T7 AIN Problem | LabJack
 

T7 AIN Problem

7 posts / 0 new
Last post
Carlo
cd-tronic's picture
T7 AIN Problem

Hi,

we are trying to capture data of an acceleration sensor. It uses an AC-Coupled Measuermentamplifier to convert the data to a voltage value.

There seems to be an issue with the MUX causing current floating out of the AIN Channel what affects our Mesaurementsignal of the accelerationsensor (see LabJackStream.jpg) to an absolutely unusable signal. After many hours of testing and researching i figured out that there are strange things going on. 

I put an 8KOhm resistor between GND and AIN and measured the signal with an scope. See Inmage Measurement8K.png. This is how it looks like when "pulled" to GND.

I don't think that this is the normal behaviour cause that makes the Labjack absolutely unuseable. 

I also tried a nother T7 but get exactly the same results. 

Any Idea?

LabJack Support
LabJack Support's picture
The following app note

The following app note discusses settling:

https://labjack.com/support/app-notes/SettlingTime

Summary:  If your signal has a source impedance of 1k or less, it will settle within spec before the A/D conversion happens for every combination of Range and ResolutionIndex.

If you think your signal has a higher source impedance or otherwise is not tolerant of the analog input bias currents, try the LJTick-InBuff:

https://labjack.com/support/datasheets/accessories/ljtick-inbuff

 

Carlo
cd-tronic's picture
Hello,

Hello,

i have seen all that information before but had been confused measureing a current of 7uA coming out of the AIN channel while streaming. 

We are using this signal conditioing device: https://www.mmf.de/pdf/4-2.pdf (M29), the left one in the file. it is AC coupled over 10 µF; 220 kΩ parallel.

What do you think?

I just ordered the ljtick-inbuff in case there is no other thing i can do to make it work. Time is getting less on that project...

 

Cheers

Carlo

LabJack Support
LabJack Support's picture
I assume the 2 wires you get

I assume the 2 wires you get out of the BNC are connected to AINx & GND?

I will see if we can find sometime to set up a test with a 10 uF series cap and 220k to GND.

 

Carlo
cd-tronic's picture
Hello,

Hello,

 

yes. exactly. AIN2 and GND.

But i also tried useing AIN3 as negative input.

Thank you so much!

LabJack Support
LabJack Support's picture
Hello,

I was able to reproduce the signal seen in your Measurement8K.png image and adding an LJTick-InBuff eliminated the problem. Please try adding the LJTick-InBuff to your T7 and let us know if any problems persist.

LabJack Support
LabJack Support's picture
We looked into this a little

We looked into this a little bit further and have an experiment that may be worth your time to try if you are trying to avoid using a LJTick-InBuff.  We also were wondering, are trying to read multiple channels at the same time?

Our thoughts right now lead us to say that reading multiple AC coupled signals at the same time is not going to work correctly unless you use an InBuff.  The switching topology that we describe in our analog input settling time app-note makes it so that any non-driven or un-buffered signal needs to be providing current into the analog input channel in order to be properly read.  

Experiment: Without an InBuff, stream only 1 analog input.  AIN0 will do.  Check to see if the data is being read correctly and keep an eye out for the data becoming off-set.