but not in the noisy Lab environment. So I need to increase the 20 Ohm resistor to improve the S/N ratio. Subsequently I need an overvoltage protection to protect the board in cases when the 0.5 A currents flow.

[AnalogKid2DigitalMan]

For 2uA accuracy, then off the cuff you need 0.2uA (200nanoAmp) resolution!

Personally I would suggest not using a DAQ card. Go with a lab quality DMM (Fluke, Agilent, etc) with proper specs on the current functions outfitted with a serial or GPIB interface. You wont have to worry about overvoltage issue with 0.5 amp short conditions.

[Gogo-1962]

Yes, but I need to measure 2 additional voltage channels simultaneously. Even the NI cards can not get synchronized, e.g. their DMM card with the DAQ card. 2 uA is actually the resolution needed.

[AnalogKid2DigitalMan]

Gogo-1962

Resolution is not equivalent to accuracy.

What type of voltage source is supplying the current you need to measure? Is it really capable of supply the burden voltage drop across the sense resistor? For example if you use a 200 ohm resistor with 0.5 amps of current, that is a 100V drop. If it is a 12V supply for example, you will never come close to generating 100V across the sense resistor.

You did not mention the need to measure other voltage channels simultaneosly, thus I threw out the DMM route.

Can you provide a clear description of the entire task and hand and the measurement requirements (sampling rates, AC or DC signals, voltage ranges, accuracies/resolutions required, etc)

[Gogo-1962]

The task in short is to characterize organic LEDs. One channel is the driving voltage in the range 0 - 60 V which through 1:10 devider is entering one channel of the board. The second channel is from a photodetector with 0 - 100 mV range. The third channel is the current through the LED. Typically the LED starts around 15 V with 5 mA current. But sometimes the silver paste that is used for contacts can give a short circuit that reportedly can produce no more than 0.5 A current. The power supply is a pretty big one with manual controls for the voltage and some 10 A capability. The current can be limited manually but not to a very small value. I do not see the relation between voltage drop on the resistor and power supply provided that the power supply can output 0.5 A. On 200 Ohms it is 100 V. So the measured signals are DC, the aquisiotion rate is once/second.

Some friends suggested using in the current channel a diode after e.g. 10 kOhm resistor to a 10 V supply, produced from the +12 V of the computer with a tsener diode and e.g. 200 Ohm resistor. So if the voltage excedes 10 Volts the diode will open and the current will flow through it and not through the board. But this probably will introduce noise, there is some leak current, and the system will need callibration. And I am not a hardware man. For any ideas I will be very thankfull.

[Jeff_T.]

With regards to the relationship between the power supply and and the resistor.  In a short circuit, very little voltage is required to drive 500mA.  For a power supply to drive 500mA of current across a 200 ohm resistor, the power supply must generate 100V to push that much current across that much resistance.   AnalogKid was referring to the volt = current * resistance relationship.  If your power supply cannot generate 100V, then it cannot push that much current across a 200 ohm resistor anyway so you don't have to worry about  getting a 100V drop across the resistor.