problem with the depth reading

[dmainou]

Hi Jan,

This may be useful.

From the predator manual:

Pressure:
The pressure in millibars. Two values are shown, the surface (surf) pressure and the current (now) pressure.
The current pressure is only shown on the surface (i.e. when not diving).
The surface pressure is set when the Predator is turned on. If the Altitude setting is set to SeaLvl, then surface pressure is always 1013 millibars.


Emphasis mine.

The predator does not adjust the "zero" depth reading based on the current surface presure. Therefore that will explain some of the variance.

Also, there is no (or couldn't find any) salinity adjustments which will account for some variance as well.

2ft (66cm) "zero" error from manufacture on either sensor could contribute to variance

+-1% "measurement" error could account for the rest.

Please let us know what was the "raw" reading on JDiveLog once you've downloaded the dive.

D

[dmainou]

Hi Patrick,

I think that it is important to know if the average sensor, the hardware and the algorithm are OK.

Then we can go into checking if particular sensor are OK or not.

I think that it is important to have confidence in the MK2 design first of all. There is always the possibility of a failed sensor but that is a warranty issue.

One of my q's is if the 50cm difference observed by you is constant? For example 50cm at 10m and 50cm at 20m and so on. Does it increases as you go deeper? Is the increase linear? Is the increase linear once you exclude the first 50cm?

The other thing that we haven't considered is that the whole decompression model is based on pressure not meters. That is why its so important that the sensor is OK. So as long as the 2bar reading is relatively right I don't mind if it is at 10m or at 9.5m. The ppo2 of air at that "pressure" will be 0.42.

So, in a sense, you are right. I am contributing to the forum for the evil reason that I want to know if MY computer is properly designed and to make an educated decision if MY sensor was a good apple rather than a lemon.

If I was so concerned about an error on MY specific sensor I wouldn't be positng here and claiming that Matthias hasn't posted back. I would be on the phone to Manfred, Christian and Matthias until someone helped me with a warranty claim.

Happy?

Patrick75 Wrote:
-------------------------------------------------------
> Hello,
> dmainou you do not consider what the others have
> to say, you are only continuously interested by
> the normal operating characteristics of the
> sensor.
> In my own opinion it is not the problem : we are
> speaking on anormal readings, not on standard
> ones.
> I think that MY Mk2 does not measure the pressure
> with a sufficient accuracy : for example I dove
> twice in pools with paint marks "5M", "10M", "20M"
> and MY Mk2 displayed "9.5M" instead of "10M". That
> means -5% and -3M at an actual depth of 60M. I had
> also an Uwatec bottom timer which display well
> "10M" at the same depth (salinity setted at
> 1.00).
> This is only an example, at sea the same drift
> occured also by comparison with others divers
> instrument.
> Considering "border line" the -5% depth accuracy,
> I proposed that HW implement a correction factor
> function which could allow me, and also the others
> in the same case to adjust the display (linear
> corrective factor 1/0.95).
> Again, I have also compared the displayed surface
> pressure with official meteo services information
> : MY Mk2 display 1035 or 1040 mBar when the
> official information is around 1015 or 1020 mBar.
> Thanks every one for your information.
> My goal to inform HW is achieved even with no
> answer received from them : the 1.82 beta method
> was not properly stated in details in my own
> opinion.
> I very much appeciate my Mk2 but definitively not
> the depth accuracy of MY set.
> Best Regards Everyone,
> Patrick OSTC 464

[Regthing]

Repeated the test today with three comps in the pot at once.

OSTC set at 1.00 salinity
Apeks Quantum set for Fresh water
UWATEC BT which is calibrated for fresh water (TBC)

Took them to 40m for a few mins. (4 bar according to the calibrated gauge)
[IMG]http://i197.photobucket.com/albums/aa14 ... /Gauge.jpg[/IMG]




The results are:
[IMG]http://i197.photobucket.com/albums/aa14 ... OSTC-1.jpg[/IMG]
[IMG]http://i197.photobucket.com/albums/aa14 ... uantum.jpg[/IMG]
[IMG]http://i197.photobucket.com/albums/aa14 ... UWATEC.jpg[/IMG]




Profile
[IMG]http://i197.photobucket.com/albums/aa14 ... rofile.jpg[/IMG]




Raw data (after the "dive")
[IMG]http://i197.photobucket.com/albums/aa14 ... /RAW-1.jpg[/IMG][IMG]http://i197.photobucket.com/albums/aa14 ... screen.jpg[/IMG]

Again the OSTC looks like it is right on the money!

[dmainou]

Wow!

Thanks mate. VERY reassuring.

Couple of ratios (which most probably may mean nothing but never the less)

Seiko 41.3/40.1=1.03 (If it wasn't on the pic I would be asking if you were sure about the fresh water setting. lol)
BT 40.8/40.1=1.0175 (seems in line/close to what I have observed on my own BT which has no fresh/salt water adjustment)

I think there is a straight forward reason why the Mk2 is capable of such accuracy. I presume that the Mk2 takes the last "dry" reading to "zero" the gauge where other computers such as the predator (see quote from the manual above) assume that the "zero" is 1013 mBar, Therefore they are unable to get such an accurate reading.

I presume that this repeats across many other computers. Especially if there is the quite real possibility to go from being off to wet to being underwater within seconds. In contrast the Mk2 has a larger rechargeable battery that allows a sleeping computer to continue to sample the ambient pressure even when not in use.

Well done Matthias and Christian! Love your work.

D

[Chilihead]

comparing the air-presure:
[pre]
OSTC Mk.2 1.83 shows: 965mbar - 969mbar
"STRATOS" Funk-Wetterstation shows: 943,7hPa (=mbar)
Garmin GPSmap 67CSx shows: 963,3mbar
Wetterstation-Rosenheim shows: 961.0 hPa <- is at 473m - i'm at 453m
[/pre]
i think that's exakt enough

[dmainou]

Hi Patrick

I have been thinking on how best to explain the observations so they are easy to explain and therefore you can get an educated guess of whether your sensor is off or not and why a cf that allows you to adjust the sensor is not the best idea.

If you think it's off then please contact the factory so you can get a working sensor just like everyone else.

I hope that the following example helps a bit.

Any computer works like this: input>process>output
In the same line if the sensor is not right then we have garbage>process>garbage
and therefore why you need a good sensor rather than an algorithm to adjust the error based on the diver's appreciation.

In the case of the Mk2 the inputs are two buttons and a combined temperature and pressure sensor. The process is the ZHL algorithm and the output is the data on the screen.

The ZHL algorithm is based on pressure not on meters.

In general terms it is accepted that 1 Bar equals 10 meters of water. This is just an approximation, the true measure is that 1 bar equals 0.98692 ATA. Or in other words one bar is equal to 1.01325 Bar. One ATA is equal to ? 1 kgf/cm2.

Therefore, a column of ten meters of DISTILLED water will have one ata of pressure (1 kgf/cm2).

There would be no argument about the output of the computer if we were only to discuss pressure. Breathing 100% O2 at 1.6 bar will give you a ppo2 of 1.6 bar (surprisingly). However, this is slightly more complicated and people prefer to have a distance APPROXIMATION in place of the pressure.

The reason I mention approximation is because basically we never dive on distilled water as all other liquids are likely to have something dissolved into it. Whether it is chlorine, calcium, potassium or sodium it is quite likely that we are diving in liquids that have dissolved "salts" in them. The question is how many of them.

An easy test would be to buy or rent a precision scale and weight 100 litres of water to measure the extra weight and therefore adjust the algorithm to display a better approximation of distance. However the ZHL algorithm, itself, does not need to be adjusted as any extra weight added to the solution by the salts is immediately incorporated into the calculations.

However, this is not practical and most of the time we conform with an approximation. That is, if we don't adjust the reading for the added weight provided by the heavier/harder water, the computer, for example, will show 10m when in reality you are only at 9.5m. If we over adjust then it will show that we are deeper than we really are.

We obtain comfort from knowing that whether we really are at 5.5 or 6m if the computer says that we "are" at 6m we can breathe our 100% O2 and know that we are breathing a ppo2 of 1.6.

You will remember that you received your Mk2 with one dive in the log? That dive was the pot test they did before leaving the factory and therefore we know that on that day the pressure sensor was working OK.

Of course there is the possibility that it failed after you received it. But how to test it if it is practically impossible to do a dive in distilled water and without sending it back to the factory?

Well, I think that there are is at least one simple test you could do:

There seems to be some evidence (anecdotal if you wish) that those bottom timers that don't have a salinity adjustment are set to something that we could call the mid point between simple fresh water (as in distilled) and sea water. Scott's test sets his Uwatec BT at 1.0175 and the analysis of my dive logs suggests that my Hollis BT is set at 1.015.

To me, this sounds reasonable if you consider that distilled water is 1.00 and, according to my local water department, the water at port Jackson has a salinity of 1.036. This will also explain why they give a deeper reading vs the MK2 when its salinity is set at 1.00

Therefore, one thing you could try is doing one or two dives with a borrowed uwatec BT and with the Mk2 salinity setting at 1.01 and 1.02 and see how you go. If they match or stay within the same distance through the whole dive then you are well ahead. The deeper you go the better.

Remember, it's not important if you really are at 10m or at 9.5 the important thing is that you stop where the computer asks you to stop.

If you want that the computer to closely match the real distance you will need to know the exact salinity of the water every time (by salinity I mean anything that is dissolved not just sodium it could be calcium, potassium, chlorine or other matter including plankton and other things). However, it will still be an approximation as you'll only be able to approximate to 1.03 or 1.04 not the average 1.036. Also, in reality the salinity of the sea changes with the currents and the various tides so it is always a distance approximation. The pressure isn't and that is why it's sort of irrelevant if you really are at 10m or not as long as the pressure is spot on.

I hope this helps and if in doubt call the guys at the factory. I'm sure they will be happy to pressure test it for you. It is likely that it will only cost you the return postage.

Best regards,

D

[Bulek]

Looks like I'm a lucky one, my Suunto Vyper and Mk2 shows exactly the same depths

[dmainou]

If you read your suunto manual you'll see that it days fresh water calibrated.

[dmainou]

Hey Scott,

Can we please get you to do a last test?

Could you please test the same three computers to 33m. However, this time, could you please set the Apex to salt water?

Would be interesting to see if the answer is 35m

Don't want to argue with anyone. If you tell me that you've personally measured the bloody pool or that the person who you trust the most in this life told you XYZ I'll believe you. This is not about that.

I'm simply interested in finding the outcome of the test and the "hidden" padding that manufacturers add to their hardware and software to introduce conservatism.

Best regards,

D

[Regthing]

Hi dmainou
Yeah I can do that; but purely for the science, I couldn't give a sh*t how deep someones pool is

[JeanDo]

@Scott

Very impressive tests !
And right on spot too, as I corrected a minor bug in temperature compensation of the captor, in the OSCT code.

As a matter of facts, I would be deadly interested if you could make exactly the same test (40m) at zero degree or so: I guess you might fill your pot 1/3 water (salted), 1/3 ice cubes, and 1/3 computer mix, and wait for it to melt gently before serving.

I made various test like that for the 1.81beta, but at 1bar because I don't have any compression chamber... You will have to wait quiet a long time before the child reach the inside of the OSTC, and the pressure/temperature sensor !

And just for the sake of Science, the same test around +40 degree (as temp compensation neutral point is 20 degrees) will be bright too.

Thanks,

[Regthing]

Hi JeanDo.

I can certainly look into it. It's quite a small chamber I'm using, just big enough to get the three comps in. I have much bigger ones though that might be better for an in water test. The other option might be to pop it in the freezer.

+40 will is also possible but I would have to wait until we where doing some more environmental testing at work, or stick it in the oven at home.

I can also pop a thermocouple probe into the chamber to confirm the temp inside.

[Michael]

Hi all,

from what I´ve been reading here, the MK2 shows most of the times a shallower depth that other computers....Well, my MK2 shows always a grater depth!! In that order: 1. Suunto D9
2. (approx. 1m deeper) Uwatec Bottomtimer (300m)
3. (one more meter deeper) MK2

Will do some testing next Saturday with another MK2 and one or 2other computer....

[dmainou]

Michael Wrote:
-------------------------------------------------------
> Hi all,
>
> from what I´ve been reading here, the MK2 shows
> most of the times a shallower depth that other
> computers....Well, my MK2 shows always a grater
> depth!! In that order: 1. Suunto D9
> 2. (approx. 1m deeper) Uwatec Bottomtimer (300m)
> 3. (one more meter deeper) MK2
>
> Will do some testing next Saturday with another
> MK2 and one or 2other computer....


Hi Michael.

Would you mind providing us a little more info. For example, which is your salinity setting and where are you reading you dive results (i.e. The computer itself, JDiveLog, etc)?

Try dividing the results you have directly on the ostc by it's salinity setting and the results of the uwatec timmer by 1.015 and 1.0175. See if they match your suunto.


Btw. The outcomes of the tests is that the pressure reading of the mk2 seems to be quite accurate. (by corroborating the results of the first pot test done at the factory)

now, once you have an accurate pressure reading then it's easier to estimate depth by adjusting the salinity settings.

Read the manual of your suunto. The ones I have read say it is fresh water calibrated. The uwatec bottom timmer seems to have a 1.0175 salinity setting. (my hollis bt seems to be calibrated to 1015)

Let us know how you go.

One thing to note though. If they stay within a metre each. Then don't worry too much as this is well within the reach of your hand. (think how many times your brain or your feet are not at the same height of your wrist?) Real problems could be signaled by big volatile differences in the reading.


D

[dmainou]

Hi Michael,

From the Suunto D9 manual (p 97-98, emphasis mine)


[hr]

Depth Gauge:
-
-
-
Temperature compensated pressure sensor.
Salt water calibrated, in fresh water the readings are about 3% smaller (calib-
rated complying with EN 13319).
Maximum depth of operation: 100 m [328 ft] (complying with EN 13319).
97
-
-
-
Accuracy: ± 1% of full scale or better from 0 to 80 m [262 ft] at 20°C [68°F]
(complying with EN 13319).
Depth display range: 0 ... 200 m [656 ft].
Resolution: 0.1 m from 0 to 100 m [1 ft from 0 to 328 ft].


[hr]

Therefore, if you set the OSTC to a salinity of 1.03 the reading should be very similar. The reason to be similar and not the same is that both sensors have error tolerances AND that the OSTC self adjusts to the required altitude. (i.e. adjusts for the current pressure before the dive).

In contrast, the D9 seems to have a number of settings for altitude. So the sea level setting is quite possibly set at 1013mBar regardless of the actual pressure on the day. If it is rainy and the pressure is around 1005 like this morning in Sydney it will still use 1013 as the baseline. The OSTC will use 1005.

This will translate into depth differences.

From your D9 manual: (page 106, emphasis mine)


[hr]

ALTITUDE DIVING
The atmospheric pressure is lower at high altitudes than at sea level. After travelling to a higher altitude, the diver will have additional nitrogen in his body, compared to the equilibrium situation at the original altitude. This “additional” nitrogen is released gra-dually in time and equilibrium is reached. It is recommended that you acclimate to the new altitude by waiting at least three hours before making a dive. Before high altitude diving, the instrument must be set to the Altitude Adjustment mo-de to adjust the calculations for the new altitude. The maximum partial pressures of nitrogen allowed by the mathematical model of the dive computer are reduced accor-ding to the lower ambient pressure. As a result, the allowed no-decompression stop limits are considerably reduced.


[hr]

Best regards

D

BTW. This is one of the reasons why I don't like suunto's. Each of them operate in a different way. The Hel02 is fresh water calibrated. And doesn't seem to have salinity adjustments. All have the latest "Suunto RGBM" algorithm available at the time of construction but there is no disclosure on the differences achieved.

Anyway enough rant. I hope this helps you a bit.

[peppe.marsala]

Dmainou,

I would bet my OSTC that you got the point!
The one you explained is the most logical explanation about this dilemma, also with the results of the pressure chamber tests.

Just to make an idea, atmospheric pressure gradient is -1mb/10mtr, so if the dive computers would have a tollerance of 0,1% instead of 1%, they would show quite a great difference between them even in shallow water (always in percentage and at the same salinity value and temperature).

In aeronautical enviroment the aircraft onboard altimeter NEED TO BE adjusted every time the ground atmospheric pressure changes to show the real aricraft's height from the field, otherwise if the altimeter is set to standard atmospheric pressure (1013,25 mb) it will show the height from the 1013,25 isobar (in this case we call it "flight level", used on high airspace to have long range aircrafts pointing to the same reference). A fixed calibrated altimeter will be extremely dangerous (recreationa aviation).

The same principle applies to our diving computers. If the are not able to read the present and real atmospheric pressure at the sea level but they refer to a fixed pressure (which can be 1013,25 mb @ sea level, 980 mb @ 300 mtr, 950 mb @ 600 mtr) they are not able to show the real deplth.

My 2 cents.

[Michael]

Hi Dmainou,

sorry for answering right now; but I was busy with "my other life" (work, you know )

Well, I think you REALLY got the point as peppe wrote....I slightly changed the salinity as you mentioned and the results are really good, not that difference anymore!!!

Thanks for your help and support!

Michael

[dmainou]

Hi Guys,

No worries. The only thing I could say is that I am truly interested in knowing that our computer is well designed and works properly.

I am no programmer so reading the assembler code is beyond my skills. Therefore I must arrive to educated evaluations through other means.

It's good that we all got something out of this long post. It is not often that people question how their "life support" equipment works they just assume it does. This was a nice exercise to confirm it actually does what its supposed. I still think people should do it more often.

So all good.

Best regards

D

[BANDiver]

Many people found that Mk.2 show less depth then number of different computers.
For much Mk.2 depth reading we usually set Salinity to 1.00, and this works.

I made tests with precise pressure gauge and looks like Mk.2 don't
convert Bar (natural output of the Sensor) into Meters of water:

1 meter of water == 0.10 kgf/cm2 == 0.09806379 Bar (meter of water conversion chart)

In case of my experiment with 100m depth (10 kgf/cm2) I have about 98.1...98.3m on Mk.2 (salinity 1.00).

Think need to enter correction into source code: multiply averaged sensor output by 1.02.

Am I wrong? May be this isn't bug but Feature?

Please help me to understand!

Best regards,
Serhiy.

[heinrichsweikamp]

Hi,

This has been discussed intensively, the OSTC assumes 1mbar=1cm depth. If you want more depth adjust salinity to 1.02kg/l and you have the requested result. Decompression is based on pressure, not depth, so the salinity has no effect here.
In fresh water, which is usually colder, the density of water is nearly 1 and this is accurate enough.

Cheers,
Matthias

[BANDiver]

heinrichsweikamp Wrote:
-------------------------------------------------------
> Hi,
>
> This has been discussed intensively, the OSTC
> assumes 1mbar=1cm depth. If you want more depth
> adjust salinity to 1.02kg/l and you have the
> requested result.

Right, as for now need to compensate 2%: setting salinity to 1.02 should be correct with actual water salinity of 1.04.
But with fresh water need to have salinity setting 0.98...

> Decompression is based on
> pressure, not depth, so the salinity has no effect
> here.

Partially agree with this, 2% error is small enough and overall conservatism in 20% cover it.

> In fresh water, which is usually colder, the
> density of water is nearly 1 and this is accurate
> enough.

Link that I refer, is for water temperature of 4 Celsius, and for this case 1 meter is around 0.098 Bar.
For case when 1 meter == 1 Bar, the water should be about 60 Celsius

Serhiy.

[sailor]

I believe density "1" means 1 KG per Liter.
And i doubt 1 mtr is 1 bar at 60 Celsius.

[heinrichsweikamp]

Here's a list from the PTB (www.ptb.de):

Temp(°C) Density(kg/l)
0 0,99984
1 0,99990
2 0,99994
3 0,99996
4 0,99997
5 0,99996
6 0,99994
7 0,99990
8 0,99985
9 0,99978
10 0,99970
11 0,99960
12 0,99950
13 0,99938
14 0,99924
15 0,99910
16 0,99894
17 0,99877
18 0,99859
19 0,99840
20 0,99820

1 is a good approximation for fresh water. Pressure (Depth) has no effect on the density in ranges where we dive.

regards,
Matthias

[BANDiver]

sailor Wrote:
-------------------------------------------------------
> I believe density "1" means 1 KG per Liter.
> And i doubt 1 mtr is 1 bar at 60 Celsius.

1 mtr == 0.098 Bar.
Water density over temperature Wikipedia: Properties of water:
Temp (°C) Density (kg/m3)
+100 958.4
+80 971.8
+60 983.2
+40 992.2
+30 995.6502
+25 997.0479
+22 997.7735
+20 998.2071
+15 999.1026
+10 999.7026
+4 999.9720
0 999.8395
?10 998.117
?20 993.547
?30 983.854

So water density about 0.98 should be when temperature above 60 C:
still think that +2% correction for depth should be added into source,
surelly for real diver life (0...30 C) .

[StefanT]

Hello all,

this thread is somewhat old already and the firmware v2.00 has 1m (fresh water) corresponding to 0.98 bar, so the issue may be considered as closed. But I thought I might add a small note to the discussion above.

While most of the discussion was about the density of the water there is one fact that I haven't seen mentioned at all -- the gravitational acceleration constant. In fact, the kg has historically been defined as the mass of 1l of water [ NIST, kilogram ] and today's more precise measurements are quite close (see Matthias' post above). However, this does not imply that 10m of water depth correspond to 1bar. The reason therefor is that a mass of 1kg exerts a force of 9.81N [ NIST, gravitational acceleration ] only but not 10N. Thus, sticking to a density of 1kg/l simple calculation results in a pressure of 0.0981bar per meter of water depth.

Based thereon, a true depth of 35m in fresh water at 20°C (0.9982kg/l) would result in a reading of 34.3m if the computer assumes 1kg/l and 10N/kg.

It's up to ourselves if we really need that little bit of seemingly more precision as there are a lot of other factors as discussed above... (but we have it now).

Regards,
Stefan