Question about Cells life

[MeRodent]

Bottom line is that "these" particular cells tracked both in Millivolts and PP02 perfectly, and held that output for an hour. I'm wondering if this sort of longer test might not have some value... just scratching my head here.

Having a chamber big enough to put the entire head into is a handy thing... but a cell checker would have done the same thing.


Dave

.

Given the similarity of a cell (or oxygen powered fuel cell as they are) to the tank on your car it's more likely that you have pushed the cell closer to failure (emptier) than anything else.

The most likely reason that a cell fails during a dive rather than in the checker is that it spends more time using more fuel here. The time between dives it's idling along, in the checker or on the dive it's travelling at full throttle.

 

[paulraymaekers]

^^ cells are about the cheapest part of diving...


This discussion got me interested in a few things though, so this afternoon I did a little experiment just for grins.

Have a trio of cells that are 20 months old here, so I put them into the Meg, did a very careful head-only calibration, and the tossed the entire head into my hyperbaric chamber. Ran it down to about 160 feet in air, watching the PP02 and millivolts on the handsets. Settled in at about 1.6 or so (that was the target) and then left it there an hour to see if the cells might weaken after a time. Bottom line is that "these" particular cells tracked both in Millivolts and PP02 perfectly, and held that output for an hour. I'm wondering if this sort of longer test might not have some value... just scratching my head here.

Having a chamber big enough to put the entire head into is a handy thing... but a cell checker would have done the same thing.


Dave

.

if you do the test with current limited sensors, you will see that at the start they will do nicely 1.6, and after a short time the current limited ones will go down
sensors seem to behave a bit like sorb: there is a little 'regeneration' taking place: meaning when you have your sensors for 24 hours at PPO of .21, and then you go for high PPO2, the sensors will work for a short time, but then go over to their 'current limited' stage...

don't trust cell checkers, unless you leave your sensors in for the duration of a normal dive... and at the temperature the sensors have during the dive

 

[Dave Sutton]

if you do the test with current limited sensors, you will see that at the start they will do nicely 1.6, and after a short time the current limited ones will go down
sensors seem to behave a bit like sorb: there is a little 'regeneration' taking place: meaning when you have your sensors for 24 hours at PPO of .21, and then you go for high PPO2, the sensors will work for a short time, but then go over to their 'current limited' stage...

don't trust cell checkers, unless you leave your sensors in for the duration of a normal dive... and at the temperature the sensors have during the dive

Paul, That's exactly what I suspected from earlier reading and discussion, and exactly why I prolonged the test. I recorded Millivolts every 5 minutes and the cells were flat across the entire test, which is confidence inspiring.

Given the similarity of a cell (or oxygen powered fuel cell as they are) to the tank on your car it's more likely that you have pushed the cell closer to failure (emptier) than anything else. The most likely reason that a cell fails during a dive rather than in the checker is that it spends more time using more fuel here. The time between dives it's idling along, in the checker or on the dive it's travelling at full throttle.

While I read and at first have some sympathy for the idea that doing this "pushes the cells further towards failure", I really don't see it that way. A cell that cannot stay flat at 1.6 for an hour does not deserve being used, and it's really just "one more dive" out of it's life. Like aircraft, we do not have confidence in things that we "baby": we have confidence in things that we robustly use and see by personal experience have extra performance that we do not need. If a cell is "close to empty" (which is not a very good analogy, frankly, but we can use it), I want to know that well in advance of any dive. I am pretty sure than recording Millivolts every 5 minutes in a 60 minute exposure to 1.6 or 2.0 PP02, with no Millivolt slide lower means that we are "still at 3/4 a tank" to use the anaology. I don't want to get to reserve...


I am going to do this test as a matter of routine, after taking cells from storage etc. Keeping careful records of each cell logged in a folder, and more carefully watching them.


Now maybe some cooling-coils in the hyperbaric chamber connected to an alcohol/ice slurry in a beer-cooler with a recirculating pump. If we want to simulate, we may as well do it correctly, with variable temperature capability.


Dave

.

 

[MatthewA]

Now maybe some cooling-coils in the hyperbaric chamber connected to an alcohol/ice slurry in a beer-cooler with a recirculating pump. If we want to simulate, we may as well do it correctly, with variable temperature capability.

Try heat. And Humidity.

 

[Dave Sutton]

Try heat. And Humidity.

Heat and humidity is easy. Heat is self-generating in the chamber when it's blown down to depth and easy to maintain with a heating pad on the outside.

Humidity... that's not a variable I can control in my small chamber so other than tossing in a wet sponge I'm not going to address it. I wish I could, but in defense of not being able to do so, that's more about condensation (which is a rig design criteria) than about cell current limits, which is the point of the test.

Cold: We often put on (literally) freezing rigs to start the dive, so the ability to cool the chamber would be fun. Plus I've already got the hardware, a cool-suit system that I use in the MiG and on the racetrack in the summer. It's just a six-pack cooler and a small circulating pump that you fill with ice/alchohol and then pump thru a suit that has tubes sewn into the interior. I can quickly install tubes and an entry/exit port on the chamber and set it up.

Actually.... you did just give me an idea though... Once I stick tubes into the chamber there is no reason not to circulate hot water thru there instead of trying to heat the chamber externally. Thanks for the brainstorm! Well done. Hmmm... internal heat exchanger, constant temperature valve... workshop sink and hot water five feet away... I think you've started something.


I'm gonna need a bigger chamber.....


Dave


.

 

[MatthewA]

Cold: We often put on (literally) freezing rigs to start the dive, so the ability to cool the chamber would be fun.

Please tell me that your "literally" is really "figuratively"!

Your cells will initially see a swing in temp when the reaction starts but once that occurs and the internals stabilize, there shouldn't be much of a change in internal temp unless something is happening with the scrubber stack.

If you are starting a dive with freezing rigs, the temp compensation network of the cells are going to have real problems. Fast temperature changes of 60+ degrees F can take up to an hour for the compensated signal output to equalize because the thermistor reacts immediately to compensate for the change, but the sensing membrane and electrolyte warms much more slowly, and as we know temperature plays a big role in current generation.

Might be fun to play with in a lab.

 

[WFO]

But where is the thermistor sited?

Where I'm going with that is if this is truly a problem you'd expect large sensor errors during "cold start" of a unit.

 

[Dave Sutton]

Please tell me that your "literally" is really "figuratively"!

No, literally freezing... well, "stored in the back of the van overnight" cold anyhow, and we are ice diving with ambient temperatures as low as, well.... brass balls and monkeys and all. Long prebreathing is the rule.

If you are starting a dive with freezing rigs, the temp compensation network of the cells are going to have real problems. Fast temperature changes of 60+ degrees F can take up to an hour for the compensated signal output to equalize because the thermistor reacts immediately to compensate for the change, but the sensing membrane and electrolyte warms much more slowly, and as we know temperature plays a big role in current generation.Might be fun to play with in a lab.

Hmmm... had not previously considered the time to come to temperature at that level... interesting to contemplate it. That's another topic to contemplate in detail. Thanks for the ideas. Definately worth a look. Freezer and multimeter here we come.


Really though, as you say: The sweet spot for testing for "classic current limiting" is really at the other end of the spectrum, when the loop comes up to temberature. So heating the chamber is the better point of interest, and truthfully? It gets pretty darned warm inside when you pressurize the chamber to start with. I'm not sure the difference between 20C and 30C means than much but I'm all ears if there is data to show differently.


Paul? Temperature considerations when doing current limiting tests?


Dave


.

 

[MatthewA]

I'm not sure the difference between 20C and 30C means than much

look at 40 - 60C. Unless you have a large heat sink nearby or have the RMV of a chipmunk, that should be the range the cells would see (in most designs) at working temps.

Freezing Van... overnight? You should treat your cells like you treat your girlfriend. (Strike that.fftopic I mean, your dog . You're inside, the cells should be inside. Sleeping in the freezing van? Your cells should be in the sleeping bag with you! In the end, they are really the ONLY source of information we have, and we know from testing that freeze/thaw cycles makes cells very grumpy and unpredictable (just like your girlfriend... or dog).

 

[Bobjr]

How about trying a drop of water on the front of one of the cells?

 

[MatthewA]

But where is the thermistor sited?

Where I'm going with that is if this is truly a problem you'd expect large sensor errors during "cold start" of a unit.

On most sensors in use in RB's they are mounted on the PCB on the back of the sensor.

And yes, you can have big problems if you calibrate the unit at 35F, or if the sensor is stored at a cold temp then put to use with a large temp swing. Eventually the electrolyte will warm to the ambient working temp, but until that happens the PO2 readings can be erratic, and/or just plain wrong.
Added for clarity: If you calibrated at cold temps, bringing the electrolyte up to working temps will not solve the issue. You should calibrate as close to the working environment as possible. Everybody... into the steam room to calibrate!

 

[Dave Sutton]

You should treat your cells like you treat your girlfriend. (Strike that.fftopic

Good point.

I'll teach her to bring in the cells at night... after all, three of the six are hers... :kiss:


Hmmm:..... I see in the crystal ball a 6 pin conector installed into a Meg so that the entire sensor carriage can be unplugged and brought inside... Hmmmm:.... <scratches head>... would work on rEvo too (sorry Paul). Who has expert advice on the best miniature gold plated connector to use? Would have to be small and absolutely reliable....



Dave


.

 

[MatthewA]

How about trying a drop of water on the front of one of the cells?

A thin layer of condensation (water) over the hydrophobic membrane will reduce the signal output of a sensor as much as 2 mV over time. Fully occlude the membrane and the signal will drop precipitously almost immediately. The former is the more insidious issue because in the latter, a drop from 1.2 to .17 should be fairly obvious. If the water is removed, either by gravity or drying, the signal will return to normal within 5-10 seconds, depending on the reaction time of the sensor (slows with age).

 

[MatthewA]

Good point.

I'll teach her to bring in the cells at night...

Smart Dog! What??????:rotate:

 

[Dave Sutton]

Smart Dog! What??????:rotate:

Oh boy, I'm gonna get it now as soon as she sees this... :whip:


It's funny, actually.... today I just sent in a deposit for a Boykin Spaniel puppy for us to have as a boat-dog, so maybe we can add that to the puppy-training (along with how to crap on the swim platform...) ...



Dave

.

 

[paulraymaekers]

Paul? Temperature considerations when doing current limiting tests?


Dave


.

ofcourse: working temp of a sensor is around 40-55°C

at that moment the galvanic part must produce far more current for the same PPO2 as at 20°C

so to check your sensor in 'working conditions': at elevated temp and giving time to equalise between PCB and galvanic part

... you see why I'm not so an 'advertiser' for cell checkers...?

 

[Dave Sutton]

Thanks Paul,

And "Yes", I see clearly that they are a tool to be used very carefully if you do not want false assurances.

Perfect full bench calibration and chamber-test for linearity and current limiting takes an hour. Not something for the back of the boat.


Dave

 

[emmbee]

Hi,

Sorry, I'm missing something...

From what I (understood of what I ) read over the past week or so, in normal operation the chemical reaction rate (read as current on the wire) at the anode (which increases with electrode surface area and temperature) is enough to consume all the oxygen flowing into the cell. The current is then limited by this flow (diffusion limited), which increases linearly with the pO2 outside the cell (Fick's law), and that's what's needed, so we're golden.

In a current limited cell, on the other hand, the reaction rate is too low to consume all the oxygen flow past a certain critical value. This can be due to too much of the electrode having been consumed (less volume -> less surface), or to the electrolyte being saturated with lead oxide, resulting in PbO coating the electrode (kind of like aluminium oxide on cylinders), which lowers the effective reactive surface (I read both, I don't know which one applies here). Either way, past that critical pO2, the limiting factor is not the oxygen flow anymore, but the reaction rate, so we get the current for that regardless of how much more oxygen there is.

It would follow, then, that if the temperature is increased, and so the reaction rate, the cell will stay diffusion limited until, and become current limited at, a higher pO2. So it could make sense then to test for current limitation at a lower temperature, since that would show potential issues sooner?

Except: not.

So where did I effed up?

Cheers,

Matthieu

 

[Paul Haynes]

Good point.

I'll teach her to bring in the cells at night... after all, three of the six are hers... :kiss:


Hmmm:..... I see in the crystal ball a 6 pin conector installed into a Meg so that the entire sensor carriage can be unplugged and brought inside... Hmmmm:.... <scratches head>... would work on rEvo too (sorry Paul). Who has expert advice on the best miniature gold plated connector to use? Would have to be small and absolutely reliable....



Dave


.

Mate don't go building in another failure point - put that Walmart DIY home electric tool kit back in the cupboard where it belongs before you do yourself some harm ;-)

 

[Dave Sutton]

Mate don't go building in another failure point - put that Walmart DIY home electric tool kit back in the cupboard where it belongs before you do yourself some harm ;-)

I'm way too lazy for that, mate. The idea is more for use by a friend of mine: my college room mate is scientific diving coordinator for the US Antarctic program. With the Meg being the NOAA approved kit, they will be there for research diving use. Having spent two seasons there myself, logistics in the field are based on all cargo being expected to be frozen to -40 routinely. Being able to quickly remove the entire cell tray would be a huge benefit. I'm thinking military avionics grade materials. Ill discuss it with Leon.


Dave



.