I’m looking at a project that would test multiple sensors using a GRV-EPIC-IV-24, EPIC 24 channel voltage input module.
Each sensor can have a different output, for example 0-5 VDC or 0-10 VDC.
The required accuracy is 0.1% so for the 10 VDC sensors, that’s 10 mV and the 5 VDC sensors, 5 mV. The module spec says it supports 0.1% accuracy for the configured range, so if the input channel is configured for 10 VDC, the accuracy is 10 mV, etc.
I don’t know which sensor will be connected to each input until just before the test so if an input is configured for 10 VDC and a 5 VDC sensor is connected, it won’t have the required 0.1% accuracy. If configured for 5 VDC and a 10 VDC sensor is connected, out of range.
Is it possible, programmatically, to change the input range at runtime before the test is performed so I can get the correct accuracy?
If it can be changed, is that automatically written to flash? Flash has a limited number of writes so I don’t want to do that.
If it can’t be changed or it is written to flash, I think another solution would be to allocate a set number of inputs to use 5 VDC sensors and the other to 10 VDC sensors. That would half the number of channels available per test so I’d need double the number of available channels.
I would test the GRV-IV-24 module carefully if using at the lower ranges. On the GRV-IVI-12 module for voltage input at the 1.25V range, the results are disappointing. The module does NOT meet the published spec due to signal drift. I found that the ambient temperature affects the modules reading. I have trends showing this affect below.
The reading drifts about 0.5mV/1°C which is unusable in my application. Opto22 has been made aware of the issue and has no fix at this time.
Purple is open circuit channel configured for 1.25V
Red is shorted channel configured for 1.25V
Blue is shorted channel configured for 10V
Bottom chart shows the ambient temperature
We need 26 input channels in total, 25 for the sensors under test and 1 for a master sensor used to verify the others.
I think you are on to something. Instead of (2) IVI-24’s, we can use (4) MM1001-10’s. That will give us 32 channels, more than enough.
I didn’t mention it before, but the sensors can also have mA output. I was going to use the IMA-24’s for that and they do have the required accuracy so 2 of those would also be plenty.
Thanks, that’s good to know. I think the smallest sensor range is 5 VDC so I don’t think this will be an issue. I will confirm that though before continuing. If they do have sensors at that range, we may have to include a mV input module.
Yes, the mV input module is a better option and doesn’t have the issue I reported. I haven’t tested the MM1001-10 at the mV level so I don’t know on that one.
At first I thought you were measuring mV level signals, for the voltages you are looking at >5VDC, you will be fine.
Thanks, @JeffS. The OptoMMP guide is very helpful and extensive. The things I wasn’t sure about was if the range change would take affect immediately after writing to the memory map and if the change would automatically cause a write to flash. The answers are probably yes and no, but I wanted to be sure.
