Nvram battery life

[Tony B.]

I have 2 that are used regularly that are 15 years +

Well over the 10 years of rated battery life lol

[vlad01]

I've watched a youtube video where these was one of these exact NVRAMs in a lab digital bench multimeter (suitcase sized thing).

The NVRAM was still working at 27 years at the time of the video which was about 8-10 years ago now.

[vlad01]

Been having sporadic Code 55 error from ECM (ECM Failure - check grounds. If good, replace ECM)

Engine will either run poorly, or start, then stall immediately.
Usually check things. Find nothing obviously wrong. Things return to normal and engine runs again.
On weekend, it did the start/stall thing and had to get towed home.
Next morning it started and ran fine for the next few days.... until today.

ECM won't even flash a Code 12 (that it is working)

Has the NVRAM battery finally died?

Going back to this, I had the same thing happen a few days ago, but this time the cause was having a VR manual loom plugged into a VN car, the ECU would spit chips, code 55 and would not run, backfire and flood the car, however it would start but then die quickly most times or run extremely poorly for up to maybe 5-10 sec as most. So it was something to do with the wrong wires going to the wrong places as the plug above the ECU that goes into the main body under dash loom have most of the wiring pins in the wrong location between these two models. I removed all that was not needed and re-pinned the rest to the right spots, the ECU and car run fine now, no more 55 and weird TP dash data like inverted TPS. I also found bad battery grounds on the body and engine block later on, but those seemed to only make it act like a flat battery although it may have contributed, but for sure the wrong pin out made it act like the ECU died and went bat shit crazy. If I only supplied 12v on that plug and nothing else, it was fine and that's how I figured the mismatch was the issue.

So def double check stuff like this.

I am tuning a VN with a built VR v6 with crow stage 3 cam for abd old friend of mine, and it took me days to figure out the electrical and mechanical mishmash that was causing all sorts including a "faulty ECU" symptoms.

[immortality]

I just used my NVram board yesterday, it's from the early days and it still worked. Last time around it seemed dead so I bought a new one but I tried the old one first yesterday and it worked fine so now I'm thinking it was the GQ4 programmer playing up previously.

[Tony B.]

Going back to this, I had the same thing happen a few days ago, but this time the cause was having a VR manual loom plugged into a VN car, the ECU would spit chips, code 55 and would not run, backfire and flood the car, however it would start but then die quickly most times or run extremely poorly for up to maybe 5-10 sec as most. So it was something to do with the wrong wires going to the wrong places as the plug above the ECU that goes into the main body under dash loom have most of the wiring pins in the wrong location between these two models. I removed all that was not needed and re-pinned the rest to the right spots, the ECU and car run fine now, no more 55 and weird TP dash data like inverted TPS. I also found bad battery grounds on the body and engine block later on, but those seemed to only make it act like a flat battery although it may have contributed, but for sure the wrong pin out made it act like the ECU died and went bat shit crazy. If I only supplied 12v on that plug and nothing else, it was fine and that's how I figured the mismatch was the issue.

So def double check stuff like this.

Good advice for anyone doing a new install.

In my case, the car had been running fine on the NVRAM for about 10 years before the engine starting acting up throwing a Code 55.

My first thought was the NVRAM battery has finally crapped out, but swapping to another ECM "solved" the problem (for a short time).
The intermittent nature of bad one day/corrects itself the next had me blindly chasing Code 55 "check engine grounds" points throughout the car.

It wasn't until I noticed Service Engine Light WAS NOT ON with Key On whenever engine would start/stall, that I was able to find a diagnostic chart to systematically follow.
And yes, it was a ground issue, but in the multi-pin connection to the ECM itself. After weeks of frustration, problem was solved in 1 hour.

Five days on and engine still fires and runs properly every time <fingers crossed LOL>

[vlad01]

Yep! Those pesky grounds! I didn't checked if the CEL was doing anything or not on this car I am working on atm, but TP item list logged 55 and another code I can't recall the number but said unstable voltage. The unstable voltage I reckon in my case was the grounds, but the weird harness pin mismatch was what was causing the ECU to go mental and spit a 55, the TPS appeared faulty and sometimes inverted.

[Tony B.]

I've still been chasing this Code 55 problem from 2024.
Most of the time I do NOT get any codes.
I've had a Code 55 before, and another one today.

Code 55 ECM Failure
Check all ground connections of the ECM. If all are OK, replace ECM.

Very random. Can be weeks between the problem arising.
Most times there is NO issue with driving and everything works normally.

Until: Car drives fine. Park and do your business.
Come out to start car, turn key to RUN position - CEL (Check Engine Light) flashes once and goes out.
THAT is not correct. CEL should remain ON until engine starts and runs.
If you start engine now, it stalls after a couple of seconds, over and over. Won't stay running. NOT driveable.

Get towed home intending to trace and fix the problem the next day.
Next day, put the key in and everything is back to normal.
CEL comes on and stays ON until engine starts (as it should).
Car drives normally again - until the next time the problem arises.

Yes - I have checked wiring over and over. Can't find anything visually or with testing wrong.
I've swapped in spare ECMs with no change. Everything was working fine for 10 years prior.

But obviously an intermittent problem somewhere.
Someone suggested that my NVRAM battery is close to dying.
I'd expect a "good or dead" battery - not a "slowly dying" scenario where the next day it's fine.

[vlad01]

I doubt it's the chip or it's battery itself, but I suspect a problem with the NVRAM module(board), maybe a bad receptacle on the ECU header/connector or a cracked solder joint or nearby track. There could be an issue with one of the receptacle in the chip socket itself. I have seen old chips from old XT PCs from the 80s where the ram chips had cracked/broken pins while still in the socket, just from the environment and time, temp cycles etc... And they were not obvious until I tried to lift the chip to find a number of pins not even attached to the chip, many were detached from the top of the lower thinner part of the pin.
On the NVRAM the pins are round and uniform all the way into the package and I have heard them breaking off at the package before from being removed to roughly.


I would go over the NVRAM module carefully and even prod, tap and wobble it while in use to see if you can coax the fault to appear on demand. If it does, then for sure there is a continuity issue. When doing this, use something non conductive as using bare fingers can corrupt the chip as I had that happen before simply by touching the solder pads on the back of the ECU, they are very sensitive to stray currents.

[Tony B.]

Vlad...


I have the NVRAM in a second car as well.
I could swap between cars and reprogram them as needed.
Then see if the Code 55 problem "moved" from one car to the other.

[vlad01]

That sounds like a plan.