Throttle position sensor

[chrism]

11.zip

Hi All, after some info from those in the know, please...

Today visited pal, car was running fine. Started up to go home and lumpy shaky...when on the road max 30mph jerking shaking etc but no CEL so I nursed her home and did a short 2min EScan.

Comparing before and after EScans my before TPS voltage was steady at ~0.4V but this after EScan shows loadsa fluctuation.

I'm supecting TPS sensor but I know nothing about it apart from where it is

Can anyone offer advice about how to remove it and test it please?

ps I haven't done a paperclip test to look for engine fault codes (yet) cos there was no CEL light...

[matts1972]

Hi Chris,

Cannot open your file right now, but if you need a replacement, you can get one cheap from China:
https://forum.lotuselancentral.com/viewtopic.php?f=29&t=10435&p=381231#p380742

Cheers,
Matt

[KenS]

Hi Chris, be cautious in investigating the TPS, the plot auto scales the axes so its zoomed in on the short run plot which makes it look quite erratic, but in fact its only jittering between 0.40V and 0.42V which is not that unusual. But after looking at many plots recently it hard to know what is normal! I think it might be worth getting a longer cold start warm up plot before going too far.

[KenS]

Chris, do you have a reference set of plots for your car which would allow injector pulse widths to be directly compared, just looks like on this short run the pulse widths are around 3ms which may be a bit wide?

[chrism]

Thanks for that Matt + Ken. Yes it was a short, panicky, warm Escan as soon as I got back to see if there's anything obvious. I'll do a cold scan tomorrow.

Quite disconcerting to suddenly go from 'normal' to a pretty good impression of a tractor tho'...

[chrism]

ps re inj pulse widths, today's is in blue and my 9/9/26 one is in black:-



Ken perhaps, as you've suggested, it's a bit early to point the finger at TPS, then. Is there a way of testing it through EScan, I wonder?

I seem to remember Tuga mentioned looking at an EScan live view throttle%(?) in the past but haven't been able to tie it down...

[KenS]

Is this the TPS diagnostic procedure (from Engine Management Manual section EMK.4 page 38)

tps.jpg

[chrism]

Ah that's brilliant Ken, thanks a lot. That'll be my first job tomorrow.

Earlier tonight I whipped out the plugs; they were all evenly freshly sooty from the short limp home. I nylon-brushed them and checked their gaps. I nearly took out the O2 sensor to clean or replace - cos now it's not swinging as widely as before - but can't think why this should be so, and don't want to unnecessarily tinker.

ps Tuga must've been using that 'Tech 1' device then?...

[KenS]

I guess Tuga ran the same procedure but using the ElanScans left live pane to monitor 'TPS Voltage' and 'Throttle' which should replicate the 2 Tech1 results.

All 4 plugs being sooty doesn't make it easy, does that rule out an injector or ignition issue? Hopefully an ElanScan will reveal something.

[chrism]

Thanks to Ken and still thinking I compared MAPS graph on EScan:-


...my (from cold) before is black, my today's (warm) tractor is blue...

My local pal 'Peteoddcar' hadn't seen this but thought the manifold press sensor hose could have come adrift somewhere - so I compared my own EScans shown above.

He told me where to find the MAP sensor - that swine's really well hidden on the bulkhead right near the headlamp lift module - and I found the pressure tube was okay but the 3-wire connector m/plug had been pushed out at an angle by a couple of stiff wires (the speedo cable being one I think...) I re-routed the stretched MAP wire and plugged it in properly.

Can't wait to try it tomorrow. Even if it wasn't totally the cause of the rough limp-home running it could have happened in the fast lane or at night, even... rather than when I parked up to see a pal.

ps could it be coincidence that this MAPS and my previous injector pulse width cycle seems to be similar'? The injectors could be opening longer cos the airflow's more, I wonder? They should be ~1.6ms (as Ken said) but they were ~3.6ms - causing drastic over-fuelling and sooty plugs to try to get to the right mixture?

[chrism]

Well, the good news is that the TPS sensor test (engine off, ign on, depress acc pedal Escan talking...) worked okay and went from 0.4V->4.06V on TPS voltage and 0->100% on Throttle so I think I can rule it out and apologise to the TPS gods...

Bad news - tractor tickover still + smelly exhaust etc even after cleaned + gapped plugs and (today) cleaned sooty O2 sensor...

I've attached healthy 9/9/26 and unhealthy 4/10/26 EScans in case some benevolent soul (with too much spare time) can spot something amiss by comparison; that is, if they want to share in my stess levels

I've noticed MAP is up, as is inj pulse width and a 'from cold' Escan showed a really slow temp increase. I stopped it on 12 min and the t'stat hadn't even opened!! Surely not temp sensor?

I didn't check the dash temp gauge on this morning's EScan but as an afterthought I've just started up again for a min or two and it seems to want to rise slightly from cold. I'd have expected more of the water temp to be retained so I'm thinking more bad temp control thoughts(?) CTS? Aaarrghh

BTScan 9-9-26.zip

EScan 4-10 -26.zip

[dapinky]

I'm starting from a default position of "It's probably the CTS, because that sounds like CTS failure symptoms"...... and looking for evidence to support the theory (which may well be the wrong starting point).

Anyway, I see on the older (September) scan that the coolant temp got up to the point of the thermostat opening at 80C(ish) just under 10 minutes - but on the latest scan at 10 minutes (600 seconds) the temperature is 'only' reading IRO 60C and still climbing steadily.

Obviously, the ambient temperature in Sept was a tad warmer, but both scans seem to start from a couple of degrees difference, so that can be discounted a bit.

I have overlayed the 2 results and the blue trace is your September run, black is the October.

I have also added a couple of old results I have on the laptop...... the pink (low) line is from when I had a defective CTS earlier this year...... and the Green is from the first run after I replaced it.

The darker blue is from a different car, when I changed the coolant, pump, sensor and thermostat many years ago.

CTS comparison.jpg

I notice that your graphs both take significantly longer to get warm - and the graph 'shape' is different, rate of change is slowing on yours, but increasing on the 2 "good" ones...... clearly the duff CTS is a lot different, with absolutely nowt really happening!

Basically, I would expect your car to get hotter, quicker....... but either way, the scans don't show the thermostat opening now, when it just about did in September........

I have no idea what this information "proves", if anything, but it shows that your CTS is doing 'something' - even if not enough!

[chrism]

Thanks for your input dapinky.

Yes I noticed how today's scan showed that the temp was nowhere near 82C therm opening when I switched it off. I didn't like for the car to 'struggle' so much but I needed to a cold warm up to compare with my previous 'good' scan. You know 'fair test' and all that...

I'm a bit stumped atm - esp since it was okay one minute, then severley disabled 30min later However, I've discovered the the TPS 'test' and I'm now confident that it's not defective.

It's gotta be something simple like a disconnected vacuum hose I reckon. I think there are 3 and I've found the MAP sensor hose, so I'm looking for 2 others - and open to suggestion (please folks)

As Ken has pointed out, injector pulse width is probably a relevant symptom but I'm on the hunt for what's causing it. I have a gut feeling that, if not a vac hose, it might sensor-related rather than electrical...(in the hope that replacing a small part will bring huge success )

Perhaps I should start by throwing a new thermostat at it; I'm pretty sure it's still running the original - and they are 'service items'

Edit:- yep, took the 'stat out tonight to replace with shiny new tomorrow. I'm pretty sure it was the original...

[DeStijl]

I spent several months this year chasing an issue with the car stumbling at steady throttle. ElanScan showed the TPS reading correctly between 0-100% the whole time so I ruled it out and ended up replacing all the other sensors trying to get to the bottom of it with no luck. I had the engine idling one day whilst just poking about at things and noticed that applying light pressure on the back of the TPS affected the idle. Immediately ordered a replacement TPS and the car has been running (almost) perfectly since.

[chrism]

Thanks...deffo try that tomorrow...

[chrism]

dapinky, grateful for your temp/CTS input which has certainly crystallised my t'stat replacement thoughts.

You commented about my rate of temp rise; it was literally a 'from-cold-10-min-scan'. I notice thar yours jumped in higher up the curve - mid-50Cs for your pal's, and late 60Cs for your own green trace.

Would you happen to have a 'from cold' trace, please, so's we can 'fair test' rates of temp rise? (Or anyone else on the forum for that matter?)

And also when you replaced your CTS (I might - at the very least - remove mine for testing, or even replace it since I'm up to my armpits in new thermostat/blue coolant stuff) did you need to disturb the CAS (which is one of my great fears )?

[KenS]

Chris - comparison plot for coolant from cold. your recent plot in Black, your older plot in pink, mine from this summer Blue.
So does look like yours has become a bit sluggish. since the CTS it is still recording sensible temperatures and does give a correct room temperature start value of about 16 degC, I would think the CTS is OK. A stuck open, or leaky poor thermostat would be a good reason for the slow warmup.

coolant compare.jpg

More digging into the plots shows not only are the Manifold Pressure (highish) and Injector pulse widths (wideish), the IAC is also highish. All these are still high even if you take into account the relatively low engine temperature. I suppose these will all be inter-related, if the engine is struggling to maintain tickover revs it will increase the IAC to get more air, the more air will push the maniford pressure nearer to 1 bar and the ECU will increase the injector pulse width the maintain a near 14.7% air fuel ratio.

4-10-26B.jpg

So trying to conclude something from this ( and probably more for discussion rather than costly action):
1/ a vacuum leak could be a possibility, but if there was one, why is the IAC also being set to allow even more air? Does this make a vacuum leak unlikely?
2/ could the MAP sensor be faulty? The MAP does record a pressure of just over 1 bar when switched on, so its recording a correct value initially. I am not sure what the effect would be if its gain were to be in error, could that cause the IAC count and injector pulse widths increase - I don't know.
2/ I guess all 4 cylinders are not firing regularly, is this apparent during idling? If thats the case can the missfiring cylinder be identified? Maybe disconnecting each fuel injector plug one at a time to see if one has no effect on its poor running, then similarly disconnecting each plug lead in turn. A poor fuel injector can lead to sooty plugs but not sure if all 4 sooty plugs shoots that thought down. Is there any back firing of popping in the exhaust which might indicate a injection/ignition issue rather than a sensor issue?

[chrism]

Hi Ken thanks for such a quick reply and it's comforting to have a 'from cold' coolant plot for comparison.

Yep ny coolant temp rise is/has been a bit sluggish (new t'stat should correct this ) but (significantly?) so is my current MAT sensor plot compared to the healthy one a month ago:-



The car's really lumpy lumpy and won't rev, smelling of fuel etc. It's so distressed on start-up that I'm avoiding running it; top speed's ~30mph . All plugs and O2 sensor get freshly sooted. There's no popping or backfiring so I'm thinking/hoping that you're right that there might be no inj/ign issue. I've just done the paper clip test to verify there's no CEL registered - and nothing stored.

I have a new 82C t'stat but it's a Landrover part that that doesn't have a jiggle pin, so I'm relutant to fit it but awaiting a call after lunch to see if a Dayco DT1019V (which I can get today - other local parts people say n/a...) has one (Google says Yes...)

Next I think I'll remove and test the MAT sensor - there must be info somewhere on resistance values... and the top feels a bit loose and leaky. I get a feeling in my water that sensor temp data could be being misinterpreted by the ecu (in fact I have a RonR 1499 standard/Everest prom I could substitute to eliminate for ecu)

Thanks again

[KenS]

The MAT sensor does report room temperature as 18decC at startup, so it looks like it is basically working OK. The throttle valve (next to the MAT) is water cooled, so if the new thermostat increases the coolant temperature maybe the MAT plots will be brought more in line with the older plots. Fingers crossed that the new stat helps cross some possibles off the list.

[chrism]

Great point, thanks. I've taken out the MAT and the plastic part is a little loose. I can suck/blow air through it with difficulty but that shouldn't happen imo.

A new Delco part for GM cars is only £7.99 so I'm gonna replace it as a matter of course...I expect it'll test okay electrically but if there's a chance it's leaking...

Can I get away w/out a jiggle valve in the t'stat I wonder?