Reports 1-1 of 1 Clear search Modify search
DetChar (General)
takahiro.yamamoto - 20:18 Thursday 08 February 2024 (28555) Print this report
Sanity check of DAC saturation bit provided by a new model

[Yuzurihara, Yamamura, YamaT]

This work is related to klog#28498.

We checked behavior of a new model to detect DAC saturations
and confirmed that DAC saturation bit for ETMX could detect the saturation on DAC channels for MN, IM and TM actuators.
For ETMY, we will do a same check in other day.

-----
Before starting this work, Dsub cables to coil actuators were unplugged at the output port of AI chassis and all de-whitening filters were disabled.
In order to check behavior of online data quality state vector for the DAC saturation, we injected a DC offset from each COILOUTF and confirmed that saturation bit was surely changed.

Figure 1 shows the some time series signals when offset is added. Top panel and the 2nd top panel shows the given offset and an actual DAC output on the user model, respectivly. Note that value and sign of the actual DAC output are different from ones of the given offset because of the COILOUTF_GAIN.

The 3rd top panel and bottom panel shows an overflow counter for each DAC channel and online DQ state vector. We can see both two flag changed when DAC output becomes lower than -31490ct.

Same measurements were done for all channels and for both sign (positive and negative large count). Finally we confirmed that DAC saturation bit changed when absolute value of output count on any of DAC channel reached 31490ct. The reason why the threshold have been 31490ct not 32768ct is to consider the effect of the digital AI ripples.

Finally, we concluded a new model for providing online DQ state vector works well.

Images attached to this report
Search Help
×

Warning

×