EPIRB BATTERIES

[hellosailor]

OK, right at the top: This is not meant to be a discussion about philosophy and "Is it safe Kenneth?". Please, argue elsewhere about whether anyone in particular is competent to service their own life support equipment. This is a thread about TECHNICAL ISSUES ONLY, *PLEASE*.

In the past, I have replaced EPIRB batteries that literally consisted of just three SAFT lithium primary "D" cells, available pre-wired with a simple plug connector for all of $75. Old pack out, new pack in, ignoring any other aspects of the whole procedure it was just like dropping a new battery in any other appliance.

Now I'm looking at an EPIRB that was supposed to be made with user replaceable batteries--except, it turns out the maker was FOS and the batteries were potted into the housing. It was not pleasant chewing those batteries out, the maker is out of business, hallelujah. These are similarly three lithium primary D cells, but the destructive removal made it impossible to see a brand. Something came out in the potting which also may have been a fuse or diode, about 1/2" of metal cylinder that was down in the potting. And as I research this I find some packs apparently do have diodes in them.

Diode? In a sealed battery pack? Does that make sense to anyone?

I can understand a fuse, I like fuses, I've just never seen an all-metal-cased mini cylindrical fuse. I'm not sure I'd call one essential in this application, as most makers don't seem to bother with them. (If your EPIRB is trying to set itself on fire, I'm not sure you'd want to save it anyway.)

But then there are lithium primary batteries themselves. They're still the most stable batteries out there, and their ten year shelf life was why we had EPIRBs rated for 50% battery capacity after five years. Well, now the folks at Tadiran have gone and upped the game. They make new batteries with 10, 20, or 40 year life, depending on the drain they're used for. (Which is why some new beacons are saying "twenty year battery life".)

This means that using the new Tadiran batteries, going ten years or longer between replacements could be quite acceptable!

But fuses? Diodes? Versus the just plain packs? Anyone have any thoughts on that? Or on whether the drain in an EPIRB allows for the 20 and 40 year batteries? I'd put an ammeter on it, but I'm short on Faraday cages, and I really hate to annoy those SAR folks...

[Wotname]

Make, model of EPIRB?

Is that device in the battery is series with the power leads?

Might it be a temperature sensor (although I can't see why)?

I'll have a look at some tech manuals on ELT that might shed some light on the battery pack.

FWIW, Canada require much more in depth testing of an Canadian registered ELT than a USA one and thus the manuals give more detail of how it works if used in Canada!!!

[Wotname]

Originally Posted by hellosailor

............. Or on whether the drain in an EPIRB allows for the 20 and 40 year batteries? I'd put an ammeter on it, but I'm short on Faraday cages, and I really hate to annoy those SAR folks...

As you know, reducing current drain is one of the major design parameters in any new electronic device. Just like your smartphone manufacturer trys hard to reduce size of battery while increasing usage time, so does the EPRIB manufacturer.

In past, they used creative ways to achieve the 48 hour requirement as battery size is very important limiting factor in the EPRIB form factor. They gotta make it float!

Newer designs have significantly reduced current consumption and this is seen in smaller and smaller PLBs / EPRIBs. I note KTI now has a 10 year battery life.
EPIRBs | kti

However if your EPRIB is an older type, then I suspect you won't get any better battery life even if using the latest battery technology as your current consumption hasn't improved. Of course, I might be wrong .

As for testing, I have always used a good 50ohm dummy load connected to the antenna terminal. With an EPRIB you will almost certainly have to unsolder the existing antenna and solder in a short length of RG58 or similar. Way easier for me as I am usually dealing with an ELT which has a antenna port (usually BNC).

When determining current draw, remember these are 406 MHz pulse transmitters so you need a peak reading ammeter. The Tx pulse is ~450 - 500 milliseconds repeating every minute. If you confine the current measurement to the first pulse at initial turn on, you won't be alerting any RCC / SAR folk as this pulse does not constitute an active alert. It is coded as a test pulse. Just be sure you disconnect the power within 40 seconds of turn on. This way you will get the peak current draw of the first pulse as well as the continuous current draw of the 121.5MHz analogue transmitter during the first 30 or 40 seconds. I'm sure can handle the maths after that .

[Eigenvector]

Most likely a diode is to eliminate a failure of one battery to a dead short taking out the others.

[belizesailor]

If it looks like the image below then its a thermal fuse.

https://goo.gl/images/E9YxWp

Ive seen those in EPIRB battery packs.

[hellosailor]

Could have been a thermal fuse. "Removing" the batteries involved some serious attacks on the potting compound including (at wide intervals spaced over months of casual get-around-to-it) baking, heat gun, drill attacks, hot water immersion...pretty much everything except thermite or an oil well drill head. Needless to say, the ten year old should-have-been-dead batteries at some point also got insulted and outgassed, wonderful stuff. So by the time the potting compound was chewed out, the remains of the wiring (five leads, presumably one pair was for water activation, one pair power, fifth lead an extra ground?) and that silver thing were all borderline "charred debris".
But a thermal fuse does make sense, I suppose to prevent an out of sight EPIRB from burning up a boat. I suspect the battery wiring itself would have made a suitable fuse though.

Wotname, good points. As said the silver thing will never be identified closer than "charred debris". The EPIRB is a "Nat Satfind" and their documentation was apparently dumped along with the product line years ago. So, no help there.

I'm aware of the peak-vs-average current loading, but have nothing to measure short peak loads that way. Tadiran still rates their "new" D cells at 19Ah, no change there, just that the self-discharge rates are down to incredible low numbers giving them up to 40 year lives, depending on whether it is the high-current capable or standard battery. And apparently part of that depends on whether the device they are being used in has an outboard capacitor to help deal with peaks, or relies on the batteries themselves.

I suppose that IF I had a way to measure the peak draw, I could also bury the antenna in the center of a roll of aluminum foil, that'd make a nice Faraday cage.(G)

If I just figure conservatively, with roughly a 6W transmit burst and a ~11 volt pack (3 times 3.7-ish volts) I'd be wanting batteries that were able to put out 550mA for the 500ms burst period. Although I haven't had a chance to ask Tadiran directly yet, I think that still confines it to their "traditional" cells.

I think KTI and others claiming long battery lives must be using the "capacitor assisted" design to lessen the peak loads. Surprisingly, when you've got a sealed container, even with three D cells in it, it doesn't take much empty space to give it positive buoyancy.

[hellosailor]

Wotanme-
As I was getting all the pieces together, I realize the antenna on this one unscrews completely. Literally, as if it was one machine screw. How would you put a dummy load across....A single wire antenna? Or just try a test with no antenna, in a sorta faraday cage, in the (brisky?) assumption that with no antenna, it can't be transmitting very far?

[Wotname]

Originally Posted by hellosailor

Wotanme-
As I was getting all the pieces together, I realize the antenna on this one unscrews completely. Literally, as if it was one machine screw. How would you put a dummy load across....A single wire antenna? Or just try a test with no antenna, in a sorta faraday cage, in the (brisky?) assumption that with no antenna, it can't be transmitting very far?

I'll get back to you with a reply later, time doesn't permit a reply ATM.

[Wotname]

Originally Posted by hellosailor

Wotanme-
As I was getting all the pieces together, I realize the antenna on this one unscrews completely. Literally, as if it was one machine screw. How would you put a dummy load across....A single wire antenna? Or just try a test with no antenna, in a sorta faraday cage, in the (brisky?) assumption that with no antenna, it can't be transmitting very far?

The proper way would be to unsolder the wiring or cut the track on the PCB before it gets to the actual antenna screw mount and connect a 50 ohm resistor between this point and the ground track of the PCB. But maybe the transmitter is potted into the case of the EPRIB and I guess you are then SOL.

Turning it on with the antenna unscrewed probably won't kill the transmitter but I would still stick it in a home made faraday cage because it is surprising how little radiated power is required to trigger those satellites. Maybe have handy the phone number for your RCC. In Oz, there is a free call number to use if you accidentally turn on an EPRIB. They appreciate getting a call before they have crank up their protocol when the system is triggered and IME are always friendly and understanding. A case of no harm, no foul.

[Wotname]

FWIW, I checked a couple of manuals re current draw for some ELTs. In essence they all use the same technology as EPRIBs so I reckon the current draw will be similar for any given generation of beacon.

They all used D sized MiLiO2 cells.

An older Artex (about year 2000) had 4 cells and drew 200mA steady state and 3.5 A during the pulse transmission.

A later model Kannard had 3 cells an drew 1300 - 1700 mA continuous and required a terminal voltage of >8.3V (under load). Initial terminal voltage was 9.5 - 10.1 V. This unit did have capacitor smoothing. Interestingly the manual required a new battery to passivated before fitting by giving it a 10 second discharge though 4.7 ohms and the voltage had to maintain 8.3 - 9.2 volts during the 10 second discharge. It also emphasised the 10 second discharge must only be conducted once. I hadn't seen this test required before or elsewhere...

[hellosailor]

Yes, putting a 50 ohm ground on the antenna contact might be possible. If nothing else, by running it from the antenna directly to the battery negative. I haven't looked into opening the head unit yet. And in the US, NOAA actually tells folks very clearly that IF you somehow trigger a false alarm, they will be quite happy to answer the phone and dismiss it. Much happier than when they don't get a phone call and have to burn assets. (NOAA, and the USCG, are both incredibly good at using telephones. Answering them, transferring calls, doing all sorts of things that BUSINESSES in the US seem to think is impossible these days. You know "there's something ringing" "Eh, yeah, ignore it, it will stop soon.")

This unit uses three "D" size cells for a nominal 9-10V pack, same as the older ACR's. From SAFT and Tadiran, they both use a LiSOCl2 chemistry these days, lithium thionyl chloride I think they call it. (SAFT quietly owns Tadiran, btw, but sells under both brands.)

Their "new" product actually is a D cell with about an extra ~20mm cylinder tacked onto the top end, which contains a special capacitor. Allowing the batteries to handle high intermittent pulse loads. If the height works out, I may look into that as well.

Thanks for the sleuthing.

[Wotname]

Originally Posted by hellosailor

Yes, putting a 50 ohm ground on the antenna contact might be possible. If nothing else, by running it from the antenna directly to the battery negative.........

I don't want to tell you how to suck eggs but beware it is possible the leads of the resistor could act as an antenna (i.e. radiating elements).
Keep those phone numbers handy

[hellosailor]

Certainly possible. I'd keep the whole thing in a tin, or the microwave, or wrapped in foil anyway. Or maybe shoot a YouTube on how to make a Faraday Casket for testing EPRIBs? (G)

[hellosailor]

Wotname-
I did a little more homework and had my memory dehazed. (Distilled water and a microfiber cloth are good for that.(G)
I would also note, I do have a fairly new GPS-enabled EPIRB, that's the primary. Fuddling around with old stuff is strictly a side issue, not a life-betting thing here.
Apparently the standard 406 transmits a 1/2 second burst every 50 seconds, and the first burst when it is turned on, is an "ignore me, just testing" burst. So even if that burst was sent immediately (as opposed to after 50 seconds of warmup) that would mean that if a 406 was actuated and then the power was killed at 30 seconds, nothing but the "ignore me" burst would get out, yes?
And that in any case, whenever that first burst got off, it would be somewhat like Red October: "Give me one ping and one ping only" and there's no harm to be done.
Presumably that one ping could be confirmed by one flash of the strobe, or one noise squawk on a 406 MHz receiver, as well. And the 121.5 locator tone should be transmitting as soon as the EPIRB goes hot as well, so there's yet another means to safely determine a 30 second count, well inside of 50 seconds.

With EPRIBs only transmitting for 1% of the time that they are on, that also means it is very unlikely that two of them both activated at the same time, would cause any interference with each other. (Well, I guess slightly under 2% or so, because of possible overlaps.)

If the old specs are still valid, that also means the plain 406's are still being monitored by weather satellites, with 3/4-2 hours expectable for a first location and only a 2km (nm?) circle of position generated by that, all the more reason to use the GPS-enabled new units, with much faster acquisition times as well as 200-meter positions.

I think it is that difference in accuracy and speed that puts the more compelling case on "get new equipment" for the folks who only have older units, huh?

[Wotname]

Originally Posted by hellosailor

Wotname-
I did a little more homework and had my memory dehazed. (Distilled water and a microfiber cloth are good for that.(G)
I would also note, I do have a fairly new GPS-enabled EPIRB, that's the primary. Fuddling around with old stuff is strictly a side issue, not a life-betting thing here.
Apparently the standard 406 transmits a 1/2 second burst every 50 seconds, and the first burst when it is turned on, is an "ignore me, just testing" burst. So even if that burst was sent immediately (as opposed to after 50 seconds of warmup) that would mean that if a 406 was actuated and then the power was killed at 30 seconds, nothing but the "ignore me" burst would get out, yes?

Yes, this correct.

And that in any case, whenever that first burst got off, it would be somewhat like Red October: "Give me one ping and one ping only" and there's no harm to be done.

The first ping typically occurs within <1 sec after turn on. The transmitter requires almost zero time to warm up and stabilise and even if it isn't 100% stable for the first ping, it doesn't matter; it's frequency (thus doppler effect) isn't used for location purposes unlike the following pings which are used for location.

Presumably that one ping could be confirmed by one flash of the strobe, or one noise squawk on a 406 MHz receiver, as well.

Yes, correct.

And the 121.5 locator tone should be transmitting as soon as the EPIRB goes hot as well, so there's yet another means to safely determine a 30 second count, well inside of 50 seconds.

The 121.5 transmitter is active whenever the unit is turned on and again, it only takes milliseconds to fire up after tun on.


With EPRIBs only transmitting for 1% of the time that they are on, that also means it is very unlikely that two of them both activated at the same time, would cause any interference with each other. (Well, I guess slightly under 2% or so, because of possible overlaps.)

There are several channels that are used and collectively they are considered to be 406MHz. Going on memory alone, I seem to recall they are at 25KHz spacing. So one model might transmit on 406.025 MHz and another on 405.975 MHz. I don't know the exact number of channels but I think it is 6 (or 8).
The 50 second quiet period is not fixed and can be anywhere between 50 and 60 seconds.
Both of these aspects reduces the odds of adjacent EPRIB interference considerably


If the old specs are still valid, that also means the plain 406's are still being monitored by weather satellites, with 3/4-2 hours expectable for a first location and only a 2km (nm?) circle of position generated by that, all the more reason to use the GPS-enabled new units, with much faster acquisition times as well as 200-meter positions.

I think it is that difference in accuracy and speed that puts the more compelling case on "get new equipment" for the folks who only have older units, huh?

AFAIK, all non GPS embedded 406 units perform to the same standards, new or old and give the same uncertainly of position.

The story is different with GPS embedded units. While the 406 Tx aspects remain exactly the same, the GPS engines and the way they are used to provide location data is wildly different between first generation units and current ones. It also varies between manufacturer and models by the same manufacturer.

The was no standard for how GPS data was used in the first generation units and AFAIK, there are still none. Historically manufacturers struggled to fit a GPS engine and provide power for it using the existing battery technology of the beacons in question without impinging on the 48hour requirement for the 406 beacon. Some dodgy techniques were used like only powering the GPS engine for 5 minutes and then shutting it down and/or transmitting GPS data once every 30 minutes and similar ways to shorten the GPS engine "on" time" thus reducing power consumption.

The real problem was (is?) the lack of standards for GPS embedded beacons as far as the GPS aspects go. The manufacturers aren't keen to tell you if they have taken a shortcut and there no easy way for the consumer to find out. I do know that KTI beacons have a GPS engine that provides an updated position every 5 minutes and this is much better than many other manufacturers.

If I was buying a beacon today, I would ask how often the GPS data was updated and if the supplier / manufacturer was not willing to given a written answer, I would look elsewhere.