05-07-2012 08:49 AM
Looks like stray capacitance is affecting the resistance measurement I'm trying to make with the PXI-4070 DMM. Any ideas on how to improve my measurerment?
Thanks in asdvance.
05-07-2012 10:04 AM
How should we know what might be improved if you provide no information on what and how you measure the resistance?
Cablelength? 2 or 4 wire? Range? Method?
Is it a resistor, sensor, cap, batteries, liquid, skin, brain 😉 ? Stable or dynamic?
05-07-2012 10:49 AM
In addition to what Henrik said, the obvious answer is to reduce the stray capacitance.
Unless you are measuring very fast and very high resistances, I doubt stray capacitance is the problem.
Please tell us what you are trying to do, in some detail, along with why you think it is not working. What kinds of effects are you seeing? Why are they problems?
Lynn
05-08-2012 08:24 AM
Hi Folks
I'm working with VI Joe on this and here's some more info on on the issue:
This type of problem is typically solved by adding a delay to the measurement. This forces the DMM to apply its test current for a period of time to charge up parallel capacitance and allow the test current to stabilize - with the caps charged the only current flowing will be through the resistor.
We have tried adding niDMM Configure Trigger to the vi and specify a Trigger Delay (sec) but this doesn't help. Using an external o-scope we have observed that the 4070 DMM is applying a test signal for the specified delay times (we've tried a few) but still gives us 8k results. Also, the scope capture does indicate the typical RC time constant charging of a cap. I want to believe that the 4070 makes its measurement at the END of the delay and not the beginning.
Any insights into this type of measurment using the 4070?
05-08-2012 10:48 AM
Hey jimvi,
Thanks for the additional information on the measurement. In the future, you may want to consider posting to the DMM forums as you will likely get a quicker response. In circuit measurements can be a little tricky sometimes, especially when using auto-ranging. The 407x DMMs has a 10 k and 100 k resistance range, so when you pick 20,000 it get's coerced up to 100k. Have you tried using the 10 k range?
What makes the measurement tricky when using auto-ranging is that anytime the DMM changes resistance range, it can change the test current applied to the DUT. When charging a parallel RC circuit, the DMM starts off seeing a very low resistance because of the discharged cap, but once it starts charging, the resistance "seen" by the DMM increases and the DMM range-changes to a higher range. When this happens the test current becomes smaller and the cap begins to discharge through the parallel resistor. The DMM then begins to "see" a lower resistance and range changes back to a lower range. Sometimes this cycle goes on indefinitely and other times the DMM will settle to a final range; it's really dependent on the DUT.
Have you tried taking a resistance measurement in Soft Front Panels in the 10 k range? Do you still see 8k after letting the measurement settle out? Are you expecting to see 10k or are there other parasitics in your circuit that could be skewing your measurement? Let us know what you find when you get a chance.
For reference, below is a table of the test currents used for each resistance range of the 4070 DMM.
Thanks!
Brandon G
05-08-2012 02:14 PM
Thank you Brandon G, this appears to have fixed our problem.
I've inherited some of the code here and I'm just now diving into the LabView realm. This is where the 20k range came from.
One thing of note, the measurement passes the first time through but if I step into this measurement and run it multiple times, it fails after the first time. We get about 8.3k instead of 10k and it continues to fail on subsequent tries. Not sure what this is, maybe we're fighting the cap discharge.
Thanks again for your insight.
Jim
05-08-2012 03:35 PM
Subsequent runs may indeed fail due to the cap holding a charge.
If this circuit is unpowered, here is a little trick I devised and have used in the past:
When connected across the R/C combo, write code to measure DC current (Amps) and wait until current has approached near zero. This should not take long at all. Current measurements use a low impedance shunt which will act to discharge the cap. Then switch the meter function over to ohms and quickly obtain your resistance readling.
-AK2DM
05-08-2012 04:00 PM - edited 05-08-2012 04:04 PM
Another point to ponder and a little known gotcha in NI DMMs.
when configured for resistance the test voltage is NOT applied until the FIRST measurement is triggered. Conversly, the test voltage STAYS ON between measurements Even if you change configuration: (e.g)
A better plan:
So if this is a problem
We have tried adding niDMM Configure Trigger to the vi and specify a Trigger Delay (sec) but this doesn't help. Using an external o-scope we have observed that the 4070 DMM is applying a test signal for the specified delay times (we've tried a few) but still gives us 8k results. Also, the scope capture does indicate the typical RC time constant charging of a cap. I want to believe that the 4070 makes its measurement at the END of the delay and not the beginning.
you simply need to take a dummy resistance reading to make sure the source is ON at the start of the arming cycle so that it is applied for the delay.
05-10-2012 09:28 AM
Jeff
Thank you for this insight.
It looks like we are going to have to carefully plan our sequence of resistor measuring - should we embark on this. Typically we verify shorts which tend to trump anything measurement-wise. Verifying resistors in-circuit is obvously more of a challenge.
What is a good way of restting the DMM?
05-10-2012 10:14 AM
@jimvi wrote:
Jeff
Thank you for this insight.
It looks like we are going to have to carefully plan our sequence of resistor measuring - should we embark on this. Typically we verify shorts which tend to trump anything measurement-wise. Verifying resistors in-circuit is obvously more of a challenge.
What is a good way of restting the DMM?
works pretty well