2 Master Handsets at 65m.....(!)

Padowan

I was diving yesterday, enjoying a nice, rarely dived wreck miles out in the channel in about 65m. I descended no problems, started laying line, got to the bow about 15 minutes into the bottom time, and my unit emitted a "beep-beep". Strange, I thought, that's usually a battery failure and that master battery has only been in a few dives so glanced down at my master handset, expecting to see it dead, but it wasn't it seemed still alive and well:
MASTER 1.3
1.3 1.3 1.3

So I grabbed the other handset, and looked at that one:
MASTER 1.3
1.3 1.3 1.3

Did a bit of a double take, initially thinking "Did I really read that correctly?" So I looked again, yep, definitely 2x Master handsets. At this point I wished I'd bought my video camera to film it as evidence, but I didn't.

OK, I think, I don't want 2 masters they might fight, or confuse each other, I'll power down the new master (my usual slave) reboot it, and let it pick up where it left off as slave again. Switched the handset off, turned it back on, "beep-beep" I expectantly look for the "DIVE-NOW" message, only to be greeted by the sight of "RESET OTHER HANDSET". OK, now I'm starting to worry a bit - I'm now running on one handset, that possibly has a power issue, and can't get the other handset into dive mode, without powering the only source of PPO2 I have. I knw of people who have been unable to get the unit into dive mode underwater after powering down both handsets, so I bite the bullet, and thumb the dive. Heading back to the shotline, I drop the SP and do a dil flush, just to validate that my single controller is controlling correctly, or at least giving me good PPO2 readouts, and start my ascent.

After doing most of my deep stops, and arriving at the station, having kept an eagle eye on the handset, I decide to now run the unit in O2 mode, and power down both handsets - I do this, and both handsets boot up correctly, one into master, followed by the slave.

Back on land (well, boat actually), I strip the unit down, inspect the handsets for cracks of moisture, checked out the battery contacts, metered the batteries with and without solenoid load, and the only conclusion I could come to is that there was a small power dip (probably solenoid induced) that dropped the voltage significantly enough for the slave to promote itself, but not sufficiently for the master to power down, or be demoted - both batteries were reading just under 6.0v without load, and approximately 5.8v when the solenoid was firing. I put 2 fresh batteries in the unit, metered these under load, and no load, and got figures of about 5.85v and 6.1v respectively.

Dived the unit today, and had no problems, except a single errant "beep" on the boat before jumping in, that may have in fact come from another unit on the boat, or if it was from mine, it gave no messages.

During the dive today, I was regularly checking the slave handset instead of the master so that any unusual events I would pick up before finding myself 2 handsets down...

Anyone expereinced similar, found a cause, or can give me some idead of other things to check?

__________________

TheMacallan

What handsets do you have?

Non back lit by any chance?

Padowan

Nope, I have backlit ones. Unit bought brand new in July 2004.

__________________

divetheworld

When powering up the handsets, each handset polls the memory. When the master is writing to the memory blocks at the same time that the slave requires information, the 'reset other handset' message is shown. This is indicative of a communications breakdown from the handsets trying to access memory locations and failing to do so.
If the error message had not been shown, then I would be more worried.

The reason that both handsets showed MASTER is not something I am familiar with as a common mode fault. Ergo, I have no idea why but I hope the above sheds some light on the symptoms of the error.

The error can be simulated sometimes by switching on both units almost simultaneously and scrolling through the options as fast as you can. Both handsets try to access the memory blocks at about the same time and bingo - same error message.


Brent

p.s. speak to MP about this if you can, most of the knowledge I have on the logic of the classic handsets is from long discussions with AP.
And dont panic. Captain Paranoia is tapping on your shoulder, no need just yet.

Green anyone?

__________________

Divejunkie

I had a similar problem but at least it happened on land. When I switched the right handset on as usual it would come up master, when the left handset was powered up it also came up master.

I couldn't believe it at first so I changed the order and switched the left one on first as master and surprise surprise the right one said... master!

I spoke to Martin Parker and he said send it back and they would sort it out. Which I did, two weeks later I got it back switched it on and it still read master/master.

I spoke to Martin again and he assured me they had tested it and found nothing wrong, they even took it for a dive. I did a video of the two handsets powering up and sent it to him as proof.

He said send it down again and they would sort it out, which I did. 3 weeks later I got it back and I'm so pissed off with the electronics I haven't even taken it out of the box. I must admit I've been enjoying the old twinset again.

I have to wonder it AP are finding out problems they should have flushed out during the R&T phases. God knows the units are expensive enough, including the parts and servicing, which is the usual horseshit you get from AP when you complain about the cost of things- 'its because of the R&T which went into the unit'- should it not work then?

Mark Chase

I have to say my faith in AP to sort out issues like this is almost non existent.

I was joking with Simon earlier that it would be a cracked hand set

ATB

Mark Chase

__________________
See my "Doing It Chasey" video where I'm locked into a padded room, naked, with two ball bearings and within an Hour, I manage to lose one and break the other!!!

Kevin Juergensen 16/11/08
[/quote]

divetheworld

Traitor

__________________

Mdemon

I've had this once and only once. "Rebooting" fixed it, never to be seen again.

__________________
www.southwestmafia.com

"small minds talk about people, Average Minds Talk About Events, GREAT MINDS TALK ABOUT IDEAS!"

The WRONG Attitude will get you killed.

"Once the agenda-monkeys and perfect-worlders have moved on, perhaps we can do some diving?"

billy

Hello,

I don't know if the post of divetheworld decribes exactly how the software of inspiration works but if it does.... I would like to do some comments.

The year is 2006. Memory pooling shouldn't be used. Especially on life critical designs.

You don't access the memory directly, use the functions provided by the OS. (If you have one....)

Sorry, wrong by design. That's why people uses semaphores, mutexs and so on. (deadlocks are waiting you but this is another story)

Everything should be a resource:
The ADCs, the displays, the circuit that fires the solenoids and so on.

Does the inspiration has a mini RTOS build in or we still play with the usual if, then else, and we try to do sync to something that it could be a task/process in the real world ???

We are sharing memory over the wire.....

It could be better if we could use semaphores. That's why we have them.

If you don't have, play with flags, it is easier. And easy to write it functions that use them..

See above.

Now the big question:

If somebody has designed a PPO2 meter and an onboard computer capable of doing all of the above using a RTOS, he releases the software under GPL and he says in order to compile it you need to buy a book that costs $70 http://www.amazon.com/gp/product/1578201039/sr=8-1/qid=1156066587/ref=pd_bbs_1/002-5172298-2482443?ie=UTF8 and contains the source code for the RTOS
would you help him ?

µC/OS-II is Certifiable for use in Safety Critical Systems why don't use it on our rebreathers ????

divetheworld

Hi billy, your quite correct, but- there is some context to consider.

The classic electronics was designed nearly ten years ago, not a bad accomplishment when you look at it from this perspective. It has been redesigned in part in the guise of the Vision electronics but give respect where it is due, dispite its significant flaws it is still more sophisicated than some 'bang-bang' controllers still in production.

Brent.

__________________