New 1.84 beta available for the Mk.2

[heinrichsweikamp]

Hi all,

Finally, we uploaded a new 1.84beta for the OSTC Mk.2! It includes the important bug fix that caused a wrong average depth. This was related to a routine that violated it's max. CPU usage time (The OSTC Mk.2 uses a simple scheduled multitasking). Additionally, more features to the internal logbook were added - especially for multigas-divers.

Download it here: http://www.heinrichsweikamp.com/en/ostc_mk2/firmware/ - make sure to install the correct language HEX file!

So, die neue 1.84beta für den OSTC Mk.2 ist nun zum download bereit! Sie enthält einen wichtigen Bugfix in der neuen Deko-Routine, der zu Fehlern u.a. bei Durchschnittstiefe geführt hat. Die lag an einer Routine die zuviel CPU-Zeit benötigte (Der OSTC Mk.2 benutzt ein einfaches, präemptives Multitasking). Zudem wurde das interne Logbuch mit mehr Funktionen ausgestattet - insbesondere für Gaswechsel-Taucher.

Download wie immer von hier: http://www.heinrichsweikamp.com/de/ostc_mk2/firmware/ - achtet auf die korrekte Sprachversion der HEX file!

Gruß,
Matthias

[Solodiver]

Thx, the new scale on the logbook looks nice. For everybody else who wpnders how to find the additional information in den logsbook: press the left Button (SET/MENU) to toggle them.
There is still no indicator if it is a CC or OC dive, so adding the used deco model would be _really_ nice. It would even be better to see the set point(s) and maybe even more settings like GF...

Thx a lot,
Jan

[peppe.marsala]

Hi all,

just flashed 1.84 beta, logbook was checked for the new features and I noted a discrepancy with the avarage depth, please see attached files.
I also noted the new CFs 57 & 58, gas air consuption bottom and ascent gas... what they are used for? Where do they apply?

Thanks!

[peppe.marsala]

Another problem:

I also tried to simulate a dive with the following parameters:

Max Depth: 69 mtrs
Bottom time: 15 min
GF on
CF 32: 30% (default)
CF 33: 90% (default)
CF 55 (not sure): 1 (gas switch time)
GAS#1 active (air)
GAS#4 active (EAN40)

When I selected "calculate deco" the OSTC counted down to 1 min and theN freezed, orange led flashing quickly, "WAIT..." still there after 15 minutes.
Could it be because of the high PPO2, maybe beyond the max on CF18?

[dmainou]

Hi Peppe,

I haven't tried the new beta (I'm at work now as its already Monday)

Make sure you add the right info to the new SAC cf's (one for bottom gas one for deco)

The cf55 is to extend the stops at gas switches for n number of minutes.

Try to reboot the machine with the magnet and start again?

I hope you're OK

D

[Olivier]

Hi all, same trouble as Peppe: my mk2 is still stuck after 12 hours trying to calculate the deco time...

Peppe, how did you reseted it ? Magnet reboot ?

[heinrichsweikamp]

Hi,

For a reset, check "comment 2" here: http://www.heinrichsweikamp.com/read.php?2,4071

regards,
Matthias

[peppe.marsala]

Good day all,

the OSTC is still thinking and I'm still at home without any magnet

Matthias,

the two SAC were set to defoult (I noted the CF blinking in red sqare after flashing, excellent feature).
Oliver, I'll reset with a magnet later on.

Ciao!

[dmainou]

heinrichsweikamp Wrote:
-------------------------------------------------------
> Hi,
>
> For a reset, check "comment 2" here:
> http://www.heinrichsweikamp.com/read.php?2,4
> 071
>
> regards,
> Matthias


Hi Matthias

I think they upgraded fine but they run into trouble whilst trying to simulate a dive. So it may be a bug?

I'll try tonight.

[heinrichsweikamp]

Hi,

In the beta version, we do not reset any CF, so this may be the issue here. A new 1.85beta will follow, still some work to do until we can think of a 1.9 stable...

Regards,
Matthias

[peppe.marsala]

Matthias,

can you please stick the link to the picture showing the position of the reed switch? I do not find it with the magnet \

Thanks

[heinrichsweikamp]

It's attached to the link above: http://www.heinrichsweikamp.com/file.ph ... 7_reed.jpg

Note that this will _only_ work of USB power is attached.

regards,
Matthias

[Bardass]

same trouble for me when I start my first simulation (15mn / 45m in OC ZH-L16, one gaz)
i need to eject battery to stop calculating

works fine now after MD2 hash

[peppe.marsala]

Ok,

hard reset done, the OSTC came alive again.

I tried the simulator again with the same parameters and it stuck.
5 min, depth 15 mtrs and it stuck...
5 min, depth 15 mtrs only air and it stuck...
5 min, depth 15 mtrs only air, CF54 OFF and it stuck...
5 min, depth 15 mtrs only air, CF54 OFF, ZH-L16 OC and it stuck...

Now I have to return the magnet to the owner...

[dmainou]

Hi Peppe,

Did you reset the cf's before?

I just arrived home and will try now (magnet is ready)

D

[nakatomi]

Something else is odd in 1.84, if I remember correctly I observed (and reported) this with 1.83 already.

This happens using the simulator function built into Jdivelog, one active gas only (air)

1. While descending it will show deco at a very early stage (e.g. after diving straight down to 40 meters). A couple of seconds later the deco disappears and the no deco time reappears

2. When simulating a deco dive to e.g. 60m and 20 minutes bottom time, and completing the first stop, let´s say at 33 meters, the computer will send you to the next stop, let´s say at 29 meters. Shortly after that stop has been reached, the stop is again lowered to 33 meters and appears in red letters.

I suppose that the high descent/ascent rates in the simulator could maybe trigger this behaviour?

Did a lot of other testing with the simulator and the bug in 1.83 with the stops getting longer and longer, from one stop to the next, seems to have disappeared.
Also the CPU hang issue is much less of an issue. While the seconds still don´t rollover smoothly, it is much better than before.


Oliver
OSTC Mk2 1830

[dmainou]

Yep, same result. It hangs on the simulator.

Tried a number of things without success.

BTW, very nice graph and info on the logs.

D

[peppe.marsala]

peppe.marsala Wrote:
-------------------------------------------------------
> Hi all,
>
> just flashed 1.84 beta, logbook was checked for
> the new features and I noted a discrepancy with
> the avarage depth, please see attached files.

Ok, then I belive the sim has something wring
Have a look at the avarage depth on your logbook too. As I posted above (pics attached) it looks strange to me. That dive was dowloaded on Jdivlog before and avarage was 5.6 mtrs.

[dmainou]

Hi Peppe,

Mi log book reported the correct average depth for those dives that I checked.

BTW, my Mrs bough these tiny Japanese butterfly fridge magnets. I stacked up three of them to use them on the back of the mk2 the worked a treat.

D

[heinrichsweikamp]

Hi,

I'll upload a 1.85beta tomorrow before any more users hung-up their Mk.2 in the simulator...
Maybe I'll get this "average depth in logbook-bug" fixed, too.

To all: We do not give email support for beta (or even alpha firmwares), sorry.

Cheers,
Matthias

[Heiko]

Hello community,

just to let you know how I fixed the running summilator, as I do not have a magnet at home:

I removed the battery, before starting a new simulation I changed the CF32/CF33 and the simulation and everything else is working fine now.

I'm using GF model. Maybe it helps also to change the model back and forward.


Cheers

Heiko

[JeanDo]

@Nakatomi

2. When simulating a deco dive to e.g. 60m and 20 minutes bottom time, ... Shortly after that stop has been reached, the stop is again lowered to 33 meters and appears in red letters.


Yes I am aware of that bug. This is due to some bad rounding mixed with varying gradient factor (eg. won't hapend in ZHL16 mode, or with GF 50/50 for instance).
This is already corrected in 1.85beta.

Thanks for the report.

[JeanDo]

@Pepe, dmainou, Bardass, Oliver:

Sorry for the bug !

I tryed several settings, and WAS NOT ABLE TO REPRODUCE this bug on my test mk2 unit.
Maybe because of some reset (its my testing units, so it get a hard/soft reset very, very often...)

Just to make sure, I tryed various combinations of :
- Freshly compiled 1.85beta of 1.84beta from the HW depot,
- 1min @ 15m, 5min @ 15m, 15min @ 42m, 15min @ 69m
- ZHL16 or ZHL16-GF
- All settings reseted to default (Reset menu)
- Defaults GF (30/90), 50/50, 30/30
- Air, Air + G4: Nx40 @30m
- CF55 (switch delay) of 0min, 1min, 10min
- CF54 (stops order)
- CF29 (last stop depth) at 3m (default) or 6m.
- English and Deutsch version.
And never get a crash...

Other suggestions will be of great help here !
Many thanks,

[JeanDo]

Hi Pepe,

[quote]I also noted the new CFs 57 & 58, gas air consuption bottom and ascent gas...
what they are used for? Where do they apply?[/quote]

Once the OCR simulation is done (hum...), 1.84 now show you gas usage from all you mixes. It is also a quick way to cross-check that your Gas List is correctly set for deco computation.

Do do so, it need usage rates (litter/min, or bar/min if all tanks have the same volume) both for the bottom rate [CF56] (from start to the first stop), and for ascent [CF57] (from first stop to surface).

Note this function is disabled in CCR mode. Because bailout gas usage means we would have to compute the bailout runtime (starting with tissue saturation after a constant CCR ppO2, but using OCR ppO2 and gaslist for the ascent)... I still have to think about it !

Cheers,

[Olivier]

JeanDo Wrote:
-------------------------------------------------------
> @Pepe, dmainou, Bardass, Oliver:
>
> Sorry for the bug !
>
> I tryed several settings, and WAS NOT ABLE TO
> REPRODUCE this bug on my test mk2 unit.
> Maybe because of some reset (its my testing units,
> so it get a hard/soft reset very, very often...)
>
> Just to make sure, I tryed various combinations of
> :
> - Freshly compiled 1.85beta of 1.84beta from the
> HW depot,
> - 1min @ 15m, 5min @ 15m, 15min @ 42m, 15min @
> 69m
> - ZHL16 or ZHL16-GF
> - All settings reseted to default (Reset menu)
> - Defaults GF (30/90), 50/50, 30/30
> - Air, Air + G4: Nx40 @30m
> - CF55 (switch delay) of 0min, 1min, 10min
> - CF54 (stops order)
> - CF29 (last stop depth) at 3m (default) or 6m.
> - English and Deutsch version.
> And never get a crash...
>
> Other suggestions will be of great help here !
> Many thanks,

Jean-Do, no worries... My unit is still calculating (2 days after) deco time as I have no magnet at home (and pretty busy right now to buy one). But as far as I remember, I launched a simulation with following parameters:
- 1.84 english
- ZHl16
- G1 air and G2 Nx 82@9m, others inactive
- GF11 115%, others default
- tried to simulate 21mn @ 75m, or something like that. Minutes went down quickly as expected and is now stuck with 1min left

Good luck when hunting !