mp3.1 Idle woes from a really strange issue

[309jazzpanda 41]

Evening collective. My 309 has been running MP3.1 now for a few years and while it is a marked improvement over the Jetronic for driveability, one thing that has always annoyed me is the idle. I hunts on tickover once it comes off cold start phase, either with or without SAD valve it still hunts. It basically sounds like it has a really lumpy cam in (it has a standard cam) which does sound cool but isn't much fun sat in traffic. 

 

This week I decided to get my finger out and finally nail down what's going on. Just for reference I am using the thermostat housing with a 306 coil back bolted on.

1st thing was the lambda sensor, I found out it was incorrect as it had a zirconia fitted rather than a titania. Swapped that out and it improved but the fault persisted. Next I noticed the TPS, IAT, ECT wires were all looking pretty ropey at the connectors from being stabbed by a multimeter at some point in the looms life (pet peeve of mine) so swapped them all out for some I had left over from a jetronic loom years a go, no difference. 

Got the multimeter out and checked some readings, loom has good continuity between sensors and ECU, load tested them and all okay. wires good, plugs good, happy days. Checked the voltage readings at the IAT and ECT, ignition on both read 3.5v seems correct for the temperature around here. 

 

Now it gets really interesting. Just out of sheer f*** around and find out, I took a spare ECT I had and plugged it in to the loom, just resting it against the head (not fitted into the stat housing) and start the car, runs absolutely spot on with a good smooth solid idle. Ah okay the ECT wants changing, happy days.

I now fit my spare ECT (that the car just ran perfect with) into the stat housing reconnect the loom as it is supposed to be and run the car, fault returns, poor idle.

Plug the loom into the now old ECT that I just removed/replaced, hold against the cylinder head and it runs perfect, hmm so that sensor is fine?. 

SO, using the ECT that is loose and plugged into the loom I hold it against the thermosat housing and start the car, fault returns. pull the sensor away from the housing, idle is spot on again. For some unknown reason, when the ECT is in the thermostat housing as it should be and the loom plugged in, it will not idle correctly almost like the housing is causing the fault.

 

My theory is that the coil pack is causing a voltage error when the car is running, possibly from back EMF going through the stat housing somehow (coil is not cracked or damaged at all) but surely the housing of the coil should contain this?. I also can't remember off the top of my head but do these coils earth through the stat housing when bolted up like the Vauxhalls used to?. Should there be a surpressor fitted in this bit of loom like on the later TU engines?, as mine doesn't have one.

 

The issue I have is if this theory is correct, where to go next. I don't really want to mount the coil elsewhere but to me it 100% is an issue with the housing altering voltage at the ECT, which is just melting my brain a bit right now.

 

P.S Before anyone says it, the engine earth is fine. 

Many thanks in advance. Andy.

[welshpug 1,644]

there is no electrical connection from the ecu Coolant sender terminals and its body.

 

evidently making it run richer/leaner than the ecu thinks it should be is causing the problem, so there's something else making it run richer/leaner.

 

 

how does it run with the lambda unplugged?   is this a CO pot ecu originally?

[DamirGTI 340]

Which ECU are you using ?

 

I had (still have but "fixed") kinda similar issue with Motronic 3.1 .

 

On mine it starts out really rich from cold , barely firing on all 4 cylinders the colder outside ambient temp. the worse , but it lasted just a minute or so till the engine warmed up a bit afterwards it's ok. .. starting lukewarm engine also ok. .

 

Now that's with the *214 605 ECU , i ditched the *212 ZX ECU as on my 205 it wasn't really suitable to the engine specs (ignition map was quite a bit off .. had often detonation occurrence even on 100 RON petrol) , so i swapped it for a *214 ECU - much better , goes better/stronger/faster , an no ignition detonation but using 100 RON fuel and cold grade plugs (coldest per NGK BCPR9ES) .

 

But with the onset of colder weather i noticed that cold start enrichment issue , tried some stuff like you did but all seems good so it's enrichment map within the ECU which needs tweaking to suit 1.9 engine ... i made an bodge fix - soldered linear potentiometer on the CTS wiring and added around 4.700K Ohm's on the circuit and problem solved (tricking the CTS/ECU that the coolant temperature is warmer than it really is , around 15C ... once it passes 50C it doesn't really matter anymore) . 

Thus , with that , so far seems fine ... no ECU error codes , runs and drives fine .

 

But i'm still on a hunt for more closer spec wise stock ECU (ignition map mainly) , as it's impossible to find someone to remap stock stuff , there's at least 5 or 6 PSA Mp3.1 ECU's , one of them might be close enough , never perfect obviously but the closer the better .

Next up , i'm waiting for XM *202 ECU , will see how that one works .

 

D

[DamirGTI 340]

.. oh , and btw , Motronic 3.1 should run on green CTS not the Jetronic blue one .

 

I have both , seems better on green one .

 

D

[309jazzpanda 41]

57 minutes ago, welshpug said:

there is no electrical connection from the ecu Coolant sender terminals and its body.

 

evidently making it run richer/leaner than the ecu thinks it should be is causing the problem, so there's something else making it run richer/leaner.

 

 

how does it run with the lambda unplugged?   is this a CO pot ecu originally?

So it's originally a CO pot loom but I went lambda, as I had the wiring for it from an old loom in the shed. Using your thread with Graham's input and my own from when I fitted it I'm 100% confident the wiring is as it should be. I was originally heading to lambda fault as it only operates on idle/very light throttle so seemed the most obvious thing to me.

Saying that It's definitely a rich/lean issue but what's confusing me is the fault only happens when the ECT is in the stat housing. I can immerse the ECT in the header tank and it's fine, while yes the coolant is lower temp in the bottle compared to out of the cylinder head, the fact it will do this fault  between 10°c onwards, means that difference in temp (possibly 10°c ish difference) shouldn't have this much of an impact compared to just being unplugged/plugged in. 

I'll try to get a video of it this afternoon as it's a proper head scratcher. 

[309jazzpanda 41]

42 minutes ago, DamirGTI said:

Which ECU are you using ?

 

I had (still have but "fixed") kinda similar issue with Motronic 3.1 .

 

On mine it starts out really rich from cold , barely firing on all 4 cylinders the colder outside ambient temp. the worse , but it lasted just a minute or so till the engine warmed up a bit afterwards it's ok. .. starting lukewarm engine also ok. .

 

Now that's with the *214 605 ECU , i ditched the *212 ZX ECU as on my 205 it wasn't really suitable to the engine specs (ignition map was quite a bit off .. had often detonation occurrence even on 100 RON petrol) , so i swapped it for a *214 ECU - much better , goes better/stronger/faster , an no ignition detonation but using 100 RON fuel and cold grade plugs (coldest per NGK BCPR9ES) .

 

But with the onset of colder weather i noticed that cold start enrichment issue , tried some stuff like you did but all seems good so it's enrichment map within the ECU which needs tweaking to suit 1.9 engine ... i made an bodge fix - soldered linear potentiometer on the CTS wiring and added around 4.700K Ohm's on the circuit and problem solved (tricking the CTS/ECU that the coolant temperature is warmer than it really is , around 15C ... once it passes 50C it doesn't really matter anymore) . 

Thus , with that , so far seems fine ... no ECU error codes , runs and drives fine .

 

But i'm still on a hunt for more closer spec wise stock ECU (ignition map mainly) , as it's impossible to find someone to remap stock stuff , there's at least 5 or 6 PSA Mp3.1 ECU's , one of them might be close enough , never perfect obviously but the closer the better .

Next up , i'm waiting for XM *202 ECU , will see how that one works .

 

D

Yeah I'm on a 212 ZX ECU here, I do have a BX one in the garage somewhere but can't see it right now to get the number. 

 

Edit: just checked my build thread and the BX one 0 261 200 206. 

Edited June 14 by 309jazzpanda

[Tom Fenton 1,536]

1 hour ago, welshpug said:

there is no electrical connection from the ecu Coolant sender terminals and its body.

 

evidently making it run richer/leaner than the ecu thinks it should be is causing the problem, so there's something else making it run richer/leaner.

 

 

how does it run with the lambda unplugged?   is this a CO pot ecu originally?

This would be my initial thought. The coolant temp sensor reading ambient say 15 deg C when not fitted to the stat housing will be making the mixture significantly richer, at which point it runs better. Indicating to me that when it runs rough its really lean. Be easy enough to prove the theory if you can beg the use of a MOT emission tester for 10 mins to see what the exhaust mixture is doing in each case. 

[309jazzpanda 41]

1 hour ago, Tom Fenton said:

This would be my initial thought. The coolant temp sensor reading ambient say 15 deg C when not fitted to the stat housing will be making the mixture significantly richer, at which point it runs better. Indicating to me that when it runs rough its really lean. Be easy enough to prove the theory if you can beg the use of a MOT emission tester for 10 mins to see what the exhaust mixture is doing in each case. 

This is the odd bit, had it in work a few months back when I changed the lambda sensor, it was a steady 10% Co on idle after putting a titania in, but it was still rough on idle whereas, previously on the incorrect lambda sensor it was on average 13% but did fluctuate wildly. Which makes sense as the zirconia generates a voltage so will throw the ECU a right fit lol.  

I'm going to have to get it up to work on Saturday and put it on the gas tester while messing about to see what changes I'm getting, when Lambda unplugged makes it even worse so it's definitely giving a reading to the ECU. Might throw the Pico on the signal as well see what wave form is like. 

 

Odd thought but I got this ZX ECU from Spanish eBay, could it be possible there's a different map for the hotter climate?. 

Edited June 14 by 309jazzpanda

[Tom Fenton 1,536]

Sounds like it is majorly rich at 10/13% CO. Injectors and fuel pressure correct?

 

[DamirGTI 340]

2 hours ago, 309jazzpanda said:

Yeah I'm on a 212 ZX ECU here, I do have a BX one in the garage somewhere but can't see it right now to get the number. 

 

Edit: just checked my build thread and the BX one 0 261 200 206. 

 

How did it run with the *206 ECU  ?

 

Check if you've maybe mixed up O2 sensor wiring connectors , ie. plugged signal to heater and vice versa ...

 

Titania is very different to Zirconia in terms of mixture reading and signal generating , Motronic ECU feeds 5volt signal to the O2 and looks/measures voltage drop and uses that for calculating mixture so Titania O2 is more like variable resistor while Zirconia generates voltage kinda like battery .

 

Though , on mine was running good with or without O2 i couldn't really spot the difference bar some smelly exhaust and some gray smoke (with wrong O2 or unplugged) ... and there was error code for the O2 with Zirconia and with it unplugged obviously . 

 

Try green CTS intended for Motronic 3.1 : 1920 K9 ... it's reading is little bit different compared to the blue Jetronic CTS  .

 

And pull error codes from the ECU .

 

D

 

 

[309jazzpanda 41]

56 minutes ago, Tom Fenton said:

Sounds like it is majorly rich at 10/13% CO. Injectors and fuel pressure correct?

 

Yeah it's way too high at that as though it's trying to compensate for something. Going to get the fuel pressure tester put this afternoon, it did have one of those adjustable ones so I swapped it for a standard one.

Unfortunately I don't have any spare 1.9 injectors to swap out and test, any idea on the resistance reading they should have?. 

 

Actually found the proper jetronic fuel pressure kit in parts the other week as it was about to thrown in the bin, so I kept that bad boy which was a cheeky win. 

[309jazzpanda 41]

17 minutes ago, DamirGTI said:

 

How did it run with the *206 ECU  ?

 

Check if you've maybe mixed up O2 sensor wiring connectors , ie. plugged signal to heater and vice versa ...

 

Titania is very different to Zirconia in terms of mixture reading and signal generating , Motronic ECU feeds 5volt signal to the O2 and looks/measures voltage drop and uses that for calculating mixture so Titania O2 is more like variable resistor while Zirconia generates voltage kinda like battery .

 

Though , on mine was running good with or without O2 i couldn't really spot the difference bar some smelly exhaust and some gray smoke (with wrong O2 or unplugged) ... and there was error code for the O2 with Zirconia and with it unplugged obviously . 

 

Try green CTS intended for Motronic 3.1 : 1920 K9 ... it's reading is little bit different compared to the blue Jetronic CTS  .

 

And pull error codes from the ECU .

 

D

 

 

It barely runs at all on the BX ECU, I expected it to be bad but not that bad.

It's running a green ECT sensor as it should which is brand new. My spare is a blue but just on a whim swapped them and no difference to the ideal so it's back on green now.

O2 sensor wiring has been checked, double and triple checked as I was convinced it was something I've done wrong on that side of things but it's spot on. Reread Graham's posts and got the pin outs to be 100%, the Heater gets hot as well as it should. 

Only code I get is code 55 for CO pot which makes sense as it's not installed. Adaptives are reset as it's my targa car so when the master switch gets turned off it cuts power to everything, so the ten minute reset gets carried out whenever it's parked up basically. 

[DamirGTI 340]

13 minutes ago, 309jazzpanda said:

It barely runs at all on the BX ECU, I expected it to be bad but not that bad.

 

Strange .. might be faulty ECU , as they're all pretty much the same bar EPROM ie. different maps fuel/ignition for given ECU number/code .

 

Oh and yes , my *212 ZX ECU also came from Spain , and it had different maps than the ones posted here for the same ECU with the same code .

 

1.9 injectors are +/- 16/16.5 Ohm ..

 

D

[309jazzpanda 41]

41 minutes ago, DamirGTI said:

 

Strange .. might be faulty ECU , as they're all pretty much the same bar EPROM ie. different maps fuel/ignition for given ECU number/code .

 

Oh and yes , my *212 ZX ECU also came from Spain , and it had different maps than the ones posted here for the same ECU with the same code .

 

1.9 injectors are +/- 16/16.5 Ohm ..

 

D

Thank you for that, I'll give the injectors a test while it's dry outside.

Yeah I'd never tried the *206 BX ECU before and it's been sat in my garage for about 5 years so there's a chance it is faulty. 

I wonder then if it's possible the different map of being a Spanish ECU could be my issue?. 

It drives absolutely fine, even went around Oulton park last year, it is just that idle driving me mad. 

[SRDT 169]

You can have more than one software revision for the same ECU part number but it should be barely noticeable for the end user.

Significant software revisions were made around 1993 with dealers having to change EPROMs but it was mostly on new ECUs that had to be rushed into the market due to emissions laws. I don't think the XU9JAK or XU9J2 had those problems.

[309jazzpanda 41]

Unfortunately rain stopped play for the day. I'll get back to it if it dries up tomorrow and see what happens on the gas tester. 

[pug_ham 243 3 Cars]

Just wondering with regards to injectors.

 

I guess Andy is using the yellow 1.9 injectors as originally fitted to the uk spec 1.9 & ZX volcane but could the injectors be different on the Spanish cars for their emissions so it's causing this problem with over fueling?

 

g

[309jazzpanda 41]

6 hours ago, pug_ham said:

Just wondering with regards to injectors.

 

I guess Andy is using the yellow 1.9 injectors as originally fitted to the uk spec 1.9 & ZX volcane but could the injectors be different on the Spanish cars for their emissions so it's causing this problem with over fueling?

 

g

Now that's a good question. I am running the yellow top 1.9 injectors and in the very small window without rain today I managed to test the resistance of the Injectors, all 4 are reading 16 ohms. Of course I have not checked spray pattern or anything so who knows what they're actually like. 

 

Did another code reading and only getting code 55 even if I leaving it running for a good while, unfortunately I didn't manage to get it on the gas tester this morning so Monday dinner time it is. 

[309jazzpanda 41]

got the plugs out while the rain stopped for ten minutes and to me it looks like number 2 +3 are over fuelling. 

Cylinder 1 on the right. 

Edited June 15 by 309jazzpanda

[DamirGTI 340]

Nope , injectors are the same .. even on 2.0 XU RFZ engines running Mp3.1 too .

 

I tried few different injectors on this management , and no problems what so ever , finally settled with some *155 series yellow twin spray nozzle (from S16 i think ..) .. i also have adjustable FPR and tried playing with pressure , but none of these caused idle issues .. i even tried installing ISCV , but that one oddly didn't work right so took it of and fitted SAD back .

 

First when i installed the loom , i've been running it on *212 ECU without lambda , and afterwards with lambda but wrong one (Zirconia) and that too had no effect on idle . And my loom is from 605 with factory lambda wiring , so seems like these Motronic Mp3.1 ECU's aren't so lambda heavy as 3row Motronic 1.3's which do go bananas with duff or wrong lambda .

 

Haven't tried plugging it on a WB meter to see what's up with the mixture with these ECU combinations , if going by feels/sense , if nothing i'd say it was running more lean than rich on *212 ECU at certain rpm's/loads . But then , my engine is not standard so it needs and can tolerate more fuel anyhow .

 

Try just for s*its and giggles , change places of the IAT and CTS connectors ie. IAT on CTS and vice versa , and see what happens ... readings of both of them are about the same and same 2pin connector , just the CTS will be in more heated source aka coolant while IAT will stay on kinda cooler environment in the inlet tract .

 

I'd also try that *206 BX CU once again , but this time take the CTS out from the thermostat housing and leave it resting out plugged in the wiring , try starting and see what happens .

 

Also , could be vacuum leak causing this , while MAP system is not as sensitive to vacuum leaks i'd still check .. especially cracks/leaks on the exhaust upstream of the O2 sensor .

 

Incorrectly adjusted or bad TPS could be too , and that will not necessarily trip a fault code .

 

Power switching off from the electrical system , also might not be so good ... resetting the ECU over and over so doesn't have time to relearn/adapt and keep that adaptation .

 

D

 

 

 

 

[309jazzpanda 41]

Did a bit more today. Fuel pressure seems low to me at 1.9 bar idle, flicks up to 2.5 when revved. Now that's at the join to the rail, while I have the jetronic testing kit I'll test it before the fuel filter to see what difference, if any there is. From memory we put a 6 bar sytiec pump in it many years a go when the original gave up the ghost, as it was in the workshop not being used. It has had an adjustable regulator on it for years and had the same issue but it's now got a standard one on it since Friday. 

What's making me scratch my head, is on initial start cold start it's absolutely fine, it's once it's out of that cold start it then goes bad. I've also found Lambda signal is reading fine (going between 0.10-1.0v when cold/open loop), but once warmed and closed loop it just sits at 0.10V and doesn't make any adjustments. 

Whether these have closed/open loop I'm not sure as I'm used to the modern pugs, but that is the best way to describe what's happening. I'll swap those plugs and also give the BX ECU another go and see what happens. 

Oh one more think, I took the injectors out, one seal looks ropey so will order some new ones while I'm here, checked the spray and all 4 spray the same. Looking into the valve chamber the valves are spotlessly clean as are the chambers. Found the SAD valve has stuck in the open position so blanked the pipes and no different, probably won't even need it to be honest. 

I have noticed it sounding a bit "tappy" so will check the 4 branch for any cracks and may as well check valve clearance Monday night. 

Edited June 16 by 309jazzpanda

[309jazzpanda 41]

Swapped IAT and ECT wires, runs perfectly. This really doesn't make sense now .

 

Rechecked wiring:

pin 13 is ECT signal

pin 22 is IAT signal 

 

So they're now essentially plugged into the wrong sensors but it runs perfectly, O2 sensor sits at 1.0v and switches when revved as it should. 

Had a look at my ECU as well and it has some "warranty void if removed" stickers that do not look Peugeot original. 

Am I just being dumb and missing something really obvious?

 

 

ITS THE ECU WOO HOO. 

As a hail Mary I decided to try that BX ECU again and it's spot on now, it's red hot, fan kicked in and still has a steady idle with O2 sensor switching as it should. Now I need to find a ZX ECU or get this one repaired. 

Edited June 16 by 309jazzpanda

[DamirGTI 340]

Possibly some wiring loom error then ? previous repair or you're looking at the ECU terminals from the opposite side ?

Sticker wise likely something was done to it , needed repair ... i've seen a few on sale stating "defective" , and one with an ECU repair shop sticker .

 

Not sure if there's real closed loop on these like modern cars with Zirconia , me thinks not . Haven't really had time to try and play with the O2 signal and WB meter to see what this system does compared to Motronic 1.3 .

 

Not a lot info with regards to Titania O2 either , as Zirconia is more common in use nowadays on modern cars combine with WB AFR sensor .

But for what i remember and if remember correctly , readings wise it's opposite on Titania to the Zirconia ie. high readings on Tinatia meaning lean , low rich .

 

D

[309jazzpanda 41]

52 minutes ago, DamirGTI said:

Possibly some wiring loom error then ? previous repair or you're looking at the ECU terminals from the opposite side ?

Sticker wise likely something was done to it , needed repair ... i've seen a few on sale stating "defective" , and one with an ECU repair shop sticker .

 

Not sure if there's real closed loop on these like modern cars with Zirconia , me thinks not . Haven't really had time to try and play with the O2 signal and WB meter to see what this system does compared to Motronic 1.3 .

 

Not a lot info with regards to Titania O2 either , as Zirconia is more common in use nowadays on modern cars combine with WB AFR sensor .

But for what i remember and if remember correctly , readings wise it's opposite on Titania to the Zirconia ie. high readings on Tinatia meaning lean , low rich .

 

D

Going to have to get this *212 ECU tested/repaired. I haven't driven it as I know the map will be wrong but using the BX *206 ECU it now idles as it should with lambda reading and a steady 900rpm tick over. That's with the sensors plugged back in as they should be, so it's 100% an ECU issue, Swap the *212 one back and it faults again, which confirms it for me. Once the cold start mode goes off the ECU just isn't deciphering the information provided correctly. Possibly had a poor repair in the past which would explain the warranty stickers. 

Messaged ATC to see if they still repair them as it does list them on their website.

[SRDT 169]

You don't have the 54 error code so the EPROM should be fine. You could try to put it inside the BX ECU and see how it goes.