Rover 827 getting 4 blinks on diagnostic lamp?

[Cardi]

Getting 4 blinks on diagnostic lamp and measure 750 ohms testing towards crank angle sensor from connector at front of engine. Cleaned connector contacts and fault is now intermittent but that could easily be a coincedence.  She has to live outside for a while so need a break in the weather before i do anymore.  I may well need some advice from those with more experience, if so, I'll start an appropriate thread elsewhere next time I post.

Russ

[Laird_Scooby]

That's encouraging that the fault is now intermittent Russ, gives a good pointer for a few more bits to try.

First, remove fuse "S" from the underbonnet fusebox :

Ideally, disconnect the earth on the battery as well.

There are two large, round multiplugs on the R/H suspension turret - unplug them from their other halves and spray inside both halves with contact cleaner, refit and unplug a few times, give another squirt in each part, do the refit/unplug a few times more and after the last one, leave them plugged in.

You can reconnect the battery and then refit fuse "S" and give it a try now. If all is good, you can ignore the rest for now.

Next thing is remove the drivers seat for access to the PGM-Fi ECU. With the battery disconnected, remove the mounting bolts, unplug the various multiplugs and give them the switch cleaner, plug/unplug treatment as above.

Depending how brave you are, you could open up the ECU and using an old, clean toothbrush and some IPA (IsoPropyl Acohol), clean any white debris (powdery stuff) from the circuit board and if you see any signs of water ingress, clean where that has been as well. You may not be able to get to all of it with the toothbrush so repeated soakings with IPA and stand it at an angle to let the gunge run off would be the way to go.

Once you're happy that it's as clean as you can get it, rebuild the ECU, refit and the drivers seat, reconnect the battery and refit fuse"S" and give it a try.

If all is good then happy days, if not i'd suggest trying a replacement ECU before changing the CAS, you can usually find them on fleabay for about £30-35.

I suspect Stu may well move these posts to make a new thread so don't be surprised if some of your posts disappear! They will reappear in another thread though.

[Cardi]

Taking advantage of a dry day and a break in the festivities, looked at the 14 way connector under the bonnet (C22), looked fine but gave it a squirt of switch cleaner anyway, next removed ECU, connectors also looking good, opened up ECU, it was clean as a new pin, looked over PCB (both sides) with a magnifying glass no sign of corosion cracked/dry joints etc, measured resistance from 20 way connector (C197 light green and blue/slate wires)  back to crank sensor - 780 ohms which seems reasonable. Started car fine no warning light, left it idle while i moved another car out of the way, drove about a mile up the road and just starting to think it might be OK when warning light came on. Did not have time to check diagnostic led. Checked ebay for ecu for ECU assuming i may well need to try that next. Only one with same part number available and it is in Lithuania and looks to have been damp at some time. I'll check diagnostic led but am assuming, until then, that it will show the same fault. 

 

[Laird_Scooby]

Did you remove fuse "S" before starting it and taking it for a spin Russ? It won't start with the fuse removed but once you've removed and replaced it, that will clear the fault codes. It may be picking up on a historic fault and giving you a false warning light.

Also worth checking before you condemn the ECU are :

Earth strap from front rocker cover to slam panel

Earth connection between battery and ABS modulator block on inner wing.

All the metal link fuses in the underbonnet fusebox.

The CPS itself - is it dirty? Likewise the back of the cam pulley. Seem to recall though the cam pulley was clean last time i saw it but that was a few years ago.

Also check your battery terminals are tight, earth lead to the starter motor and alternator belt for tightness.

[Laird_Scooby]

https://www.ebay.co.uk/itm/Engine-ECU-Rover-827-Honda-Legend-2-7-V6-manual-1988-91-37820-PL2-B03/163246684235

One on fleabay in the UK Russ, £25, no indication of whether it's for a manual or auto but the "MT" on the label may be that it's for a manual. Several others as well :

https://www.ebay.co.uk/sch/i.html?_from=R40&_trksid=m570.l1313&_nkw=rover+827+ecu&_sacat=0

Maybe find one at a sensible price closer to home than Latvia, Lithuania or anywhere similar!

[Cardi]

Thanks for reminding me about the "S" fuse I hadn't removed it after cleaning connector C22. Did so this morning but sadly no change, fault reoccured. Seems to me that the fault comes on when the temperature gets to the point when fast tickover reduces to normal idle, when the fault is on idle is about 500rpm.

Thanks for the links to ebay but they do not match the part number on mine (37820 PL2 G12), i know the manual & auto ECU's are different but there seems to several types about and i don't know what implications there might be from using the wrong type e.g. non cat type on cat equipped car.

The list of other things to check is interesting, I did have a problem with an ABS wheel sensor (resistance of sensor was low but slowly crept up) and after i changed it i had trouble clearing the diagnostics, at various times showed pump fault, solenoid fault and even brake light switch fault, eventually i cleared them all after disconnecting and cleaning a lot of terminations  but could not honestly say which if any were the problem, this was only a couple of days before the PGMFI problem started. I do rember cleaning the eath at the ABS end but not on the wing, I think that should be the first place to look but probably not today, it's raining again.

[Laird_Scooby]

Just so you know, fuse "S" will clear ABS and EAT faults as well. 👍

[Cardi]

You mean turning the ignition on and off 20 times isn't necessary - just pull a fuse - AAAARGH!

[Laird_Scooby]

You've got it! Pain in the butt doing the ignition on/off 20 times, especially as it doesn't always work first time! It can also blow fuses because of the inrush current, many things won't be a problem but anything with a motor or coil (as in the ignition coil, injectors etc) will present a possible problem.

[Cardi]

On 12/28/2018 at 7:00 PM, Laird_Scooby said:

Also worth checking before you condemn the ECU are :

Earth strap from front rocker cover to slam panel

Earth connection between battery and ABS modulator block on inner wing.

All the metal link fuses in the underbonnet fusebox

Had a little time to spare today so checked the above and found earth strap from the slam panel hanging in the breeze, all other connections were ok.. Cleaned up the conections and connected the strap to the cambelt cover where a bolt was missing and seemed the obvious place to put it. Cleared fault code and started engine, decided not take it for a run but leave it to warm up at tickover. When it warmed up enough to switch from high to normal tickover the revs dropped to around 500 rpm but no PGM fault lamp until I blipped the the throttle. Didn't take the cam cover off to check pulley/CPS but what I could see through the inspection hole looked pretty clean.

[Laird_Scooby]

Can you remember what the hot idle speed was before this problem occurred Russ? There's a couple of things in what you've just put that make me wonder if the fault code for the CPS is only a symptom of a slightly different problem.

Also, can you remind me please - does yours have Lambda sensors or not? I don't think it does but it's a few years since i've seen it!

[Cardi]

Hot idle was about 750 and yes it has o2 sensors.  I changed an o2 sensor a couple of years ago but for the life of me i can't remember what the fault sysmptoms were or how i knew the sensor was faulty.

[Cardi]

Found the notes i made about o2 sensors at the time, erratic idle/stalling, 2 blinks = rear bank sensor.

[Laird_Scooby]

As i said before, i have a strange feeling it isn't the CPS. First can you check the cam timing? Make sure both cams are lined up spot on when the crank is at TDC, if the timing has slipped on one or other of the cams, it will throw a CPS fault.

Also, disconnect the fron Lambda sensor and run a link wire from the rear one where the sensor plugs into the cars plug to the front bank plug. Pull fuse "S" for at least 30 seconds, refit and give it a try.

If there's no difference (CPS fault still comes up, confirm by counting the flashes of the red LED) then reconnect the front Lambda (still with the link wire in place) and disconnect the rear then do the fuse "S" thing, refit and give it another try.

If either of these tests give an improvement, renew BOTH Lambda sensors.

The thing i've found with the PGM-Fi engine management system is that while it's a very good system when running correctly, something silly can throw a fault that isn't necessarily correct on the diagnosis. Generally, it will be ok but you've already tested the resistance of the CPS and it seemed good. Therefore it's reasonable to assume (for now at least) that it is working ok and the fault code for it is a red herring.

Thing is, when everything is working as it should, the PGM-Fi system expects to see a certain "picture" from all the sensors at any given engine speed/load.

When something distorts that picture, it analyses what isn't happening. It assumes that unless one or both Lambda sensors aren't altering their output voltage quickly enough for mixture adjustments that they're ok. However, if one has gone weak, it might still alter the output voltage but maybe not as quickly as is ideal which could result in a weaker than necessary mixture on one bank or other. It can also cause one bank to run rich on the lazy Lambda. All of these will make the engine run at different rpm to what the ECU expects it to be. Because of that it's expecting to see a steady stream of pulses from the CPS at a certain frequency relating to the speed it expects to see. If it's running slower than expected, it will assign the fault to the CPS if all else seems good.

Because Lambda sensors can take a considerable time to die, even after they're technically out of spec, the ECU won't always "notice" this. As such the fault gets passed onto the next likely candidate, in this case, the CPS.

Hence the test above with the link wire!

[Cardi]

That all makes sense, if it dries up later I'll give it a try.  Might it be worth dusting off the osocilloscope ( if it still works, I haven't even seen it for five years but it is in warm dry condensation free storage so should be ok) if the tests you suggest don't help ? Not sure what i should expect see on CPS o/p but sometimes it's better to take measurements before you know what the results should be. Digging out and playing with 'scope is probably just displacement activity until weather improves.😀  

[Laird_Scooby]

The scope might be an option if all else fails, i'd go for back-probing at the large round multiplugs on the suspension turret, that way you prove the wiring from that point to the sensor as well as the sensor - it's also more accessible! 👍

This pretty picture might help :

To the best of my knowledge, the CPS is a magnetic sensor  so you'll get the waveform on the right with the bit of hysteresis thrown in for fun. The best time to try testing it with the scope is once it's thrown the fault which would mean getting the scope out, connecting it and letting it warm up until the point where it comes off the thermal and ECU controlled fast idles which is when the fault seems to occur.

If there is still an output at that point, we can then rule the CPS out completely.

Just a thought - you haven't adjusted the throttle cable or idle adjustment screw or anything of that ilk recently?

[Cardi]

This morning tried linking the 2 ecu inputs to one o2 sensor, the front one, car warmed up, idle settled at about 750 rpm, blipped throttle several times, no warning lamp, idle back to 750 rpm, everthing looking good so stopped engine reconnected everthing, restarted engine low tickover warning lamp comes on as expected. Tried same test with rear sensor only, fault still on, as you would expect. Retried test with just the front sensor again fault still on, not what i wanted or expected.

At that point decided to have a coffee break and consider options. Decided to leave engine to cool and repeat tests starting with rear sensor only this time.

It's all a bit inconclusive at the moment but quite possibly an o2 sensor.

Thanks for diagram showing expected waveform, if possible i'll connect  the 'scope before i try the tests again.

The front sensor doesn't look very easy to get at which is a pity as there is nasty looking twisted wire and tape joint in the lead going to it and that will have to be sorted whatever else is done. 

[Laird_Scooby]

37 minutes ago, Cardi said:

Tried same test with rear sensor only, fault still on, as you would expect. Retried test with just the front sensor again fault still on, not what i wanted or expected.

Did you remember to pull fuse "S" in between each test?

38 minutes ago, Cardi said:

there is nasty looking twisted wire and tape joint in the lead going to it and that will have to be sorted whatever else is done. 

Inside that there should be a red butt-splice if memory serves correctly. I didn't do that but observed it being done and advised against it at the time. I was busy welding the stainless back box onto the (remains of) the old back box pipe.

The front Lambda is indeed tricky to get at, i use a 50mm Lambda socket with an "ear" on it to attach a 1/2" drive then use the 2' breaker bar. Once you've released the threads (they should be fairly easy, copper grease went on in addition to the microscopic spot of anti-sieze the makers put on!) you can get your hand down and undo it, either with the socket still on until it's about halfway out (only about 4-5 threads so doesn't take much) then remove the socket and do the rest by hand.

Are you familiar with "W" crimps Russ?

https://www.autoelectricsupplies.co.uk/product/825/category/2

You need the smaller of the two sizes there 1.0 - 2.5mm with the tabs.

Once you've removed the old sensor complete with the male half of the plug, dig out the rubber seal from the bottom end of it and slide it down the wire. Using a small, flat bladed screwdriver, go in from the top end and push the tab/tang flat against the main part of the existing crimp terminal so you can pull it out. Now cut the crimp off the old Lambda and recover the rubber bung.

Buy a new pair of Lambdas, amputate any terminal that happens to be fitted (chances are it's a single round pin) and slide the rubber bung on. Now fit a new crimp terminal, slide the terminal into the plug housing ensuring it latches, feed the bung along and fit in the bottom end and you're ready to fit back on the car.

https://www.ebay.co.uk/itm/CLEARANCE-LAMBDA-SENSOR-FOR-VASUXHALL-ASTRA-F-MK3/264054405438

That should be the right one (but you'll need two), check with the seller that it is indeed a single wire Lambda. That's about the best price i can find on fleabay at the moment for the right one. Other models using the same sensor include Suzuki Swift, most 8v Vauxhalls from about 88-96 (some later but not always), Daewoo 1.8/2.0 models with the Vauxhall engines (Espero, Nexia) and several others. It's a fairly generic universal single wire Lambda sensor so there's plenty to choose from!

Although your tests today are slightly inconclusive, it is pointing to a lazy Lambda so give them another try and see how things go. Worth sorting out the mess of wiring on that front sensor, fitting a new front one and replacing the rear one as well so you have two good ones in there.

 

[Laird_Scooby]

https://www.ebay.co.uk/itm/FOR-VAUXHALL-ASTRA-F-MK3-1-8-FRONT-1-WIRE-UNIVERSAL-LAMBDA-OXYGEN-SENSOR-EXHAUST/163058055731

Another choice for Lambda sensors - i've got a pair of these on mine.

[Cardi]

Checked o/p of CPS today with results :--

At approx 1000rpm, no flt lamp, 10v peak to peak

At approx 750 rpm, no flt lamp, 10v peak to peak

At approx 450 rpm, flt lamp on, 8v peak to peak

slight throttle very approx 1000rpm, flt on, 10v peak to peak

Good sine wave under all conditions

Measrements taken at connector on suspension turret

I think it safe to say there is no fault with the CPS so i have ordered a pair of O2/lambda sensors.

Thank you for the help, much appreciated.

 

[Cardi]

Removed rear sensor without difficulty, but how do you get at the front one, from above or below?

I've ordered a socket as well.

[Laird_Scooby]

4 hours ago, Cardi said:

how do you get at the front one, from above or below?

I've always gone from above as that is easier for me, it's a fiddle though as previously described.

Fingers crosesd the new pair of sensors sorts out the various little niggles! By the way, i found the other day (according to the Mk1c Electrical Manual) the ECU is listed as being the same for manual and auto - find it difficult to believe unless the ECU senses the lack of EAT inputs and sorts itself out accordingly.

[Cardi]

Fitted rear sensor yesterday at that has sorted the problem, however i will be changing the front as well as soon as i can. As the air conditioning is u/s the simplest method looks remove the AC pipes and go in from the top.

Thanks again

Russ

[Laird_Scooby]

Glad we got to the bottom of this Russ, it's the 3rd or 4th now i've known of to do this. Just goes to show sometimes computers don't know it all!

You can simply unbolt the A/C pipe from the slam panel if you think it will help, i found even after doing that and tieing it out of the way it was more of a hindrance than a help so now i just leave it be and work round it. What can sometimes help is unbolting the slam panel and moving the radiator forward, better than introducing a leak into your A/C system that used to work well on that car, probably just needs a re-gas. Well worth fixing the A/C as well, it's amazing how quick the screens clear with it working.

[hughezee]

Many Thanks for assisting Russ on this Dave, I did have my suspicions on it being the Lambda but thought it best left to the expert on this specific model.😁👍