hwos V2.15 bug reports

[Ralph]

Update: new findings 5) and 6), no.6) may be a potential safety hazzard, see below

~~~~~~~~~~~~~~

Hi,

i just got my new OSTC-cr (#4422) a couple of days ago and while playing around with it i discovered a few bugs:

1) In CC mode, while editing the DIL gases, in submenue Change depth: making a 'reset to MOD' does not change the depth of the DIL gas, BUT changes the Change-depth of the setpoints! In particular, depth of first DIL gas overwrites depth setting of first setpoint, second DIL gas second setpoint and so on.

There are two anomalies between doing a simulation and calculating a deco plan:

2) You can set a dive time and a depth before entering simulation as well as deco calculation. The set depth is regarded in both cases, but while in deco calculation the deco is calculated for the point in (run)time set before, the simulation disregards the time set before and always starts with a runtime of 0 mins.

3) When doing a simulation, the autoSetPoint settings are regarded - in my test case the setpoint shown in the lower left is 1,30 although 1,30 is my 5th setpoint i.e. auto setpoint switching did occure. When doing a Deco Calculation the used setpoint is not directly shown, but the resulting deco schedule looks like calculations stick with the first setpoint (0,70 in my case) i.e. not regarding the AutoSetpoint function beeing activated. This makes the deco calculation function a bit useless as one wants to know how the real dive would look like (with diving at depth with a propper high setpoint). Putting the high setpoint on the 1st setpoint entry doesn't make sense as at the beginning of the dive the ccr will be on low setpoint and you won't want your Computer to assume more oxy in the loop than there is in reality. Needing to reconfigure the 1st setpoint any time one toggels between deco calc and real diving would be no fun.

4) Compass calibration / compass readings. I don't know if it is a bug or the limits of the Hardware or something else but after calibration i encountered the following situation: main directions N, E and S where correct, but just W was about 20° out of its direction. That is between the indications S and W i had to turn the Computer by about 110° instead of 90°, and between W and N it had only 70° mechanical turning angle. How do you calibrate the compass? In shipping, a Deviation table is created that holds the correlation between compass "needel" direction and true magnetic direction per each 10° step. By this way deviations that are not uniform ('Feldverzerrungen') can be handeled. Nearby metallic parts (like the battery) always create non-uniform deviations. I could reproduce the above with other single main directions beeing off axis while the other 3 were correct - just by doing new calibrations with other turning sequences and while in different positions e.g. inside/outside of the house.

5) CC mode, lost gas handling: if choosing Bailout gases, then Lost Gas you get the Bailout gases and can set a gas to disabled - all fine. But if choosing the DIL gases, then Lost Gas you do not get the list of DIL gases but again the list of Bailout gases and will subsequently disable one of your Bailout gases instead of a DIL gas. If one has the same gas type configured in both gas lists (e.g. DIL gas onboard, same gas offboard in a stage with 2nd stage reg) the bailout deco schedule may be impacted by not making use of a gas that in fact is available - not good if you are in an emergency...
Side remark: when setting a gas to 'lost' while beeing deeper than its MOD, the gas will be colored red in the list. After disabling it, it will still be coloured red. The only thing that happens is the depth setting which turns from something to zero. If you disable e.g. the OXY gas, the tiny figure '6' becomes '0' - quite hard to see, especially when you're under water and in light stress because you just lost equipment. It would be better if the disabled gas would change its color to the standard disabled-color or at least something more visible happens (e.g. a 1 sec pop-up message 'gas disabled' or alike).

6) ##### potential safety hazzard #####
During testing the functions of the computer by making various simulation runs, deco calculations, (re)configuring gases and settings sometimes I got wired results. First i thought it's due to my limited knowledge of the Computers behaviour so far. But then (not reproducable yet) the computer calculated a deco schedule and within the gas consumption results of the bailout plan it used the names of the DIL gases. Before that, i sometimes had the feeling that the Computer is confusing DIL gases with OC/Bailout gases as well while in Simulation mode. Up to now, I have not found the pattern that triggers this misbehaviour, but it happens. I don't know and can't test if this may also happen while in diving mode, therefore i declare this snag as a potential safety hazzard. To my impression, the root of this bug may be an initialisation error, a memory overlap or a stack overflow/overrun.


best regards,
Ralph Lembcke

[heinrichsweikamp]

Hi Ralph,

Thanks for the detailed post and sorry for the late reply. Bottom Time in the simulator is only valid for the deco planner, not for the real-time simulator. When properly calibrated the compass works quite well. Since all OSTC have the same compass routines I'm almost sure the calibration was not good. We have a short video showing the process since the description in the manual is quite short: https://www.facebook.com/heinrichsweika ... 679330933/ You can see the three axis there. Auto-SP is not computed into the planner currently, sorry.

I'll check lost gas handling and the change depth settings menu. I can confirm this bug and it will be fixed in the next release.

Issue 6: Do you get different results when you repeatedly compute the deco plan? I couldn't reproduce this here, sorry. A step-by-step description how you get the erroneous result would help a lot to locate this issue.

The planner itself is quite limited and was intentionally added to make a quick computation to compare the OSTC with an external dive planner. To make sure all gasses and settings are correct. It's not intent to replace a real dive planner.

But, nevertheless, it's absolutely required that the planner is error-free, of course. So we're very interested if there is an issue in the planner.

Regards,
Matthias

[heinrichsweikamp]

Please continue here: read.php?6,17930