As of May 2, 2022 the LabJack forums have been set to read-only.
If you are in need of support, please contact us via one of the methods described on our contact page.
Is there any filter by default on AIN0 ? On AIN0 I have a smooth curve than AIN1 and AIN2 where I have 0,003V step between almost each measure (see attached curve AIN0 and AIN1).
The T4 has a 12-bit converter. Across a 20V range that's a step of 5 mV. The data in the first graph is way better than I expect from AIN0-3. But looks about right for AIN4-11. The second graph looks about right for AIN0-3.
I have 3 sensor with 0-10 V output (range 0-1 mm) on AIN0, AIN1 and AIN2. The resolution of my sensor is 1 mV (0,1 µm), can I reduce this "noise step" of 5 mV? Why I haven't got it on AIN0 ?
I use streamBurst function to collect data and store it in CSV. I made those graphs with excel. I will investigate hardware side if there is difference between AIN0 and the others.
Is there a way to reduce this noise ? Resolution index seems to reduce it but with a lowest scan rate if I understood correctly.
The T4 is a 12-bit device. That means it has 4096 steps, or a resolution of 1 in 4096. That resolution is spread across a 20 volt span (-10 to +10), so the voltage resolution is 20/4096 = 4.88 mV. You can increase resolution index to improve effective resolution, and you can do smoothing in software, but do not expect 1 mV resolution. See Table A.1.3:
Effective resolution describes 1 standard deviation of noise and is also called the RMS resolution. Peak-to-peak noise would include every single reading, whereas effective or RMS includes 1 standard deviation, so it tells you the noise band of most readings rather than all readings.
When you see an effective resolution greater than the inherent resolution of the device, the meaning is mostly just statistical. So in this case where you see an effective resolution of 13.2 bits on a 12 bit device, it is telling you that your 12 bit data will be very low in noise, not that you will get greater than 12 bits of resolution.
The T4 has a 12-bit converter. Across a 20V range that's a step of 5 mV. The data in the first graph is way better than I expect from AIN0-3. But looks about right for AIN4-11. The second graph looks about right for AIN0-3.
I have 3 sensor with 0-10 V output (range 0-1 mm) on AIN0, AIN1 and AIN2. The resolution of my sensor is 1 mV (0,1 µm), can I reduce this "noise step" of 5 mV? Why I haven't got it on AIN0 ?
Perhaps some software is doing smoothing. What software are you using to collect the data and what software are you using to make those graphs?
I use streamBurst function to collect data and store it in CSV. I made those graphs with excel. I will investigate hardware side if there is difference between AIN0 and the others.
Is there a way to reduce this noise ? Resolution index seems to reduce it but with a lowest scan rate if I understood correctly.
The T4 is a 12-bit device. That means it has 4096 steps, or a resolution of 1 in 4096. That resolution is spread across a 20 volt span (-10 to +10), so the voltage resolution is 20/4096 = 4.88 mV. You can increase resolution index to improve effective resolution, and you can do smoothing in software, but do not expect 1 mV resolution. See Table A.1.3:
https://labjack.com/support/datasheets/t-series/appendix-a-1
To improve on this you need to move up to the T7 or T7-Pro.
https://labjack.com/products/comparison
Ok on table A.1.3 it says for default resolution index (5) there is 2,2 mV of effective resolution.
What is the difference between volatage resolution and effective resolution ?
Can I configure AIN0-3 to be on 0/10 V and not -10/10 V ?
Effective resolution describes 1 standard deviation of noise and is also called the RMS resolution. Peak-to-peak noise would include every single reading, whereas effective or RMS includes 1 standard deviation, so it tells you the noise band of most readings rather than all readings.
When you see an effective resolution greater than the inherent resolution of the device, the meaning is mostly just statistical. So in this case where you see an effective resolution of 13.2 bits on a 12 bit device, it is telling you that your 12 bit data will be very low in noise, not that you will get greater than 12 bits of resolution.