LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Gain error with shunt calibration

Hi Anton,

 

That is strange. With no amplification I don't see the point in setting any limits at all. Anyway, that is a different topic that has not been a problem.

 

I made a bridge out of four 120 ohm resistors and did some tests with jumpers again. The shunt factors from five tests with each jumper configuration are:

 

no jumpers

1.066, 1.075, 1.070, 1.073, 1.068

 

jumper EX- to RS-

1.011, 1.014, 1.010, 1.020, 1.010

 

jumpers EX- to RS- and EX+ to RS+

1.007, 1.007, 1.012, 1.020, 1.010

 

jumper EX+ to RS+

1.071, 1.069, 1.069, 1.073, 1.080

 

Clearly, something happens when EX- is connected to RS- while a jumper between EX+ and RS+ seems to have no effect on the shunt factor. In this case I cannot compare the gain to any physical quantity, but the strange behaviour is still there and this configuration has only a few centimetres of wire to take into account.

 

/Emil

0 Kudos
Message 21 of 24
(879 Views)

Hi Emil, 

 

Excuse me for my late reply, but I have finaly gotten a test system to work and I have been trying to reproduce the effect that you are seeing. I have however not been able to reproduce the effect that you are seeing. There are some fluctuations in my shunt factors, but they are very small (only +-0.001) and I can see no correlation to how my jumpers are connected. I have used 500 Ohm resistors, but it should not make any difference.

 

This is how I think it should be, but I can not understand why you are seeing a consistently different behaviour. This is indeed very strange. I understand that you have several modules; does all modules show the same effect? If you measure many shunt factors in sequence; how much does the result change? Is there anything with the cabling that could cause a change in the connection or contact resistance when you apply the jumpers? Could you post a picture of the bridge setup that you have made? 

 

Which version of the 9237 do you have, is it the cat RJ50 or D-SUB? Have you connected the SC leads? 

 

Many questions, but we will need to find out what we are missing here. 

 

/Anton

0 Kudos
Message 22 of 24
(860 Views)

Hi Anton,

 

All modules show the same effect. The five shunt factors for each combination in my previous post are measured in sequence on the same channel with a few seconds interval. The 9237 modules are D-SUB with an added screw terminal block and I have one end of the jumper connected to the EX terminals all the time to avoid changing the contact resistance. SC+ is always connected with a jumper to AI+ in my measurements, but SC- is the same terminal as RS-.

 

I couldn't get a good picture on the wires with my mobile camera, so I made a drawing instead.

 

/Emil

0 Kudos
Message 23 of 24
(843 Views)

Hi Emil, 

 

Now I have an idea to what this behaviour is, finally! I have been using the RJ-50 version of the 9237 and there is difference as to how the wires should be connected for these two. The SC terminals are wired to the internal shunt resistance in the 9237 and for the RJ-50 there is two different terminals for these. On the D-SUB version however one of the RS terminals are on the same pins as the SC- and therefore if you have not connected the jumper between EX- and SC- you can not make a shunt calibration since the shunt resistance is not wired. Please see the attached picture.

 

So in conclusion you will need to use a jumper between RS- and EX- on the D-SUB version of the 9237 in order to be able to make a shunt calibration. 

 

This explains why you get different shunt factors when you have connected jumpers and it agrees with your measured data. This however does not explain why you are getting a worse reading after calibration. This issue I think is in part due to the wire resistance, but it still feels like there is something we are missing. 

 

/Anton

0 Kudos
Message 24 of 24
(824 Views)