Problem of measuring 4-20mA using SCB-68.

[Bill Choy]

Hi all,

I'm using PXI-6220 and SCB-68 to measure 4-20mA current from 6 transducers. All transducer outputs are 4mA when they are no input signal. I soldered 6 240ohm resistors on the SCB-68 board from RC12 to RC22 to make it generate 0.96-4.8V. However, the measured voltages of these 6 channels are around 0.98V when the transducers are no input signal.

I measured the voltage across the resistors are 0.96V and I found that the extra ~20mV is the voltage different between point b & c (see attached photo "SCB68.jpg"). I used a DMM to measure point a and b that is 0.98V and it should be the voltage measured by the DAQ card. The table below is the changes of voltage different between point b & c when measuring different number of channels.

No. of channels     Vab
6                            18.85mV
5                            15.77mV
4                            12.6mV
3                            9.29mV
2                            6.12mV
1                            2.93mV

The Vab is decreasing when the operating AI channels also decreased. This effect makes my measurement inaccurate when the number of operating AI channels changed. Please help to suggest a solution.

Thanks,
Bill

[Bill Choy]

Sorry that there are some mistake of the post, the following is revised.

Hi all,

I'm using PXI-6220 and SCB-68 to measure 4-20mA current from 6 transducers. All transducer outputs are 4mA when they are no input signal. I soldered 6 240ohm resistors on the SCB-68 board from RC12 to RC22 to make it generate 0.96-4.8V. However, the measured voltages of these 6 channels are around 0.98V when the transducers are no input signal.

I measured the voltage across the resistors are 0.96V and I found that the extra ~20mV is the voltage different between point b & c (see attached photo "SCB68.jpg"). I used a DMM to measure point a and c that is 0.98V and it should be the voltage measured by the DAQ card. The table below is the changes of voltage different between point b & c when measuring different number of channels.

No. of channels     Vbc
6                            18.85mV
5                            15.77mV
4                            12.6mV
3                            9.29mV
2                            6.12mV
1                            2.93mV

The Vbc is decreasing when the operating AI channels also decreased. This effect makes my measurement inaccurate when the number of operating AI channels changed. Please help to suggest a solution.

Thanks,
Bill

[Henrik_Volkers]

Assume you have a parasitic resistance of about R_px=0.75 ohm due to your traces (or whatever) .

If all channels share the same ground and you have your prasitic resistors chained up you end up with that error (if ou measure voltage from Input Ix to GND)

GND --R_p6--a6--R_p5--a5--R_p4- ...   R_p1--a1

and

ax--R260--Ix

with Ix = 4mA  that fits  (R_p6 gets all currents -> greates error )

(I hate ascii drawing ....) and sorry for not using your point names..

Solution: Use a star ground topology  or differential inputs right across your R260

Message Edited by Henrik Volkers on 07-20-2006 02:23 PM

[Henrik_Volkers]

Ups, the shematic isn't correct to describe your error exactly, however might be somethng close to it.  Have to rethink about it when I find some time.

[Bill Choy]

Hi Henrik,

Thanks so much so your reply. I can't use differential inputs because I will use more than 8 AI channels later for my application.

I want to know what is star ground topolgy? How can I implement in the SCB-68?

Best regards,
Bill

[Henrik_Volkers]

You need to avoid the current flow in your voltage measurement traces. 

You have one common ground pin for your channels and n (>8) voltage input pins.

 As a first test try out the following:

Connect all current shunts (your 260ohm resistors ) at one point together with one cable for your SCB-68 ground and one (or multible) ground cable for your voltage source(s) to power your sensors. This is your star ground . It is important to keep this point as small as possible and of best conductivity.    

Now you connect your sensors and your input channels to the open (and as short as possible) legs of the resistors.   

Take a look at your shunt resistors: At 20mA the 260R will dissipate 0.1W and will heat up, so your temperature coefficient (ppm/K) can give you additional errors. If you have a choice, chosse bigger ones with better (best) temp. coeff.! Just hook one up to a good DMM and monitor R while you touch it.

I haven't used the SCB-68 nor have the time to read the manual, however a good board design for current measurements should realise it. (Are you shure you used the right pins to solder your resistors? )

Message Edited by Henrik Volkers on 07-24-2006 10:00 AM