Hi All,
i have a stubborn P1647 code that won't go away on our 2000 VDP 4.0L. Car is getting 21mpg on highway. This is what I've done so far:
1) swapped upstream sensor bank 1 and bank 2 plugs. Code stays on P1647.
2) verified sensors are Denso and that Heater circuits on both upstream sensors have same amount of resistance (.9ohm)
3) verified heater circuit relay and fuse are working and that the heater circuit has power at ECM sensing wire.
4) checked continuity and resistance between the ECM (EM82 connector) and the O2 sensor connector (EM23).
5) did the pin point tests for the P1647 code out of service manual
6) inspected ECM board for corrosion, leaky capacitors etc. Very clean, no leaks or any visible damage
7) cleaned MAF sensor
8) ran Techron injector cleaner. Which helped with mileage and car ran better
9) cleaned body ground point on center of firewall.
10) with my OBD2 ELM scanner I can monitor the upstream O2 sensors voltage. Both are varying between .03V and .9V.
Though bank 2 tends to run .01V-.02V higher on average than bank 1. See picture. I can't tell the O2 sensors response time accurately enough to gauge if it's sluggish. And I'm not certain that a sluggish sensor would set a P1647 code.

Anybody had a bad ECM that was totally clean on the inside? Or a P1647 code from a working but sluggish O2 sensor?
Anybody have suggestions of what else I might look at? Are there other ground point that affect this? Help!



​​​​​

 

P1647 is an upstream but your graph shows downstreams.

The elec guide I have, if it's the right one, shows upstreams are linear (wideband) sensors which are sensed via current not voltage.

BTW I don't think you can just swap the plugs!

 

​​​​​​For some reason I totally missed that I was looking at the downstream sensor voltages. Thank you for pointing that out.
​​​​​​Looking closer at my OBD2 readouts I now noticed that it says it's in open loop due to insufficient engine temperature. But the temp is 198°F so that is okay. Also, I have 0% on both STFT and LTFT as well as 0 mA on both upstream O2 sensors. Is that because it's on open loop? Or does it think the engine is not up to temp because the O2 sensor is not putting out mA?

 

It seems there is a problem with the engine temperature sensor, as you mention that the CPU thinks the engine isn't up to temp. It may be that during the installation of the 02 sensors, something was not hooked up correctly.

 

ELM closed loop vs open loop detection isn’t reliable on early OBD2 cars.

 

Okay, I replaced the upstream bank 2 O2 sensor and nothing changed. Old one is in great conditionIcondition. I did a hard reset on rhe ECM. Still have the P1647 code and 0 mA. Bank 1 sensor reads .25mA but does not vary.
This all started when the surpentine belt for toasted by a bad tensioner. Previous owner replaced tensioner and belt but belt walked off right away due to not looking up belt on pulleys correctly. He left it in the garage for a year. Battery went dead. Then I bought it. Replaced belt and battery and the P1647 code showed up. Totally flummoxed on this!

 

You may like to use a meter to check whether current is flowing because if not you have a broken wire, blown fuse or some such (as the doc about that code intimate).

 

Change out the temp sensor it only costs $25 no need to drain coolant just put a rag under sensor take out the old one have the new one ready just screw it in do a hard reset this is over looked the engine thinks its not up to temp if its bad

 

My scanner shows the coolant temp at 190-199°F. Isn't that where it should be when at temperature? Is there more than one coolant temperature sensor? I read somewhere that the ECM also uses whether the upstream O2 sensors were reading within range to determine if engine was warm enough to go to closed loop. Also, I have reason to suspect that only bank 2 upstream O2 sensor is in open loop. The reason is when I monitor bank 1 & 2 downstream sensors I can see that bank 2 is running consistently richer than bank 1. If the ECM was reading using tables for both banks fuel to air ratio, wouldn't they read the same voltage?

 

So it's finally solved! I read somewhere, probably on this forum, that with issues like this the ECM is the problem less than 4% of the time. I did the JTIS pin point tests and tested all of the wiring involved and found nothing. I concluded that it had to be the ECM. I finally found a mechanic that had the ability to program an ECM and we replaced it with one of eBay. Problem solved!

 

@CaptRick:

yes I had exactly the same problem on my 2001 AJ27 X308 4.0 NA.

After chasing every problem in the book, then using IDS I found the computer was registering the signals but the CPU was not correctly interpreting them. So a trip to my local dealer who dug deeper into the ECM and found additional CPU error codes, although he did not feel comfortable to clear them without having a backup ECM. I sent my ECM to Foreign Auto Computer Repair in Texas [ForeignECUrepair.com], new ECM and codes vanished and never returned.

I feel the problem occurs when previous owners or previous mechanic's disconnect the ECM with the battery still connected and all voltage not properly discharged. That is standard procedure when unplugging any module.

Now that I am more comfortable with IDS and after reperforming the pinpoint tests, and having a proper power supply I will first re-flash the existing ECM to see if I can get all the code to vanish on my 2000.

 

Last year I had my local dealer run diagnostics and the service manager told me the MAP was defective according to the diagnostics and he showed me the printout.
So I popped the hood and knidly asked him to show me the MAP sensor on that engine a 2000 AJ27 4.0 NA
He scratched his head and looked and looked, but alas, my engine has none because it is calculated (totally guestimated) by some sort of algorithm using the MAF and throttle position.

The point is, beware of diagnostic programs. They are only troubleshooting aids and supplements to give you indications toward possible failures. Nothing replaces pinpoint tests.