Custom scaling of data without MAX?

[uscmagz]

Okay here's my problem. Hopefully you can help me or offer some helpful advice.

I am creating a user interface for data acquisition (only analog inputs are used). Thermocouples (Type J and Type K) will be used, as well as factory calibrated strain gauges. Some other sensors that can be linearized on a 0-5Volt scale might be used too.

I originally had my program set up where the user had to set up their channels in MAX and then come to the program I made. There will probably be a lot of scale/offset changes made by the users so channel configuration needs to be flexible. I changed the program to custom scaling. Instead of virtual channels the user just enters scale and offset values. These values are saved to a text configuratio
n file and later used on raw voltage.

Then I realized I would have to have the user input what type of sensor they are using because I'd have to compensate for CJC in thermocouples. AND, I dont think the SCC-TC02 is linear.

*Am I right in thinking I will have to use case structures when I finally scale data for LINEAR versus J THERMOCOUPLE versus K THERMOCOUPLE?

*How innacurate would my readings be to use the Labview's Thermocouple scaler VI and set the CJC to 1.25 volts (25 degC) rather than reading from Channel X+8.

*Is there an easier way to do what I'm trying to do? I know people do custom scaling, how is it done?
THANKS.

I have:
SC-2345 Module Box
SCC-TC02
SCC-SG24
LabVIEW 6.1

[CoastalMaineBird]

I'm not sure I see where your difficulty is.

I have always used custom scaling, MAX never did quite enough.

I have a cluster for each channel. The cluster contains a scaler (EU/Volt), an offset (EU@0V) and a channel type enum, among other stuff (channel name, units, filters, etc.)

When processing the data, look at the cluster for that channel. Use a CASE statement on the CHANNEL TYPE value. If linear, multiply the volts by the scaler and add the offset - you get EU (Engineering Units) out. If the channel type specifies some other type, use the appropriate linearization routine.


How innacurate would my readings be to use the Labview's Thermocouple scaler VI and set the CJC to 1.25 volts (25 degC) rather than reading from Channel X+8.
<
p>
Depends on which type TC, and where you are on it's range. You might expect 20 degrees or more error at the high end of some ranges, without proper CJC.

There's no easy answer to that question.

[uscmagz]

Thanks for your advice.

How do YOU account for CJC? I have what seems to be a problem. I am acquiring in a continuous buffer. To scale data to display on a graph, I will need to factor in CJC, meaning I have to get the appropriate reading from Channel X+8. I was trying to use AI Sample Channel.VI to get the average of about 10 readings of the CJC channel while the data was being scaled.

Error -10608 occurs at AI Buffer Read.

I can't read from the CJC and from the sensors at the same time. Is there a way I can do this? Does my question make sense?

[CoastalMaineBird]

It's been a while since I did thermocouple channels, but the board I used had CJC as the last channel (It was an AMUX-64 expansion board, and CJC was on channel 31).
I just included channel 31 in the channel list, and sampled N samples, and averaged them all together (for each channel).
Then I used the CJC average in the linearization process.




I can't read from the CJC and from the sensors at the same time.
I don't understand why not. At least, if you define "same time" as meaning "in the same scan".




Error -10608 occurs at AI Buffer Read.

-10608 = "No transfer in progress for the specified resource". You have asked for a DAQ read on a task that hasn't been started. (Or has already finished and stopped
).