e70 35d DDE connection issues, crank no start

[acarpenter86]

Any ideas welcome:
TLDR, intermittent DDE communication issues. Detailed thoroughly below.


I have a 2011 X5 35D that I repaired the wiring harness on. It was cooked by a pressure sensor hose blowing off. I separated each wire and re-insulated it with liquid tape and then wrapped it with self adhering silicone tape. Then the entire bundle was wrapped in the silicone tape AND exhaust wrap for heat protection in case that ever happens again.


Car started up after and I drove four short trips.


I then proceeded to a stage 2 tune, but was having issues flashing it with my cable so I had a local person bench flash it.


It started right away but then died 100 ft later. We flashed it back to stock. Same thing happened.


I found bank 2 injector faults (vehicle did sit outside and we got snow, so probably water ingress failure with terrible timing) so I replaced two injectors (5,6 of course) that measured outside of spec. injectors coded.
It started first try and I did about 3mi of driving as a test.


Then I got the tune flashed back on, installed the DDE and did another 3-4mi test drive. Parked it overnight, then the next day it would not start again! It waits about 5s after pressing "start" and then cranks but will not attempt to fire.
Ista shows NO COMM with DDE and sometimes ACSM (airbag) module. Codes also present are those from other modules expecting a signal from the DDE. (DSC, VTG, etc).


I then proceeded to check the fuses I could find, which oddly the only fuse labeled DDE was #4 and it was missing, I installed one just in case with no change. I also checked the five fuses inside the DDE box and continuity was good on them. I swapped the two relays in the DDE box for two known good ones from my other 35D. I also experimented with unplugging the intelligent battery sensor to reset, and once I even left it unplugged. Oddly, when I reset codes in ISTA, the DDE turned green as if working. But still crank no start after 5s delay.


My only remaining theories are:
-Grounding issue (what should I be checking?)
-DDE is somehow corrupt even though we flashed it a couple / few times via bench (he uses HexProg) and he didn't seem to notice anything.
-somehow my wiring harness repair is not solid


Final thoughts. Very weird that several times it drove just fine and then other times it would just refuse to turn over and only crank. Also, the delay from start button to actual crank has me puzzled. No glow plug emblem by the way. I tend to think the wiring repair is fine since it was not experiencing issues over bumps or something but like after parking overnight.

[328 Power 04]

What cable are you using to communicate with ISTA?

If K+DCAN and your latency settings are on the default 16ms, the cable will fail to communicate to modules at random because the latency is too high. Set the cable latency to 1ms im Device Manager.

That setting can really really really throw you off and chasing everything else but the problem.

[acarpenter86]

What cable are you using to communicate with ISTA?

If K+DCAN and your latency settings are on the default 16ms, the cable will fail to communicate to modules at random because the latency is too high. Set the cable latency to 1ms im Device Manager.

That setting can really really really throw you off and chasing everything else but the problem.

Thanks for the tip. I will confirm that I have those set correctly. I think I do though as I have been using this setup with my other BMWs for years. I have both a generic cable and as well a nicer one from Bimmer Tuning Tools.


I had an idea today actually. I have another e70 D that is in perfect working order. If I am starting to suspect my DDE is failing, would it be possible to put it into my "running and driving" X5 as a simple test? I dare not start it due to CAS and TCU incompatibility, but simply to test communication with ISTA and rule out relays/wiring/etc etc? I just don't want to mess up anything on the good car!

[328 Power 04]

Windows constantly changes that latency setting back to default 16ms as it sees fit, and some E chassis cars are more sensitive than others.

You can swap it into another car, but it would only be a test of bootup and communication. Nothing else.

Immobilizer will block the DDE and the transmission will throw a fit until you switch them back.

[bmwdirtracer]

Heck, I didn't even know you could run ISTA on an E body without ICOM......

[acarpenter86]

Update.
1. I found a working, non diesel contaminated DDE locally and had my bench ECU guy clone the DDE (including the stage 2 tune) to it.
I tested the new DDE and still got the same old comm errors.

2. Relay
Even though I had swapped the two DDE box relays with known good ones from my other car, I went to town today testing for correct voltage (with DDE out of the car) at the harness supplying power to the DDE. First, I found no power at the 4,6,8 pins on connect x200002 (from memory, the main power connector to DDE). OK, getting somewhere. I also had no power in the fuse holder. I trade it back to the blue relay (k2003a, main DDE relay) and found that there was supply voltage AND signal voltage to that relay and no voltage after. So, either I totally miffed before and maybe reversed the good relay and it still didn't work, or the relay failed recently. Either way, I didn't have the good relay on hand, so I soldered a 20A fuse and made a fused jumper. NOW it gets power to the DDE pins 4,6,and 8. I also heard the familiar "whine" in the engine bay that I am not sure what it is but is a good sign. BUT, STILL no comm with DDE (plus the "no message" errors from other modules). I tested BOTH of my DDEs. So, surely it is a CAN issue now?

Is it possible that the DDE and CAS need to be synchronized? I only read about that somewhere, but someone suggested it to another person with DDE comm issues.

I need to triple check the light green relay next to blue (I was so excited about finding the blue one misbehaving that I forgot to properly test it). I think it is for the fuel pump though? Should not affect DDE comm...
Any other relays, junction boxes, etc in the path that would allow voltage to the DDE relay but somehow mess up CAN comm?

I also ordered a powertrain wiring harness to have on hand. Determined to figure this out and I have to be close at this point.

I am sure I mentioned this before, but there is a 5s delay when pressing start until it will crank. I believe this is triggering a CAS error A0B3 error (starter terminal 50 expects a signal and it is delayed is my understanding)

Any help is much appreciated!

[328 Power 04]

No, there is no sync on those systems of that generation, they have a fixed ISN code, no rolling codes. Certainly a good DDE will communicate even if there was no security data changed and mismatched. Your test essentially revealed there is a problem elsewhere.

[328 Power 04]

Does it boot up and communicate on bench, not for cloning/service mode, but normal boot up.

[acarpenter86]

I need to get the person who has done my flashing and made the clone DDE to check for me but he has been very swamped.
I tried today to connect it to INPA on the bench but I did not have success. I think from some further research that you might need a gateway or a CAS?

I used a bench power supply, pin 8 for 12V and 3 for ground on x20002 (the main power connector to the DDE, plus used a switch to supply voltage to pin 13 (ignition signal) and of course CAN H and L on pins 19 and 20 (don't recall the order) on x20001. I saw the voltage draw go up slightly when the "ignition" was on, but could never get INPA to connect to it though it did detect the battery and ignition as "on".
Sorry, that second bit is kind of extra, but I thought you guys would know for sure if it could work or not and / or find it interesting.

[328 Power 04]

You don't necessarily need a gateway, nor a CAS.

You do need to make sure that if you connect CAN-H and CAN-L to the DDE CAN-H and CAN-L that you have between 60-100ohm resistance between the two CAN lines for it to talk to the KDCAN (CAN-H is pin 6, CAN-L is pin 14 on OBD end). I have a resistor added and talk to that generation of DMEs all time.

[acarpenter86]

Yes sir, I did indeed have a 100 ohm resistor between CAN H and L. I actually tried with and without the resistor.
And as luck would have it, I found an e90 today and tried both of my DDEs (original and clone) and neither communicated. Honestly, I was not surprised at this point.

I spoke to the person who did my flashes and he took them to see if his setup would still communicate with them.

I am new to using ChatGPT, but I found the ideas it gave me fascinating... at this point it is getting over my head (shoot, it already was but I have been learning), but I am posting this here just for my own curiosity and in case you all find it interesting.

1. Corrupted flash or eeprom structure in the suspect DDE — cloned into the good one


If you cloned everything bit-for-bit (both flash and EEPROM) from the suspect DDE onto the good one, any structural corruption, bad checksum block, or security sector issue would have transferred over, potentially making both non-communicative in any chassis.


BMW ECUs are picky about checksums, immobilizer sync (CAS-DDE), and sector integrity.
A corrupt area may allow one boot cycle, especially after a fresh flash, then fail once it hits a bad memory area or triggers an anti-tamper routine.


Test:
• Try flashing just the original known-good flash and EEPROM back onto the good DDE and see if it communicates standalone.
If it works, the suspect DDE’s data is bad.


⸻


2. CAN configuration/coding issue propagated via ZGM or vehicle order conflict


If there’s a problematic vehicle order (FA/VO) or a bad coding block somewhere upstream (in the CAS, ZGM, or Kombi) — it might allow a one-time handshake after flash/reset, then drop it.


Since you’ve tried in another chassis and still no comms, though — this would only be a factor if the flash image on both DDEs carries problematic CAN ID configurations or addresses that confuse the bus master (ZGM) and make it ignore the DDE.


Test:
• Bench test both DDEs standalone with a simple CAN test harness (power, ground, CAN H/L to a USB CAN interface or oscilloscope) — see if they broadcast any frames on power-up.


⸻


3. Residual security lockout or anti-tamper routine inside the DDE


Some Bosch EDC17 variants used in the E70 X5 35d diesel (especially in US 35d M57 engines) have anti-tamper routines that can go dormant after a flash but trigger after a power cycle.


This sometimes happens with programming tools that don’t correct internal sector checksum blocks fully — especially after OBD unlock / boot-mode operations.


Test:
• Re-flash the known-good DDE with stock software (from BMW ISTA/P or a verified WinKFP package) without transferring suspect data.
• Then check for communication.

[328 Power 04]

Well, at least we confirm that chat GPT is also inventing things and situations that don't necessarily exist.

The first one is plausible, except that most good tools will warn of checksum errors. And the person doing it should know better.

[acarpenter86]

Well, at least we confirm that chat GPT is also inventing things and situations that don't necessarily exist.

The first one is plausible, except that most good tools will warn of checksum errors. And the person doing it should know better.

LOL. That is part of the reason I posted this. It sounds believable the way it is presented, but it's all beyond my pay grade so I hoped you or someone would call bull crap if needed.

Not the first time Chat has straight up lied to me.

I will let you all know what happens from here. I think it is worth trying a ground-up stock flash on one of them if they are not dead. Which I really hope they are not.

[acarpenter86]

BIG UPDATE:
After testing in the e90 chassis, I sent BOTH my DDEs with the gentleman. He was kind enough to flash my "backup" with my tune I bought one more time but did not correct checksum. On the original, he flashed one of his tunes (similar performance tune with deletes). And what do you know, the car started right up first try. And also the next day- lol.

I did not test the backup with my original tune yet, though I still might. But I am also kind of hesitant to mess with it at this point.

Thank you so much for your help. I HOPE I won't be back for more...


Cheers