Some observations on random EAS and locking issues. I am hoping Gilbert can offer a theory.....
My friend is still here and using Bolt to run errands with. (He is the one who had odd issues with EAS a couple of weeks ago......)
A few days ago, he was parked next to the Borrego, and when he unlocked Bolt via Fob, whilst talking on his CDMA phone
the Borrego promptly woke up, and lowered just the rear end to the bump stops. (Timer relay is out, and just a 4 pin in place.)
It has had zero issues with EAS prior to this, and none since.
2 days ago, he was walking past the Borrego, and his phone rang. The Borrego locked itself. No fobs involved, but the dogs
were in the car, and wife had just checked on them, thus BECM was awake. Wife had left keys in the ignition, so had to use Valet key
and then fob to re mobilise.
Last night in Bolt, he turned in the seat to get something from the back seat, sat on his phone and butt dialed it. The doors locked themselves!
I am aware that CDMA phones can affect EAS as I was helping a friend in Australia figure out why his would randomly go to access
height for no apparent reason. It turned out that it was his phone and it only occurred when he was in a fringe signal area, thus phone was putting out more power to connect. Around town, no issues.
Here, his CDMA phone has effected both the 02's but only when he is in a low signal area.
I have GSM phones and have not had issues, that I am aware of. My Mate in Oz also had no issues once he switched to an GSM phone??
I know the factory phones in these were CDMA in US, so I am curious as to what has changed, if anything?
Maybe the factory phones were to blame for some folks EAS isssues?
Any clues?

CDMA? I suggest you tell him we're 21 years into the twenty first century, not still in the 1990's. Even GSM is old tech now, if your phone is capable of 4, or even 5G, it'll be LTE but with the capability of falling back to GSM if all here is available in the area is 2G. I suspect it is going to be down to the frequency band used by the particular networks having a mathematical relationship with the 315MHz remote locking receivers in the cars. So if they are using frequencies around 940-950MHz, that will be three times the receive frequency so the receiver will be responding to strong local signals from the phone. Any receiver will also respond to signals at multiples (or divisions) of the wanted frequency (odd multiples are more susceptible than even too) and will also respond to images, signals that are twice the receiver intermediate frequency above or below (depending on whether the receiver uses mix up or mix down) the wanted frequency. So when you have your marine radio on Ch 16 (156.8 MHz) and your radio uses an IF frequency of 10.7 MHz, you will almost certainly receive a strong local signal being transmitted on 167.5 MHz (156.8+10.7) but as that band would be used inland and not at sea, it's unlikely to ever be a problem.

Can't explain the problems caused to the EAS, although the ECU will have a microprocessor in it which will have a clock frequency so again there may well be a mathematical relationship. The EAS ECU will have had to be included in the EMC (Electro Magnetic Compatibility) tests when the car was originally type approved but they will be done at a specific field strength so if the phone is very close, and running maximum output when in a poor signal area, it could well exceed the level it has been approved at.

My central door locking works properly almost all the time, but I noticed that whenever I park at a parts store that I frequent, it doesn’t work. There must be some sort of interference there. (NAS GEMS). I don’t think this helps you at all Tom, but it’s interesting.

Next time you are there try putting the fob against the receiver aerial etched into the RH rear side window. You'll probably find it works perfectly then. You'll have the 315MHz system on your car so it's fairly unlikely there is anything else on the same, or a nearby, frequency, but there may a a powerful radio transmitter on a completely different frequency. This will desensitise the receiver. Think of it like you try to hear a quiet noise but someone is digging the road up nearby. Your ears will be so full of the other noise they won't hear what you are trying to hear. It's just the same when I park outside my daughter's house. There's a mobile phone mast about 150m away and the fob needs to be within 4 or 5 feet of the receive aerial yet at home I can lock it and unlock it from a good 20 feet away.

That makes sense. It took me a few times to clue in that it always happened there and seemingly never anywhere else.

Sounds like another issue our filters would solve :) when more components turn up to build a new batch at least...

When I go to the post office each day, there is one particular car park, next to a doctors, if I use it, I can only lock and unlock the RR by standing next to the RHR window. Any of the other dozen or so car parks are fine, just that one, so “something in the air” so to speak.

Do areas of interference keep the electronics awake and flatten the battery?

They can do. From my understanding, if the RF unit in the boot receives a signal it'll pass it to the BECM for the BECM to check if it's a valid unlock code.
The BECM draws a fair amount of power so if it keeps waking up, the battery can suffer.
There were 3 generations of RF unit, each with better rejection of invalid signals and therefore with each generation the BECM is woken less and the battery is under less stress.
Marty and Sloth are very very smart people and have built a unit to replace the RF unit which is as good as or better than the Gen3 unit, but without the Green Oval tax. I'm sure any explanation from them will correct my asumptions!!

Sloth wrote:

Sounds like another issue our filters would solve :) when more components turn up to build a new batch at least...

No, I don't think it would as the receiver isn't responding to the signals, they are just blocking it from receiving anything on the correct frequency.

The receiver needs to be receiving a signal on the correct frequency to keep the BeCM awake so it won't if it is simply being desensed as it isn't receiving anything. The Gen 3 receiver or the Marty/Sloth filter will prevent it being kept awake if the signal is on the correct frequency.

deleted, I'm speculating and Sloth knows so I'll wait for him :)

I'll elaborate - I'm not saying the filter will be able to receive a proper signal if it is being overwhelmed by other noise, but it will stop anything that isn't a P38 key from getting to the BECM. Including potentially this CDMA phone noise. Bearing in mind the original receivers would send out pretty much anything that came in be it on 433/315 or harmonics thereof.

At some point we'll put one on the scope and show the crap that comes out.

Amusingly enough, I was looking at electrical engine issues on a P38 yesterday that I have a suspicion might have had one of our filters fitted to it by its previous owner - when I go back with some bits to look at the engine loom issues, I'll have a peek and see if it does. They've gotten quite far and wide these days and I'm happy we've been able to help solve the old battery drain RF issue. As said - without the green oval tax.

Just realised, the filter doesn't stop desense as I've got one on my car (admittedly a prototype still)...... Agreed, it will stop anything that the receiver will respond to though.

The thing is the filter can't deal with the receiver being desensitised, as it's dealing with the output of the receiver. The LR offering is the same, it's just the nature of the whole design unfortunately.

The main thing beyond stopping battery drain is preventing memory corruption of the becm by interference, which is what screws things up properly and can cause things to go into lockout meltdowns etc.

Not only of the receiver in the P38, desense can, and will, affect any radio receiver. The filter stops received random transmissions from being passed through the receiver to the BeCM, it doesn't affect what the receiver receives, just what it does which anything it does receive.

I am aware - we designed the filter from the ground up, we have prototyped an entirely new receiver with filtering built in as part of its output, but decided it wasn't worth the hassle, because the way the system works it wouldn't help improve key responsiveness. Not because of the receiver, but the key itself.

Other marques of the time faired far better in terms of being able to still receive a transmission properly from a key even in areas of strong (or even intentional) interference. The P38 setup just has very poor error handling - it basically doesn't. So unless it gets a perfect transmission start to finish, it drops it.

We have future plans for a new fix for this, and have already prototyped a solution. Whether it will make it to production or not is another matter. The filter solves the main problems, and is well proven at this point.

Speaking of... another 60 to build in the coming weeks.

Maybe the filter does help with desense, or at least stops the datastream from being corrupted. Had to call round at my daughter's last night so decided to give the Ascot a bit of a run (and it had more LPG in it than mine). Mine, with the filter, can be locked and unlocked there but with reduced range, the Ascot, without the filter and probably with a Gen 1 receiver, wouldn't lock on the fob even with it held against the aerial on the back window.

Having spent a lot of time working on the problem when it first appeared, I'd say the worst affected marque was, and still is, BMW. Even the current Bini is affected. Land Rovers from 2010 until 2018 suffer but in a different way where they just won't acknowledge the key if you park in the wrong place and the fix isn't cheap.

Good to know it helps there :) I think its mainly preventing corruption, as it will only pass through what appears to be valid data for the P38. But its a bit of a dark art honestly in how it works. Early 90s communications... I'm just glad we're done on the development work with it these days as it was fairly headache inducing!

Makes canbus look trivial... fun fact: my own P38 has at this point three additional microprocessor based control units. One for my diesel heater control/timer, one for the inverter control, and one that reads canbus from my engine ECU to drive the P38 temp gauge, my cooling fan, and AC clutch go/nogo. All three individually have more processing power than the BeCM has... just goes to show though - they did a lot with very little power back in the day. But its all voodoo and magic if you try and figure it out 25 years later.

All three individually have more processing power than the BeCM has... just goes to show though - they did a lot with very little power back in the day.

Most of the BECM's code is down to toggling a few pins and reading a port, and then based on what it sees toggling a few more pins and writing a port! I suspect all the "hard work" like flashing the indicators and so on are offloaded to the massive custom chip on the board. The 68HC11 does handle the immobiliser "rolling code" decoding though, based on some numbers programmed into it. The decoding routine is a surprisingly small part of it - most of it is just write port write port read port compare write port write port write port, over and over and over.

These discussions always make me wonder how the useless Rcvr. in our rigs did not attract a huge Class Action in the US.

Their Cell towers can pump out some serious Wattage so the problem should/must have been much worse there.....?