LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Multiple charts Data Mixing

Solved!
Go to solution

Hello,

 

I am working on a system that can multiplex through a bunch signals while using a single voltage input channel on an NI 9205. I would like to be able to display each signal in separate graphs. I have verified with LEDs that the MUX part of my code is working properly. The PWM code is also working but is not relevant to this problem.

 

I am trying to take the RMS of each value, but even without doing an RMS calculation, the data from previous iterations leaks into the "ADC Data Input 2" chart and throws the data off or something. In figure 1 (attached), I have a 5V source plugged into channel 3 of the MUX and the rest of the channels on the MUX are left open. For some reason channels that are not left open are accurate, but as soon as they are left open, they show anything from 0-5V. I checked with a multimeter and found that the pins left open are at 0-0.1V. If I give each channel of the MUX a 5V or GND, then all of the charts show the correct values.

 

Is it possible that the buffer is not clearing and taking data from previous iterations of the while loop? I have attached the code.

 

Please let me know if this is not clear enough.

 

Thanks,

LittleBush

Download All
0 Kudos
Message 1 of 4
(935 Views)
Solution
Accepted by topic author LittleBush

If you are reading open channels in between other actual signals, it sounds like you are seeing what is known as "ghosting".  With multiplex A/D converters the charge of the previous measurement is left in the circuit.  When you go read the the open channel, it is a very high impedance input, and thus there is nothing to allow the charge amplifier to float to any other voltage.

 

Google it.

 

Ignore the data from open channels, or put a wire between the analog input terminal and ground so you have a zero impedance path to drain the charge.

Message 2 of 4
(890 Views)

Thank you for your reply RavensFan! After googling ghosting, it makes sense that this is probably the issue. I will try to find a way around this phenomenon.

 

Thank you,

 

LittleBush

0 Kudos
Message 3 of 4
(884 Views)

RavenFan was right about the ghosting. It was an issue. So, I minimized the effects of ghosting by only measuring signals that were fairly close to each other in voltage. Another issue I had that caused the ADC to appear to not read the correct signal at first was that I was using the wrong RMS function (pt by pt). So I was accidentally mixing data that way from previous iterations with the ADC. I needed to use the RMS.vi in order to only perform the RMS calculation on the data from a single set of samples.

 

Thanks RavenFan!

 

LittleBush

0 Kudos
Message 4 of 4
(836 Views)