hwOS Tech: 3.03 beta 1 released

[Ralph]

honestly, i don't know the OSTC4 code, but i know the "4" has a lot more CPU power than the "2s"...

[Ralph]

Hi Michael,

copied your settings and ran the dive as 45 meters, 18 min through the deco calculator:

18m 1' Nx26
15m 3' Nx26
13m 2' Nx61 <- that's the gas switch on that "unusual" depth of 13 meters which is the set change depth
12m 1' Nx61
9m 1' ...
6m 3'
4m 7'

Then repeated the dive in Simulator mode. Basically, the stops match pretty good with the calculator results, bearing in mind that the deco calculator computes everything with a 1 minute resolution whereas the simulator as well as the real dive mode work on a resolution of 2 seconds thus have less rounding effects.

When watching the deco stops, you can see the 12m stop frequently disappearing and reappearing again. That's because of the nearby 13 meter stop which while deco accumulates phasewise is sufficient to clear the tissues enough to be able to go straight to 9 meters. This effect is also boosted by the high delta between GF high at 85 and GF low at 30: That gives a high GF slope resulting in the M-line virtually "runnig away up" while the ascent. By the way, for that gases and depth i'd use 50/80 or alike, but that's a personal choice.

At 13 meters i got the gas change advice, and on switching was more or less instantly cleared for 12 meters. The last stop at 4 ran down form 7 to 0 minutes while in parallel the TTS ran down from 12 to 4 minutes. That offset is due to the TTS including the time for the final ascent, which is programmed to take 1 min per meter, so with last stop at 4 meters is 4 mins. You actually don't see these last 4 mins in the TTS as when the last stop is done, the display changes from deco stop + TTS to NDL mode, not showing the last minutes of the final-ascent-TTS any more.

Can you reproduce that? From your log, it seems the same sequence of events happened during your dive.

BR
Ralph

[Duglum]

Just a bit off topic about the GF's: In addition to the older DAN presentation by Simon Mitchell there was a nice little article by Dr. David Doolette a few days ago: https://blog.gue.com/gradient-factors-i ... ops-world/

So 65/80 or 70/85 might be even "better". But as Ralph already said: personal choice.

[Ralph]

Well, as long as we stay with air and nitrox, Bühlmann seems to not having been that wrong in the first place...

[Clownfish]

Ralph,

Thanks for your reply.

Yes, your simulation is approximately what happened BUT the last stop was not 7 mins it cleared after 3 mins, that is, it went from 4 mins to 0 mins in one go. This was not the TTS. You can see that in the profile I loaded.

The other questions are why the computer does not take into account your deco gases when showing deco while you are still on the bottom. It used to do this and enabled a more accurate understanding of what the deco requirement was (yes, I know that I should know what my requirement will be for the dive, I do - well I used to with the older firmware, this one is not yet that clear to me).

I dive this wreck a lot (more than 140 times over the past 28 years) so I know what to expect for my planned bottom times and my alternate bottom times.

[Ralph]

> The other questions are why the computer does not take into account your deco gases when showing deco while you are still on the bottom.

??? It does, for sure! If deco gases are picked when running the dive through the deco calculator, they will also be picked by the deco calculation while the real dive. It can't be different, because it's the same program code engaged in both cases, running from the same settings. The mismatch must be in some completely other area...

[Clownfish]

Okay, I just ran simulator for a very close dive to the one I did. That is, 45 m for 5 minis then 46 m for 10 mins then ascend at 18 mins which is 16 mins on bottom like I did.

At 18 m and 15 m stops, the simulation is very close to real dive. At the 15 m stop, the computer is showing last stop of 15 mins @ 4 m and TTS of 31 mins. When I get to 13 m and swap to 61% as planned, the deco @ 4 m drops to 7 mins and TTS drops to 18 mins. This is what happened in the real dive.

Therefore, the computer IS NOT taking into account the planned change to 61% when showing the deco and TTS before the change. This is the problem.

A few firmwares back and on all ones before that and on OSTC 2N, it took into account the planned change when showing the deco and TTS.

[geowas]

I really like the alternative layout in dive mode (layout big) as it focus on the relevant information. Looking forward to the changes there!

BUT: With my OSTC cR I can't use layout big in dive mode when diving CCR as the ppO2's of the O2 sensors aren't shown.

Diving CCR you have to see (allways) the ppO2 value of all 3 sensors in order to understand which is working correctly and which is not. To show just the derived ppO2 isn't enough as there are circumstances when the derived/used ppO2 for deco calc is just wrong!

So my feature requests for OSTC cR in dive mode would be:
1. Could you please add the ppO2 values of all 3 sensors (!) to the alternative layout big on a OSTC cR? E.g. as a row in the middle of the display like in the layout normal?
2. When diving with fixed ppO2 Setpoint it would be great to see the fixed ppO2 Setpoint instead.
3. Making the menu directly available in dive mode while using the layout big without changing back to layout normal (when pressing the left button for menu) would be good, too!

I hope at least #1 and #2 can make it in a new release!

Thanks a lot!
Georg

P.S. Just added this message as a new subject, too:
https://forum.heinrichsweikamp.com/read.php?6,21210

[Ralph]

please follow this topic / find the reply under the separate subject.

[Ralph]

Michael, if you like me to help, please follow my directions. Set your OSTC as you have described to me and then run a dive with the depth & time as i did write to you in the deco calculator. You should see a deco table whose stops change color in between. The colors reflect the gases. If all stops appear in the color of the gas set as 'First', then probably you havn't set the Nx61 to type 'deco'. Also the stop depths & times should appear as i have posted, +/- 1 minute here and there due to differences in surface pressure where you are and where i am, given that we both are at about sea level.

[Clownfish]

Ralph

When I run 45 m for 18 mins I get very similar to what you posted earlier but no 12 m stop at all. 18 and 15 m are red which is 26%, rest are yellow which is 61%.

If you run this using the "Start Simulator" option, it comes out as I noted in my previous post, that is, it shows TTS and deco till you actually change gas assuming you are not using a deco gas.

[Clownfish]

Ralph

Okay, I have now worked out why it did not work correctly.

In the gas setting for NX26, I did not have a MOD set (it was 0 m). When I Reset to MOD: 47 m and reran the simulator, it came out as it should, giving the deco taking into account the deco gas. Is this a bug or is it the way it should be?

[Ralph]

Wow - good job! That's something i did not expect to happen. In the past, the gas designated as First always had its depth set to 0, meaning "as deep as you want". But because of this, you couldn't really have travel gases. Now that's possible, but to make it work the first gas also needs to have a senseful depth limit set, zero makes no sense anymore. I think change depths of zero are color-coded in red to draw attention on. Will have a look in the code what consequences a depth setting of zero has, because actually at the moment i have no idea if it is a bug or bears some somehow useful functionality in it.

BR
Ralph

[Ralph]

Hi Michael,

i could confirm your observation from examining the code. After some thinking about it, it's not really a bug, but also bears no useful functionality. So i classified it unintended behaviour and added an exception handling to the code.

BR
Ralph

[ijdod]

Does 3.03B1 include logging tank pressure on TR models?

[Ralph]

Not yet, sorry. I'll urge Matthias to do it 'till the release latest...

[algo]

Did a dive yesterday with beta 1 3.03

This morning I waked up the computer and noticed that it shows a desat time of 9:50. I opened the logbook, left the menu and than it shows a desat time of 0:20. (10min later that same thing happened with 0:20 and than 0:10). So somehow the time seems not to be updated...

[Ralph]

Thanks! The desat and no-fly times are only calculated while the OSTC is awake, not while in sleep mode. While awake, they are (re)calculated every minute. So what you first saw where the old values from before the OSTC went into sleep mode, and then they got updated once the next full minute went by. I pimped the code such that after awakeing from sleep mode an update will immediatly be calculated to have current values right from the begin.

Because destat and no-fly times aren't exact science at all, they are only calculated in multiples of 10 minutes. So a jump from 0:20 to 0:10 is by intention. Moreover, as the calculation incorporates current ambient pressure which always goes up and down a very little bit, elsewise one would have the times going up and down a few minutes in zig-zag all of the time...

Ralph

[Kehrmaschine]

Hi Ralph

Any idea about when we can await the next stable version with the improvements of the current Beta 3.03?

[Ralph]

Well, programming works and "dry testing" for a beta 2 have just completed. The CCR parts went off to the red sea for a week of wet testing just yesterday evening. There has also been done a little bit on the TR specific code, which still needs to go through initial wet testing. And last but not least, my code updates need to be merged with Matthias' recent work on the compass part. So probably we will have a new beta out in about 2 to 3 weeks. That one should then be pretty final, but before releasing it officially as a release version we will give it some more time out as beta to eventually collect some more observations from the field.

Ralph

[algo]

Had the OSTC2 this weekend with me in Switzerland for diving in the Lake Marmorera. I don't know if this is related to diving in mountain lakes and this beta version, but i noticed the following:

When starting the dive it shows a ambient pressure of 860mbar, which sounds reasonable to me. At the begin of the dive the OSTC did not went to diving mode but stayed in surface mode with an increasing ambient pressure for some minutes. The pressure increases up to approx. 1200 - 1300mbar and than the OSTC went to diving mode and showed a depth of 3.5m. The depth also sound reasonable to me.

I this normal behaviour? I never noticed such a behaviour (increasing ambient pressure before diving mode) in my "normal" diving spot here at the Lake of Constance...

BR,
Alex

[Ralph]

Hi Alex,

that's not part of the Beta but normal OSTC behaviour since long. Due to the reduced surface pressure, the OSTC has switched to altitude mode. In altitude mode, it requires more pressure aka more depth to automatically switch the OSTC from surface or sleep mode to divemode, because the OSTC simply doesn't know if you are diving or just driving down the road into the valley again.

BR
Ralph

[heinrichsweikamp]

Hi Alex,

The "high altitude mode" was introduced several years ago, not for altitude diving (This mode has no influence on your deco, which is based on the surface pressure anyway) but to prevent the unit going into dive mode when descending with a plane and having the unit active. Some people like to check the cabin pressure with the OSTC and ended up having a computer diving at 1m until the battery is flat (OSTC 2) or you remove the battery (OSTC plus/sport).

Regards,
Matthias

[algo]

Hi Ralph and Matthias,

thanks for the (as always) excellent explanation. And i can confirm, the OSTC ist doing the same thing at lake of constance level

Btw. i did now 6 dive with the beta (two deep one; three in mountain lakes; one long, shallow one) and discovered nothing unusual... its not working like a beta!

Alex

[Ralph]

Hi Alex,

thanks for your warm words! Indeed, there are a few shortcommings in the current beta we have discovered so far, but they are all to the safe side and an updated beta will soon be puplished here. That version is likely to be the release candidate then...

Ralph