Need Opinions on AAD, Time for my replacement

[ChangoLanzao 0]

Quote

Quote

is there a good reason why Airtec can't just recertify the old units after thorough testing and inspection?).

Yeah, they want to sell you a brand new unit for $1500 bucks!

Nope. That doesn't answer my question. That's not a good reason. I would really like to know if there is one.

[PhreeZone 15]

I'd say they have to consider the availability of replacement parts. At some point the parts bin runs short and they just can not replace the same sensors and circuit boards that are cycled through on each maintenance cycle. How fair is it if they were able to extend the life on the units for the first 3-5 years and then have to stop at some point. Its really hard to come by replacement parts for VCR's that are only 3-4 years old and they made many hundreds of thousands if not millions of VCR's in a year and maybe 100k CYPRES units in the last 20 years.

Yesterday is history
And tomorrow is a mystery

Parachutemanuals.com

[riggerpaul 1]

Quote

I'd say they have to consider the availability of replacement parts. At some point the parts bin runs short and they just can not replace the same sensors and circuit boards that are cycled through on each maintenance cycle. How fair is it if they were able to extend the life on the units for the first 3-5 years and then have to stop at some point. Its really hard to come by replacement parts for VCR's that are only 3-4 years old and they made many hundreds of thousands if not millions of VCR's in a year and maybe 100k CYPRES units in the last 20 years.

That presupposes replacement parts are required to continue airworthiness.

But what about units that would still pass the maintenance checks at 12 years?

Some certainly made it through the 8 year check without needing parts.

How do we KNOW that a unit wouldn't pass another check at the 12 year mark?

We could even call for bi-annual checks past 12 years.

If a unit won't pass without new parts, and the parts are not available, then fail it.

But right now, there are still parts, and I don't see why I shouldn't be able to continue to use my AAD if it can be maintained.

Sure, at some point the parts run out.

But that isn't now.

So, for the record, AIRTEC and SSK, can you tell us if some units make it through the 8 year checks without needing replacement parts?

Thanks!

[USPA 0]

At the lab I work at, a machine called a FACS is more then 20 years old. It has a lot more components then an AAD and these machines are used to diagnose patients (and their lives). They are build in way smaller numbers then AADs (if you ask the company about a serial number, then can tell you where the machine is, and has been). And parts are still avaible. The manufactorer offers trade-in programs to get them out of service, but untill then, they are simply serviced.

The trouble with skydiving; If you stink at it and continue to jump, you'll die. If you're good at it and continue to jump, you'll see a lot of friends die...

[billvon 2,400]

>I'd say they have to consider the availability of replacement parts.

That's one factor. A few more:

Some types of computer memory simply don't last more than about 10 years. FLASH memory, for example, has both a limit on how long it can retain data (on the order of 10 years) and how many times it can be written to (on the order of 10,000 to 100,000 times.) Many small controllers use such memory both for program storage and for nonvolatile storage (like the last-jump information.)

Some components, like electrolytic capacitors, dry out after a while, especially when used in a high temperature environment (as rigs often are.) This would lead to odd behavior of the circuit as holdup times decrease and filter characteristics change. Worst case it could lead to a no-fire failure if the remaining capacitance was insufficient to source the required current to the cutter.

Some components, like wire bonded integrated circuits (i.e. most of them) fatigue over the course of their lives if they are subjected to large temperature swings. A broken bond wire will lead to unpredictable behavior.

You could replace all those, of course, but replacement at the component level costs a _lot_ of money (around a hundred per unit to do it well plus cost of components) and you're still left with a device with other life-limited parts. Plus which, the cost of a failure of one device might well be millions (in the ensuing lawsuit) and thus retiring devices _before_ they have component-life based failures might be a very wise thing to do.

[ChangoLanzao 0]

Quote

>I'd say they have to consider the availability of replacement parts.

That's one factor. A few more:

[... lots of very reasonable stuff]

OK ... BUT the units do a complete self-check every time they are turned on. If any of that stuff you mentioned goes wrong.

F - A - I - L

So .... where's the beef?

[billvon 2,400]

>OK ... BUT the units do a complete self-check every time they are
>turned on. If any of that stuff you mentioned goes wrong.

>F - A - I - L

Uh, no. No self test can detect all the component failures I mentioned above. A FLASH checksum, for example, can verify that everything is OK at the moment you turn it on - but cannot detect a cell that's about to flip states. A continuity test can test for continuity through the cutter - but cannot evaluate the remaining capacitance in the firing reservoir capacitor.

[flydog 0]

What if the pressure sensor fails halfway through a skydive. Just because it passed self test doesn.t mean it will last the day when its actually in use.
Bill made may good points and I agree wholeheartedly with him. We have cycled a 14 year old Cypress through an altitude chamber and it functioned as normal each time. The following week we used it to activate the main on a test drop. It failed spectacularly. Our electronics gave us speed and altitude readouts. Everything was in the firing parameters but it didn't.
Was this the Cypres fault? I say no. They told us they were no good after 12 years, and we thought since we tested it we were ok.
With that said do you really want to gamble with your last chance just to be cheap. My Cypres expired so now I don't have one. When I get back to jumping more and doing heavy AFF quantities I'll buck up and get a new AAD.
Which one I cant say but I have no hesitation on any or their reasoning on life limits. They all have their reasons, but I don't think its to gouge us.

[VideoFly 0]

I’ve had two Cypres units. One had to go to Germany on its second four-year check. I’m glad they fixed it. My other has never had a problem and it is eight years old. Also, I’ve seen cypres’ work when needed. My car electronics is starting to go bad at ten years. My TV’s, computers, and other electronic devices typically have electronics problems within ten years. Why would I expect an AAD not to have problems in ten to twenty years? I like the four-year checks for piece of mind and maintenance if necessary. As far as the price goes, the extra buck or so per jump is no big deal. Furthermore, my son recently started jumping and I bought him a Cypres II. Moreover, I’m sure they all work fine, but at this time, I’ll stay with Cypres.

[ChangoLanzao 0]

Quote

>OK ... BUT the units do a complete self-check every time they are
>turned on. If any of that stuff you mentioned goes wrong.

>F - A - I - L

Uh, no. No self test can detect all the component failures I mentioned above. A FLASH checksum, for example, can verify that everything is OK at the moment you turn it on - but cannot detect a cell that's about to flip states. A continuity test can test for continuity through the cutter - but cannot evaluate the remaining capacitance in the firing reservoir capacitor.

That IS scary. So this can happen any time ... it just gets more likely as time passes?

[humanflite 0]

Quote

>I'd say they have to consider the availability of replacement parts.

That's one factor. A few more:

Some types of computer memory simply don't last more than about 10 years. FLASH memory, for example, has both a limit on how long it can retain data (on the order of 10 years) and how many times it can be written to (on the order of 10,000 to 100,000 times.) Many small controllers use such memory both for program storage and for nonvolatile storage (like the last-jump information.)

Some components, like electrolytic capacitors, dry out after a while, especially when used in a high temperature environment (as rigs often are.) This would lead to odd behavior of the circuit as holdup times decrease and filter characteristics change. Worst case it could lead to a no-fire failure if the remaining capacitance was insufficient to source the required current to the cutter.

Some components, like wire bonded integrated circuits (i.e. most of them) fatigue over the course of their lives if they are subjected to large temperature swings. A broken bond wire will lead to unpredictable behavior.

You could replace all those, of course, but replacement at the component level costs a _lot_ of money (around a hundred per unit to do it well plus cost of components) and you're still left with a device with other life-limited parts. Plus which, the cost of a failure of one device might well be millions (in the ensuing lawsuit) and thus retiring devices _before_ they have component-life based failures might be a very wise thing to do.

Great info here Billvon

Thanks.

[billvon 2,400]

>That IS scary. So this can happen any time ... it just gets more likely as time passes?

Of course. Any device can fail at any time. In electronics, people often refer to the "bathtub curve" - this is a curve that starts high, goes to a very low level for years, then starts rising again. It indicates that failure rates are very high initially (either due to manufacturing error or "infant mortality" due to a bad device) levels out to a pretty low level then start rising towards the end of the device's life.

With good burn-in procedures you can minimize infant mortality, or more accurately ensure the failures happen in-factory where they are easier to detect. However, there's not much you can do about the wearout failures other than put a limit on the device life.

[billvon 2,400]

>Frankly, that sneaky business practice really ticked me off, and I do
>NOT want to give Cypress my business again because of it if I can help it.

As an alternative, get your rigger's ticket and jump your Cypres as long as you like. That way if there is any problem with it, it will be very clear whose fault it was.

[ChangoLanzao 0]

Quote

>That IS scary. So this can happen any time ... it just gets more likely as time passes?

Of course. Any device can fail at any time. In electronics, people often refer to the "bathtub curve" - this is a curve that starts high, goes to a very low level for years, then starts rising again. It indicates that failure rates are very high initially (either due to manufacturing error or "infant mortality" due to a bad device) levels out to a pretty low level then start rising towards the end of the device's life.

With good burn-in procedures you can minimize infant mortality, or more accurately ensure the failures happen in-factory where they are easier to detect. However, there's not much you can do about the wearout failures other than put a limit on the device life.

OK ... so, assuming that everyone is being honest and nobody is trying to rip anybody off ... this indicates that the materials used in the Argus are superior and have a much longer device life. The engineers of the Argus must have a better handle on the "bathtub curve" in order to be confident that their product will not be automatically obsolete after exactly 144 months (plus or minus 3) from the date they are put into service. I guess what I'm trying to figure out is just how deep is Airtec's "bathub curve".

Why do you suppose that Airtec can't simply UPGRADE the internals of their units (with components that have a, shall we say, shallower "bathtub curve") for the people who have already shelled out the $1500 for their units instead of making us throw out the units we have and purchase completely new units that are not as long-lived as their competition's.

[PhreeZone 15]

Something to remember is that the components in the CYPRES1 are also almost 15+ years old since initial design. Newer chips and parts can allow for better tolerances and life spans. The CYPRES2 currently has a estimated life of 12 years but it could be extended also if the parts hold up per tolerances.

Yesterday is history
And tomorrow is a mystery

Parachutemanuals.com

[hackish 8]

Quote

Why do you suppose that Airtec can't simply UPGRADE the internals of their units (with components that have a, shall we say, shallower "bathtub curve") for the people who have already shelled out the $1500 for their units instead of making us throw out the units we have and purchase completely new units that are not as long-lived as their competition's.

The cost of the physical board inside the unit is very small. The cost of the physical unit is also small. Probably to the tune of about $300 out the door. The expensive part is the R&D, the support, the advertising. With a staff of 5-6 people you probably need margins of $1m per year just to keep the doors open. That's like selling more than 1000 new units per year, every year.

I frequently deal with the same misconception. Hey, if you reflash the computer in my car it doesn't cost you anything - why is it $500??? Well, true but not true. What about the 3000h of unpaid time to figure out how to do the performance reflash? R&D is a bitch.

With every AAD you buy you're paying some of the labour for those engineers and programmers and testers who developed it.

If, after 20 years they just replace all the innards of the AAD then how far is that from replacing the entire AAD? It's probably cheaper for them to replace than open the unit up to replace the electronics inside. In both cases they need to run a full battery of tests. Since the unit itself is not that expensive to build the labour costs quickly eclipse that.

-Michael

[ChangoLanzao 0]

Quote

Quote

Why do you suppose that Airtec can't simply UPGRADE the internals of their units (with components that have a, shall we say, shallower "bathtub curve") for the people who have already shelled out the $1500 for their units instead of making us throw out the units we have and purchase completely new units that are not as long-lived as their competition's.

The cost of the physical board inside the unit is very small. The cost of the physical unit is also small. Probably to the tune of about $300 out the door. The expensive part is the R&D, the support, the advertising. With a staff of 5-6 people you probably need margins of $1m per year just to keep the doors open. That's like selling more than 1000 new units per year, every year.

I frequently deal with the same misconception. Hey, if you reflash the computer in my car it doesn't cost you anything - why is it $500??? Well, true but not true. What about the 3000h of unpaid time to figure out how to do the performance reflash? R&D is a bitch.

With every AAD you buy you're paying some of the labour for those engineers and programmers and testers who developed it.

If, after 20 years they just replace all the innards of the AAD then how far is that from replacing the entire AAD? It's probably cheaper for them to replace than open the unit up to replace the electronics inside. In both cases they need to run a full battery of tests. Since the unit itself is not that expensive to build the labour costs quickly eclipse that.

-Michael

I totally appreciate what you're saying. The reality is that you also have to keep your existing customers. I don't see how Airtec is going to be able to do that unless they bring down the cost of ownership for their customers. My impression is that they are totally unconcerned about that and that they are banking on the fact that they have a reputation for being in business longer than the competition and capitalizing on the fear that their existing customers have of switching to something new.

[billvon 2,400]

>this indicates that the materials used in the Argus are superior and
>have a much longer device life.

It may also indicate that Argus simply has lower standards for MTBF. I strongly suspect they use FLASH-based processors, which are susceptible to the same wearout problems any other FLASH-based system is.

Again, if this bugs you, get your rigger's ticket and use your cypres as long as you like. Then it really will be more akin to an Argus.

>Why do you suppose that Airtec can't simply UPGRADE the internals of
>their units (with components that have a, shall we say, shallower "bathtub
>curve") for the people who have already shelled out the $1500 for their
>units instead of making us throw out the units we have and purchase
>completely new units that are not as long-lived as their competition's.

Because the guts of the unit are where all the R+D money went. You're not paying for a handful of parts on a board; you are paying for a design that's been reviewed, tested and verified.

[377 20]

Quote

Personally, I'd go with the Cypres2 because I jump from pressurizable aircraft often enough that the Vigil 2 misfire issue is a problem for me. Now, once the Vigil 3 comes out, I might revisit that.

I agree with Bill. Like him, I am an engineer. I talked a lot of shop with the SSK and Airtec guys at WFFC and they are very sharp. This is not to say that the competition lacks anything, but these guys really knew about software validation and QC, component aging and environmental factors affecting reliabilty.

They also outfitted me and a few other volunteers with Airtec black box recorders and had us do a high speed C 130 exit to see what kind of pressure transients were recorded. That impressed me, constant research, always looking for bugs, even in a mature product.

Airtec also had a better field response to the Thailand misfires than their competition did. I hate 12 year Cypres life but as an EE I can see some valid reasons for it. Components, especially certain types of memory circuits and capacitors, change with age. Will a 15 year old Cypres that passes a self test work? It probably will, but I can tell you that some component values and characteristics will have changed significantly from the date of mfr. I hate the limited life, but I like the required periodic checks. It forces you to get your AAD looked at and tested in far greater detail than a self dignostic can do.

I looked long and hard at the alternatives but ended up replacing my expired Cypres 1 with a new Cypres 2. I think the other AADs are good and are quite a bit cheaper in the long run, but I like Airtec's lengthy AAD design experience.

I used to work for a company that made implantable heart defibrillators. In some ways they are similar to AADs in that they use circuits to detect a critical situation and deliver a life saving jolt of electrical power to the heart in one case and a cutter in the other. I know how hard it is to design and test for every possible combination of inputs, events and component aging. Experience counts. Newbies make mistakes that the old timers made and solved years ago. Look at the cutter design and mfg stumbles that some of the newer AAD makers made.

Anyone can hire some engineers and build a good AAD today since the tough pioneering work was done a long time ago. I'd just rather be wearing one built and maintained by the company with the most experience in the business. That company is Airtec.

377
first jump 1968
early AAD user, SSE Sentinel and Sentinel MK 2000
2 cutaways, 0 AAD fires

2018 marks half a century as a skydiver. Trained by the late Perry Stevens D-51 in 1968.