search

UCLA-Rocket-Project / OLD-Ares2022-2023

Central Repository for Ares Software
3 stars 0 forks source link

Perform calibration for all PTs to be used in Fall 2022 tests and operation. #10

Closed harrisonCassar closed 1 year ago

harrisonCassar commented 1 year ago

Motivation/Description

After completing #7, we need to actually calibrate the PTs needed/desired by Prop, obtaining the calibration values used by our internal ADC C code to convert raw PT voltage values to desired pressure readings (in PSI).

Proposed Workflow

Work with Propulsion team to complete calibration. From software side, we expect:

Following this calibration, verification evidence checking the sanity of these results should be collected/posted.

Blocked by #7.

harrisonCassar commented 1 year ago

Working with Connor on Propulsion, we attempted initial calibration on 10/21, but delayed due to the need for DAQ re-wiring.

On 10/26, completed full calibration of 9 PTs. The details are as follows:

ALL CALIBRATIONS DONE AT BUCK OUTPUT 5.01V (NO RPi plugged in, only battery!)

IN0: C
IN1: Fuel 
IN2: CC
IN3: Low He 
IN4: LOX PT
IN5: D
IN6: Purge (250 PSI)
IN7: Connor (5k?), and then switched with F (1k) to calibrate separately
IN8: High He (5k)

Calib 1K PTs (0-5, and 7 --> Connor, 5k), ~30 seconds
0 - Figure it out!
50 - 1666823840249
100 - 1666823907804
150 - 1666823983271
200 - 1666824043348
300 - 1666824115227
400 - 1666824169162
500 - 1666824234394
600 - 1666824289554
700 - 1666824361761
800 - 1666824422898
900 - 1666824476504
950 - 1666824528557 (less than 30 seconds oops!)

Calib Purge 250 PT (IN6):
0 - 1666825107747
50 - 1666825174462
100 - 1666825232563
150 - 1666825284528
200 - 1666825345033

Calib F 1k PT (IN7, second plug-in), only 20 seconds:
0 - 1666826248828
50 - 1666826327373
100 - 1666826368275
150 - 1666826406714
200 - 1666826448472
300 - 1666826486335
400 - 1666826525327
500 - 1666826564348
600 - 1666826604657
700 - 1666826648527
800 - 1666826685719
900 - 1666826753445

Calib High He 5k PTs (High HE IN8), only 20 seconds:
0 - 1666827245998
200 - 1666827339391
300 - 1666827383449
400 - 1666827420579
500 - 1666827468037
1000 - 1666827512902
1500 - 1666827573524
2000 - 1666827666456
2500 - 1666827735556
3000 - 1666827779518
3500 - 1666827825569
4000 - 1666827873684
4500 - 1666827919069
4900 - 1666827994986
harrisonCassar commented 1 year ago

The "manual" workflow for computation of calibration values are done as follows:

20-30 seconds of values at a certain pressure reading are recorded and averaged over (to avoid potential noise/spikes from dominating sample measurements). We then repeat this process for all channels (that are currently being calibrated), as well as all pre-defined/user-defined pressure readings.

We then linear fit the results, and produce a, b belonging to (a*x_avg + b).

harrisonCassar commented 1 year ago

Computations for IN0-IN5:

Averages over 20 seconds of data collection time: image

Plotted (with linear best-fit): image

Extracted best-fit values: image

harrisonCassar commented 1 year ago

Computations for IN6:

Averages over 20 seconds of data collection time, plotted (with linear best-fit), and extracted best-fit values: image

harrisonCassar commented 1 year ago

Computations for IN7:

Averages over 20 seconds of data collection time, plotted (with linear best-fit), and extracted best-fit values: image

harrisonCassar commented 1 year ago

Computations for IN8:

Averages over 20 seconds of data collection time, plotted (with linear best-fit), and extracted best-fit values: image

harrisonCassar commented 1 year ago

NOTE: Above images plot x vs y, and NOT y vs x; so we re-compute for the ACTUAL desired calibration values, and summarize them as follows:

image

For the mapping defined by:

IN0: C
IN1: Fuel 
IN2: CC
IN3: Low He 
IN4: LOX PT
IN5: D
IN6: Purge (250 PSI)
IN7: F
IN8: High He (5k)
harrisonCassar commented 1 year ago

From Connor on Propulsion, sent in Slack on 10/27/22:

For reference, IN0 (C) is going to be renamed "LOX Fill"
IN1 & IN4 should now be named "Fuel Tank" and "LOX Tank" respectively
harrisonCassar commented 1 year ago

Updated the ADC source code to include these values, and ran a sanity check at ambient pressure to confirm approximate accuracy (ideally, all should be ~0 PSI, but due to likely expected artifacts of our calibration, reasonable inaccuracy exists at extreme values).

image

Supporting Leak Checks on 10/27/2022 also confirmed reasonable accuracy for higher pressures up to 500 for 1k PTs, and 1300 for 5k PT.

harrisonCassar commented 1 year ago

Given the completion of this round of calibrations for all PTs that will be used in Fall 2022 tests and operations, we determine this calibration campaign as complete.

~Passing to @philipdoucla for concurrence of closure before closing.~ EDIT: Before aiming for closure, let's also update the ADC source code to have these newly-updated values. This depends on #7 being completed.

harrisonCassar commented 1 year ago

In preparation for 11/5 Hotfire operations, a re-mapping of these calibrated PTs was performed for active system use (as summarized in #27), and is presented here explictly:

CALIBRATED LABEL --> UPDATED LABEL/USE C --> LOx Fill Fuel --> Fuel Tank CC --> CC Low He --> Low Press LOX PT --> LOx Tank D --> LOx Manifold Purge --> Pneumatics F --> Fuel Manifold (Connor 5k not used) High He --> High Press