Request for advice, car won't idle

[Jonathan Kay]

+ Swapping in a good one and looking for changes. But that won't catch some bad connections. And it does need to be a good one.

Jonathan

[CtrMint]

Thanks and agreed Jonathan, although it would be handy to diagnose before purchasing. 

I'm only aware of the main loom connection for the O2, and of course the main ECU harness plug.  Anywhere else of concern connectivity wise?

[StevehS3]

I am no expert either but I think that if the lambda is faulty the ECU would detect that and go to open loop anyway. So disconnecting the lambda might make no difference for some or all of the time.

Personally, I would want to get to the bottom of the problem rather than just accept it. It’s a modern engine which should run smoothly, particularly if it did before. If it doesn’t idle properly then it probably doesn’t run properly when driving and the engine is on-load - it’s just far less noticeable.

[StevehS3]

Changing the lambda sensor is a bit of a shot in the dark but as it gets very hot is maybe more likely to fail, maybe. There are other sensors too such as ambient temp, engine temp, barometric pressure etc. which could be the cause. 

[ScottR400D]

Mine has its original sensor from 2015, I had the uneven idle etc until changing to RBTBs in 2016. After the RBTBs were fitted the popping and banging increased of course.

More recently I’ve fitted a decat pipe. Still the same lambda sensor. 

Still idles pretty much perfectly. 

[John Vine]

Re #25:

Without using Easimap to see what's truly going on,  debugging a problem like this is, I'm afraid, going to be a somewhat hit-or-miss affair.  For example, it could be due to a faulty lambda sensor, poor earthing, or a poor lambda connection.  But it could also be due to a TPS problem such that the idle position is not being translated into the correct mapping site -- without seeing what Easimap has to say, you won't be able to tell.  The irony is that cash spent on unnecessary replacements adds up to the point where you're well on the way to affording a diagnostic cable!

I understand your reluctance to shell out on an MBE cable (yes, they are crazy expensive, but at least the software is free!), but I feel that's really the only viable DIY solution given your remote IoM location. 

If it's any help, I have an MBE cable which I'll happily lend you, provided you're willing to pay two-way postage and insurance.   If you're interested, just PM me your postal address plus a contact email address too.  But I won't be able to send it off until the end of next week as I'm abroad at the moment.

JV

[CtrMint]

Scott,  Your experience does suggest they are more reliable than I'm giving them credit for, but I do believe something has changed, I just think its the most likely culprit.  

JV, Thanks JV, that is an incredibly generous offer, but I do agree with you from a strategic position it does make sense to arm myself with the cable, so i'll graciously decline and invest this weekend.  I spoke to SBD on Friday and confirmed the payment process, just need to commit.  It's not the cost that's holding me back, more the chance the existing CAN cable I have might work, as per the conversation earlier in the thread, and I don't want to waste 140GBP if it could be avoided.  I'm still trying to get my head around why not, but I think that topic is for a different thread.

[John Vine]

...so i'll graciously decline and invest this weekend. 

No problem, and a good decision.  You'll need a suitable PC laptop, of course.  I use my wife's old Vista PC and that works fine.  The great plus of Easimap, IME, is that you can log stuff while driving.  It may be that you'll need to do that, given that your problem occurs after a period of running.  Make sure your PC battery is in good nick, though, as if you lose power while logging, you'll lose all data, plus it's probably better not to be charging while logging.

I'll be very interested to hear how you get on -- definitely a new chapter for your entertaining blog!

JV

[CtrMint]

As a non Windows user I've opted to resurrect an old Macbook Air, it's now running Ubuntu 18 with KDE plasma, I've then setup Windows 10 Pro within Virtualbox.  Windows license cost 19.99.  The guest has x2 cpus and 4GB which is just about enough to cut the mustard with Windows.  Easymap is installed and working. Just need the cable now. 

 

[aerobod - near CYYC]

The default mapping of the MBE ECU for the R400 with plenum (assuming they are using the same or similar map for the 420R) is very hard on the Lambda sensor as it runs rich in many places. I killed my first sensor in less than 10,000km with carbon deposits, although my altitude makes it run even more rich. I also found the default map is badly tuned for idle with high ambient temperatures, above 25C when stuck in traffic it tended to stall easily when pulling away, just had to give it a bit more throttle until the engine compartment temperature dropped a little and things equalised with the air temperature sensor.

My car has none of these problems with the unlocked ECU I'm currently using with a tweaked map.

[aerobod - near CYYC]

The guest has x2 cpus and 4GB.

Easimap will run well with that resource, it is quite efficient. The only performance issue I have noticed is more around having a flash drive vs spinning disk. With large log files and 30 channels logged, the graphical analysis can run quite slowly when all parameters are shown on the screen for long time frames, the flash drive seems to help a lot in that circumstance with the performance of windows paging, speeding up the rendering a lot.

[CtrMint]

Windows has a 50GB partition on the MB Air's internal SSD so we should be good, but thanks for the advice.

[John Vine]

The default mapping of the MBE ECU for the R400 with plenum (assuming they are using the same or similar map for the 420R) is very hard on the Lambda sensor as it runs rich in many places.

Very interesting, James. 

Does that apply to TPS-based maps as well as MAP-based ones?  My R400D (with plenum) came with a MAP-based map, and the idling was never very smooth. This was changed to a TPS-based one by the Two Steves during a mapping session, which improved things a lot.  I currently have roller-barrels with CC's standard RB map, which gives a very even idle and a generally satisfactory performance.  No doubt a Two-Steves remap would make things even better, but I'm reluctant to spend the thick end of £500 to find out.

JV

[aerobod - near CYYC]

Hi John,

That is for MAP based, the way Caterham shipped it. My current mapping on the 9A4 ECU is TPS based, the MAP sensor is just along for the ride and only useful when logging data, as I'm also using an IAT sensor in my custom air intake that uses the bonnet louvres, as opposed to the integrated one in the factory MAP sensor.

[John Vine]

Thanks, James.

JV

[CtrMint]

I've been reading as much useful information as I can find on the O2 sensor, this article seems straightforward and well documented.

http://www.autotap.com/techlibrary/understanding_oxygen_sensors.asp

From what I understand when the car and O2 is cold, the ECU will go to its open loop configuration which is a default fuel mixture.  Here it will remain until the O2 and engine reach sufficient temps to switch to close loop and start to flip flop between levels. The open loop should also be the same state if the O2 is missing or disconnected.

If this is the case, is it reasonable to state disconnecting the sensor should produce the open loop default mixture, which was totally acceptable on my car at the beginning.  If the car isn't happy in the forced open loop state and when cold, then there's likely to a be a problem elsewhere and the issue isn't the O2.  If the forced open state produces a steady smooth idle back at 950rpm as per original operation then it's the O2.

Hope that makes sense, seems logical to me, though I might have over simplified the sensory ecosystem which produces correct operation.

I've bought the cable now, just wanting to learn as much as possible.

[CtrMint]

Two possible build issues relating to the O2 which might contribute to the problem.  Untested as yet as it's raining hard and I'm not eager to poison myself in the garage again.

Electrical tape on the cable.

When I built the car the protective sheath over the O2 cabling at the back of the sensor started to fray, so I placed some electrical tape around the end.  I recall at some point tidying that tape up again and affixing it to the back of the sensor.  Have I blocked the reference air flow?  I've now been and removed, plus lightly cleaned with a rag and spot of IPA to remove residue.  From what I can see the 4 wire O2 used doesn't pick up the air from the plug so it must be from the back of the sensor.  Does anyone know for sure?

O2 Not tight

I've just been and put a spanner on the O2 nut and found it more than finger tight, but not as you'd expect.  Possible it was sufficiently torqued at build and has slackened with driving and heat cycling causing an air leak?  I've nipped it up now, probably around 30 Nm, obviously can only put a spanner on it.

It would be fantastic if one of these was the cause.

[John Vine]

The sensor takes its ambient air supply through a channel where the wires enter. So, yes, the tape you applied might well have restricted the reference air flow. I'd be surprised if the lack of tightness affected things much, but it's possible.

As you've made two changes, you can't be sure which one (if any) was the culprit. I suggest you try loosening the sensor to how it was and then test again.

JV

[CtrMint]

JV I hear you, though I’ll just be happy if it solves the problem.

Have to say though even if the issue isn’t fixed or even related to the O2, while researching the problem I have learnt loads about the sensor.  Really enjoyable stuff, mind I’d be less inclined to think so if it wasn’t peeing it down.

[CtrMint]

Hi James, 

Thanks for supporting my ECU thread.

Just wondering if you've ever tried using your MBE985 cable on Linux with SocketCan and CANard etc?

Thanks

Mark

[aerobod - near CYYC]

Sorry, I haven't.

[CtrMint]

sorry wrong thread...doh 

[CtrMint]

Well after much faff, (had to install Easymap on Windows using VMware player) I've got Easymap working and connected to the ECU.  Only managed it about an hour ago, so didn't start the car.

Given the TPS is being considered as a factor in the idle issue, and I've always been curious whether the car was getting maximum throw in the throttle pedal I've tested max and mins.  I've done this on the pedal and on the throttle body/spring.

Results are shown in the video below.

Via Throttle Pedal

Max[Throttle Site 1 = 14.5, Throttle Angle Increasing = 4.15v]   Min[Throttle Site 1 = 0.0, Throttle Angle Increasing = 1.13v] 

Via Throttle Body / Spring

Max[Throttle Site 1 = 15.1, Throttle Angle Increasing = 4.70v]   Min[Throttle Site 1 = 0.0, Throttle Angle Increasing = 1.13v] 

At present I'm not sure how this impacts the map, I'm still learning, plus had about 30mins with the software at best.

---Edited---

        Lastly, any advice on the Baro Pressure Fault?  A missing sensor on Caterhams?  or a genuine cause for concern.

       Found this thread, which had the answer.  

 

Thanks for the help and support.

[aerobod - near CYYC]

I would check the TPS screws are tight, then adjust your idle throttle stop screw until you just get to site 0.1 (probably 1.25 to 1.30V from what I remember). At the top end the cable, throttle pedal (via bending !!) or any pedal stop may need adjustment to get to 15.1 on the pedal. If you have slack in the cable at idle (after adjusting the throttle stop screw), it would be best to adjust it first to get to site 15.1, leaving slight slack at idle, but don’t adjust any further than just hitting the maximum position, as you could snap the cable at full throttle. If any further range of motion is needed, I believe that you would then need to gently lever the throttle pedal upwards with a block of wood, assuming there isn’t any pedal stop (non on my car) or a pedal stop is adjusted all the way in.

[CtrMint]

Thanks James,

Do you know if there is a reference table for what those values translate to in the map, or is that part of CC’s hidden data?

thanks