# Articulated Sled Commissioning Notes

## Mobilization

### Mechanical Installation

*  Sled loaded and plumbing checked.
*  Plumbing was backwards from assumed control logic, flipped connections.
*  Pilot pressure verified after failed attempt, tank dump line not attached.
*  Sled moved well in deck and vehicle hydraulics
*  Noticeable speed difference (faster) when on vehicle hydraulics.

### Electrical/Software Control Testing

* After some tuning, software functioned properly under input from the level sensor.
* Functioning gains were Kp=1 and Ki=0.005. This will likely be different under pressure and vehicle hydraulics.

#### Issues with the software:

* Control was not returning to zero when auto was disabled. `[Resolved]`
* QDoubleSpinBox has an unfortunate behavior that partially typing in a number emits the `valueChanged` signal. Unfortunately this is the same signal that using keyboard tracking also emits, so a fix will require a special derived class from QSpinBox to resolve the issue. `[Next Release]`
* Changed the default input to the loop control to the level sensor, we’ll see how it performs at depth. 


## Day 1 Tests

* 0905L Onsite at shallow water (90m) location
* 1027L After relocation due to crab pot and lots of lead dunks, we are on bottom at 231.7m depth.
* 1032L Vehicle stable on bottom, all systems functioning, setting up at 3m altitude for tests.
* 1053L Kp 2 Ki 0.016
* 1112L Kp 10 Ki 0.016 FOR KEARFOTT
* 1116L Kp 10 ki 0.016 FOR CALCPITCH
	+ test didn’t go well, it was highly oscillatory
* 1130L Test Complete. Log files TESTA closed
* 1130L Risi testing autocontrols.
* 11??L Recalibrated level sensor, adjusted gains for servoing on level sensor.
	+ Kp 2 Ki 0.025

* 1215L End log testb
* 1220L Set angle 27 and articulating vs gravity
	+ Kp 8, Ki 0.025, Kd = 0.5
	+ Results were good, still some consistent higher frequency oscillations we may tune out.
* 1326L Survey Complete, still in test C, so now we are servoing on level sensor. 
	+ Kp 2, Ki 0.025
* 1413L Survey Complete, breaking files
* 1416 Files broken, testD will be more tuning for kearfott
	+ kp 8 0.04 0.5 still the best for kfott control
	+ The cushioned caused serious problems when working against the ends of the ram, associated with the null of the ram being such a large portion of the dynamic range of our control (i.e. we thrust to 10 and the null is -6.7)