Open Source Rebreather Design Idea

onetime

Not sure if this is the right place to put the question, but, what's the status of the idea floated a while ago around an open source rebreather design here on Rebreather World? Any news?

--Paul

AD_ward9

Perhaps a Moderator should best comment. But from a contributor viewpoint:

1. A 40GB area has been created on this web site to house it.
2. We have uploaded 4 presentations, and are just checking over the FMECA Vol 1 to ensure it does not make comparisons with kit from any third party, then that will be up, and also the Handset simulator which will be up.
I believe the plan is to go live as soon as both of these are up. FMECA we have.

Simulator we have (its called the Rebreather Code Validation Tool X1, as it is one of quite a few stages used in automatic code verification), but we added some features recently so we would like to remove a few bugs first.

The simulator is a professional validation tool, running Monte Carlo checks of the code, dive profile/result comparison, both hardware and user fault simulation, etc. Just we do not want to disgrace ourselves putting it out when we know we just added a fairly major function block that it uses on every tick: the respiratory monitoring using the data from pressure sensors on the scrubber stick.

Cheers,

Alex

onetime

Thanks for the update!

canuckdiver

any ideas on when we will be able to view this?

AD_ward9

Sorry the delay in replying. I didn´t see the thread for a while.

It is all happening. We are just really busy on some commercial work at the moment, which is why it is going slowly.

The FMECA had a lot of references to something that was covered by an NDA, so on my list is to go through it and remove those references. There is also a patent we need to file on one thing before we publish it.

The simulator with the respiratory monitor included is also progressing. It is just the pressure of work right now. It will be there, sometime soonish.

Probably Jan 06 it will all be ready.

Cheers
Alex

2stoned

I really hope that there will be some more news prior to January...
That is too long to wait. Give us just a little bit here and there, please...
This is truly and interesting project.


Some thoughts on the Open Source Revolution Rebreather:

Scrubber canister should probably not be metal: A plastic would offer better insulation, keeping the scrubber nice and warm and effective.

Are the CO2 laser readings and the He2 readings reliable in 100% relative humidity?
Could you CO2 monitoring system be integraded in other Rebreathers?
- I believe that is what is needed most currently.
- Should have been standard by now.

In the pictures, it looks as if this machine has only one Counterlung?
How will the addition of gasses be done...
Dil via the ShuDBoF, O2 via Scrubber-head?
ADV directly into Counterlung , or as Flushing through ShuDBoF?

HUD:
Please, when all is right, then only dim LEDs.

Computer:
Large display. Color, if readable/bright enough.
Would be nice to read a book during deco, video in future.
Since you included a speaker, how about some music, that would be unique.
Computer Gas:
Integration for multiple gas selections, show active Dil selection
Via flow/pressure monitoring to show witch cylider, if on CC/OC, Deco O2
Computer Gas:
Option of adding helium from separate cylinder into the mix for deap dives

Extras:
Very loud surface alarm via piezo units and run on batteries not air,
with build in time-delay to give you time to cover your ears.

Waterproofed electronic compass and GPS, so you can see how far you drifted when you reach surface.

DSV for Kirby Morgan M48 FFM
Poseidon Besea BCD and xstream Regs (just a thought)

Neck pocket/surface-bag, positioned on top of scrubber, between hoses; for
SMB, EPIRP, strobe, flag, snorkel, water, chocolate, pyrotechnics, what ever. (Quick release surface bag)

AD_ward9

Thanks. You have thought about this very well.
It is plastic. Only proto II was metal because it was used just for flow measurements, scrubber life testing and worst case condensation test. If you look at the proto III pictures you will see it is machined from Delrin and the real thing should by moulded from Kynar.

It works even immersed in water! Hydrophobic membranes are really tidy, plus it is mounted on the scrubber stick which gets hot and drives out the condensation from inside.
It was standard. That is the problem. A competitor got EN14143 to say if the CO2 sensor is fitted it must be accurate to within 3% across all depth and gas mixes. It has been 10% for years. They basically kept CO2 sensors off the market, because only recently has 3% been achieved.
You got it in one.
User can vary the brightness, and they also dim automatically with ambient light (same as the backlight)
Colour displays add to the power and disturb one´s night vision. So red is the backlight colour.
Is this the modern version of porn on deco? You´ll want DVD´s next. Incidentally, it supports MP3 because we have voice annunciation of alarms and the quality is not bad. Not wonderful, but not bad.
There are no gas selections. It works out what gas you are using so you can plumb in whatever you want, hit all the wrong buttons and still live. It has full gas sensing. If you see the display it shows the He levels. One has to measure He to measure CO2 correctly (it is an iterative process, because the CO2 sensor false reads if not compensated for He, and He false reads if there is a lot of CO2, so an iterative loop goes round reducing the error each time until there is none.
Thinking about. Not simple. It introduces extra failure points.
Thought about the compass already, but the magnets in the slider switch through it out. Pity as Earth magnetic sensor chip is really cheap.
It has the DSV built into the loop.
First stage can be anything (the injector is a variable orifice valve that tolerates 300 bar above ambient, though at above 30 bar it acts a bit like a solenoid, oscillating).
Good idea. We have a space under the scrubber for lift bags etc. Anything else behind is a problem. You use a snorkel?

Things are moving along. Just we have so much commercial work on right now we are up to our eye balls for about 2 months. I have had an FMECA sitting on my desk for the Open Revolution project since the start of September, and not been able to find a day to go through it to remove some stuff covered by an NDA. It is happening though. We have two customers who have both agreed to the Open Revolution concept.

Thanks for your thoughts.
Cheers
Alex

2stoned

Wow Alex
- So that is why I liked this one so much; it's close to perfect
Thanks for your thorough answers. They cleared some of my misconceptions.

Very good concept.
Very good explanations.
Bravo!

About mp3; wma is better (better compresses ion with better sound Quality)
How about the reader for eBooks?

CO2 sensor:
Waiting impatiently, like everyone else, i guess.
Glad to see that it is coming along though.
A 3% accuracy seems unneeded though.
-Edited:
Obviously:
We are talking about a spread of over 33 times normal value before effect is notisable.
- If CO2 represents 0,03% of atmosphere (0.03 kPA), meaning ppCO2 is 0.0003 at sealevel...

kPA Effect at ca 1bar (1 kPA is equal to approximately 1/100 atm)
>1-4 Mild to moderate dyspnea (ppCO2>0.01 to 0.04)
4-6 Dyspnea, anxiety
6-10 Impaired mental capability
11-15 Severely mental impairment, discomfort nausea
16-20 Loss of consciousness
> 20 Uncoordinated twitching and convulsions
(still below fatal levels)

So I guess that means keeping ppCO2 below 0.3?
Your display shows .00 didgets (to show from 33 times normal surface air ppCO2 and upwards. (0.01 and up)
-Here comes the question...
? Should there not have been an exstra didget or two in there?
That way you would also have a number to read in surface air.
Are my numbers wrong?
-Edit end

Piezo-alarm:
The surface piezo-alarm was meant as a diver2boat alarm instead of the ones that work on air from your tanks. Totally independent system. The delay would give you time to cover your ears and not risk going deaf. A single battery would go a long way on these piezo-alarms. The air-driven alarms use a lot of air. It could even be sealed inside a box, and opened on the surface, thus just needed to be safe against water spray or temp. submersion. Could be simple.

Electronic compass:
Magnetic interference from the switches on the Computer Unit, on the arm?
How far away from those contacts does the chip need to be? Shoulder?
If so then; you could put the magnetic sensor chip with battery inside a small orb.
This orb can be rotated to calibrate. Like they do with the handheld's or those inside watches.
Data transmission to its "holder" via IR.
Battery inside the "holder". Data via cable for display. Put the holder by the shoulder or somewhere you can reach it to check/calibrate.


The Surface bag:
Putting the bag behind your neck could make you more streamlined.
The bag can be released by pulling a cord/handle on the surface. Staying attached to diver by a string, you can blow some air into a pocket to keep it afloat. Instant dining/diving table. Now open your lunch box and a small plastic bag of fine wine while you wait to be picked up.

The Kirby Morgan M48, was for the ability to change to a new mouthpiece (alt bailout/snorkel) and not loose your mask at the same time. (using FFM)

More pics?

Thank you very much for doing all this. Keep up the good work.

Sten.

The only man in history that ever got all his work done by Friday, was Robinson Crusoe!

2stoned

Sorry to be such a nag...

The Computer:
I understand that it reads pressure from O2 and Dil Cylinders...
Would be nice if it was prepped for extra cylinders.
Incorporating all cylinders, the computer could calculate and alarm if:
You are getting low on one tank, meaning not enough to bail-out and also do your stops.

Scenario:
[You are using too much of your Dil, You need some for the wing up top.]
[Stop putting all that air into those lift-bags. Leave the gold or use some O2 instead.]

Deep scenario:
[Neon supply is critically low, ascent]
[Stop manual adding of O2 to your hydrogen mix, or you will soon be breathing rocket fuel]
[50/50 mix supply is low, surface now, else not enough for stops]
[Helium supply is plenty, put some in a lift bag with a note, and send a message flying around the world.]
[O2 supply is critical low, told you to ascent! Nah! never mind, Fish food!]

Like that, fully integrated, ready for anything. Full control.

Sten.

__________________
Sten.

The only man that always had all his work done by Friday, was Robinson Crusoe!

www.rebreatherworld.com - Is going down in silence - Staying longer in stealth and style - Diving first class, all the way!

AD_ward9

Nice idea. A few things mitigate against it:
If the tank sensors were RF one could do this, but they are wired (with a custom plug on each end of the cable) so the number of extra hose ports would add a lot of cost to the basic unit otherwise: each sensor itself costs $45 as well.
There is also the issue of ensuring the user relates the right tank to the right sensor. There is the other issue that at the moment the philosophy is not to enter gas mixes but the system works it out as it is used, and this would go against it a bit for a more traditional approach.
Taking the user completely out of the gas equations, and anything they or their blender does wrong out of the loop, should add to the overall safety. If there are extra tanks, just putting a contents gauge on each extra tank is probably the simplest solution for now.
The simulator that goes with it shows how much DIL one uses. In fact, I have a document to post as soon as the Moderators/Web Designers restore the Open Revolutiona area that is a very detailed model of rebreather operation and you can run any dive profile and it tells you how much gas you use and when.
If you plug neon into it, it will give a warning on the surface and won´t let you dive.
It calculates CO2 and He levels in an iterative process: without knowing the He % the CO2 sensor has an error. Without knowing the O2, N2 and CO2, then the He sensor has an error. This is because the thermal conductivity of CO2 is very different from other gases. If if cannot close the iterative loop on 20 cycles, then there is an unknown gas in the system.

It does not support Hydrogen, because of the injector options. A solenoid will ignite a H2+0 mix. Around the injector the PPO2 is very high. The overall PPO2 is high. Therefore once a large enough fire starts, it will burn nicely. Plug in H2 and the system will refuse to dive, but if already diving will ask you to abort because it cannot resolve the gases or the CO2.

The Open Revolution is about making it as safe to go diving with a rebreather as getting on a shuttle aircraft for an hour: in fact, safer. It is not about eliminating some risks so users can add others to keep the margin, in their view, constant.

In terms of moving forward with OR:
1. The system itself needs to be spec´d. We have a full spec. We are agreeing the publishable version with the primary customer (yes, someone has agreed to make and sell this beast, in fact two companies will be making it and selling it in different guises).

2. Prove the verification framework. I have uploaded a series of presentations to the OR area, and also put them onto our web site www.deeplife.co.uk, setting out the ground rules we are working to.

3. Continuing that theme, I attach the pdf file for review of the rebreather model used in the rebreather simulator.

4. The FMECA of the whole system has been written. There is a few days work involved in taking out a small amount of material that is subject to an NDA we have but appears in many places, then it can be published. It is a big document, with circuit diagrams of the whole rebreather as submitted (though reviews always ask for changes, lots of them, these are set down in the document: one part of the Open Revolution objectives is to educate manufacturers on how it should be done and seeing designer mistakes in the raw is a helpful tool in doing that).

The FMECA is 10 volumes. The circuits and so on, are in the first volume, which is easy to read. The others are somewhat dry.

5. A simulator of the whole rebreather, using the model attached for the Rebreather, driven by the actual code (Verilog HDL and C Code in the microcontroller). The published version will just use the C code, to save people having to load up Verilog tools for co-simulation. The simulator is part of the Rebreather Virtual Test Bench we have, that is being cut down in this manner to make it accessible. Some bugs still in the cut-down version, but a month or two will see these out and you can play with it.

6. The FMECA volumes then continue with the test data for the system and raw compliance data with its analysis.

Cheers,
Alex

Attached Images

Rebreather Test Bench_OR_051106.pdf (690.4 KB, 69 views)