Hi,
we are doing a design with the OPT3101 and we are having serious problems with temperature.
Our sensor board, which is already in the fifth redesign, is working very well at room temperature during short time after calibration (some hours). We have an electrical crosstalk of around 22-24 in all cases, and we have an illumination crosstalk in the highest HDR current level (173.5mA) at room temperature below 300 in most cases (we have built more than 1000 sensors already and we have a strong statistical database).
We are trying to measure a target at 3.5 meters with very low diffuse reflection. Our typical amplitude of the received signal ranges from 70 to 100 counts, which is low and we are aware of it. In fact we obtain that level because we have built our system with two SFH4550 in series with a supply of 5V for the diodes, following the instructions of SBAA303 Application note "Extending the Range of OPT3101 Systems" (with one single emitter, our level was below 40).
Our design is very similar to the OPT3101EVM, with the sensors and the photodiodes placed at a similar distance to the EVM, and with the same photodiode (SFH213FA) and LED (SFH4550) as the EVM
Our residue signal amplitude with the receiver occluded after performing the electrical and the optical crosstalk calibration is in most cases below 3 counts, and always below 8 in all cases (we have done the test over 1000 sensors).
According to all your documents, this should be considered a good design.
We have read your application note SBOA315 "Accurate Measurements with OPT3101" very carefully and we have seen that with a residue crosstalk of 8 counts and a signal amplitude of 70, we should have a distance error of 27cm in the worst situation.
But we have to operate the sensor outdoor, with a temperature range between 0ºC and 40ºC, and we are seeing that when we move away from the temperature where the sensor was calibrated, the detected distance of a fixed target set to 3.5m distance changes more than 40cm in most cases.
We have used a climatic chamber to try to characterize the optical illumination crosstalk with temperature, that you mention in the calibration document of the OPT3101 that should be done once per design, and we have discovered that each sensor has a different behavior regarding the optical illumination crosstalk against temperature.
Furthermore, we have seen as well that when we leave the sensor running during two days at the same room temperature (controlled in the climatic chamber) the illumination crosstalk residue, which at the calibration moment was giving an output of output 2-3 counts, moves to amplitudes of 30-40 counts without touching anything (same temperature, photodiode with the same occluder, etc)
We also bought two units of the OPT3101EVM, and we have inspected them with your latte python software, and although they should have the same calibration constants, they don't!! We have seen that one of the EVM modules has
IPHASE_XTALK_REG_HDR1_TX1 = 49
QPHASE_XTALK_REG_HDR1_TX1 = 16
and the other EVM has
IPHASE_XTALK_REG_HDR1_TX1 = 29
QPHASE_XTALK_REG_HDR1_TX1 = 10
As we see that even your EVM modules have different Illumination Crosstalk Correction Registers, we have implemented a method for saving the illumination crosstalk along a sweep the 40 degrees celsius. We have done that with 70 sensors and our surprise is that each sensor has a different slope and therefore this cannot be done as a per-design basis, but instead this implicates that we have to run a temperature calibration of each and every sensor. And even with that, since the illumination crosstalk changes along the time.
Another thing we have seen is that if we heat up or cool down your OPT3101EVM, the deviation is significant as well, although not so pronounced as in our system. Your system deviates 20 cm with the same reflector we are using.
We really need to put those sensors in operation but temperature deviation and time deviation is killing us.
Please give us some advice about how we can operate the OPT3101 in a commercial temperature range (0-50 deg celsius) and how we can stabilize the residue illumination crosstalk amplitude.
Regards,
Juan Ramon
