Setting Input Resistance on the PXI-4071 using nimi-Python for niDMM

[Sparks!]

The LabVIEW NIDMM Property Node will set the input resistance on the NI 4071 to >10GOhms.

Using the following, does NOT 

import pyvisa
import nidmm
import time

NI4071 = nidmm.Session("NI4071")
NI4071.configure_measurement_digits(nidmm.Function.DC_VOLTS, 10, 7.5)
NI4071.input_resistance = 1000000000.0
NI4071.adc_calibration.ON
NI4071.auto_zero.ON
NI4071.settle_time = 2
NI4071.dc_noise_rejection.HIGH_ORDER

NI4071.initiate()
for i in range(1, 10001):
t = (str(NI4071.get_dev_temp()))
print(str(i) +', ' + t + ', ' + str(NI4071.read()))
NI4071.close()

Using SFP, I can set the input resistance as needed. {see photo}

[Sparks!]

NI I/O Trace of Python script.

6. niDMM_SetAttributeViReal64 (NI4071 (0x00000002), "", NIDMM_ATTR_INPUT_RESISTANCE (0x00118C4D), 1E+10)
Process ID: 0x00002914 Thread ID: 0x00001E84
Start Time: 13:51:27.0502 Call Duration 00:00:00.0000
Status: 0 (VI_SUCCESS)

Sure looks like it set the attribute . . .

Now I wonder why SFP gives different readings between 10Meg and >10G?

27. niDMM_Read (NI4071 (0x00000002), -1 (0xFFFFFFFF), 9.922156 (9.922156E+00))
On >10G 9.925564

Hmmmm . . .

[santo_13]

Please share NI IO Trace capture to ensure your python configurations are reflected.

[santo_13]

So, it does look like the 10G input resistance is indeed configured, please compare the SFP's IO Trace and replicate to obtain the same results.

[Sparks!]

TEST program

NI4071.initiate()
NI4071.input_resistance = 10000000000.0
for i in range(1, 11):
t = (str(NI4071.get_dev_temp()))
print(str(i) +', ' + t + ', ' + str(NI4071.read()))
NI4071.input_resistance = 1000000.0
for i in range(1, 11):
t = (str(NI4071.get_dev_temp()))
print(str(i) +', ' + t + ', ' + str(NI4071.read()))
NI4071.close()

TRACE calls

Hi Ohms niDMM_SetAttributeViReal64 (NI4071 (0x00000002), "", NIDMM_ATTR_INPUT_RESISTANCE (0x00118C4D), 1E+10)
Lo Ohms niDMM_SetAttributeViReal64 (NI4071 (0x00000002), "", NIDMM_ATTR_INPUT_RESISTANCE (0x00118C4D), 1E+10)

Same call gave same readings, from quite different code instructions. Looks like a bug; walks like . . .

[Sparks!]

"So, it does look like the 10G input resistance is indeed configured, please compare the SFP's IO Trace and replicate to obtain the same results. "

NOPE, they are not the same:

Choose 10Meg and DMM SFP calls for "1E+07"

Choose >10GOhms and DMM SFP calls for "1E+10"

And there is a difference in the measurements.

Hello Marcos . . .

[kirsch]

The attached text file shows 3 calls to configure NIDMM_ATTR_INPUT_RESISTANCE:

6. niDMM_SetAttributeViReal64 (NI4071 (0x00000002), "", NIDMM_ATTR_INPUT_RESISTANCE (0x00118C4D), 1E+10)

24. niDMM_SetAttributeViReal64 (NI4071 (0x00000002), "", NIDMM_ATTR_INPUT_RESISTANCE (0x00118C4D), 1E+10)

87. niDMM_SetAttributeViReal64 (NI4071 (0x00000002), "", NIDMM_ATTR_INPUT_RESISTANCE (0x00118C4D), 1000000.000000 (1.000000E+06))

Your test program only shows two. There are other inconsistencies. I don't think the attached text files corresponds exactly with the test program.

Would you mind attaching a full Python test program that reproduces the issue, and the exact I/O Trace calls emerging from that?

My guess given the information provided is that you are running into a reconfiguration bug in the 4072 driver and can be worked around simply, but I am not sure yet.

[Sparks! · Accepted solution]

Hi Marcos and thread,

Well, the only difference between the code {below 09.40.07} that ‘worked’ and the prior submission was the use of ‘1e+10’ instead of 10000000000.0

Then I tested using ‘10000000000.0’ and that worked (of course it would. {trace: 09.40.42) so perhaps removing code unrelated to this test?? I’ll check.

AND, the code {10.08.43} that combined the VDC reading and the thermistor did not set the input_resistance to >10GOhms.

That is a new can of worms, so I’ll post the code that worked and the trace with a ‘caution’ that there is work-in-progress . . . .

Thanks to all for your support,

Donal

Donal

[Sparks! · Accepted solution]

Put in YOUR for-loop:

NI4071.input_resistance = 1e+10

Volts71 = str(4071.read())

That works for me in a script that accesses two meters: one for DC_VOLTS and the other for TEMPERATURE. Use the 'Volt71' as needed, change the name etc.

WHEW . . .

[kirsch]

I looked at one of the screenshots (!) of your code. The forum supports codeblocks, much better.

import pyvisa
import nidmm
import time

## (Marcos) It's best practice to use the context manager (`with` statement)
## so that your session is automatically closed.
NI4071 = nidmm.Session("NI4071")
NI4071.configure_measurement_digits(nidmm.Function.DC_VOLTS, 10, 7.5)

## (Marcos) After calling configure_measurement_digits, the driver will handle
## all the low level settings in order to accomplish 7.5 digits. Don't set those
## yourself you are unlikely to get the desired results.

# (Marcos) This is non sensical. You aren't setting the value of the property.
NI4071.adc_calibration.ON

# (Marcos) This is non sensical. You aren't setting the value of the property.
NI4071.auto_zero.ON

NI4071.settle_time = 2

# (Marcos) This is non sensical. You aren't setting the value of the property.
NI4071.dc_noise_rejection.HIGH_ORDER

# (Marcos) If you use read(), then don't initiate(). Just call read().
# initiate() is typically used with triggering and/or multi-point or waveform 
# modes.
NI4071.initiate()

NI4071 input_resistance = 1e+07
print (NI4071.input_resistance)

for i in range(1, 2):
   print(str(i) +', ' + str(NI4071.read()))

NI4071.input_resistance = 1e+10
print (NI4071 input_resistance)
for i in range(1, 2):
   print(str(i) +', ' + str(NI4071.read()))

NI4071.close()