The Lotus Cars Community banner
  • Hi there! Why not register as a user to enjoy all of the benefits of the site? You may register here. When you register, please pick a username that is non-commercial. If you use a name that appears on any search engine commercially, you must pick another name, whether it applies to you or not. Commercial usernames are for supporting vendor use only. If you want to become a supporting vendor and grow your business, please follow this link. Thanks!

05' Elise surging at partial throttle

10K views 50 replies 8 participants last post by  mm3boost 
#1 · (Edited)
I have a 2005 Elise w/ ~50k miles. Purchased it about a year ago from the previous owner, of which I am number 5. Since taking ownership I've noticed that at partial throttle the car will surge. The behavior seems to start as the car warms up and it only gets worse as the car reaches operating temperature. This doesn't appear to be limited to throttle input either, just more apparent under those conditions since the car can be felt surging. At the track it will rhythmically crackle and pop w/ a hot exhaust as you decelerate and or idle. Even when the car is on jack stands I can hold the throttle down and hear the engine and watch the tachometer needle ever so slightly wiggle. This leads me to believe it's something related to the air and fuel ratio modulating, but it hasn't thrown any codes yet... and I'm all out of ideas especially w/ it throwing no codes or CEL... and trust me I've read most if not every post on here...

Modifications:
Larini Sport Exhaust (PO installed)
Sector111 Motor mount and transmission stud upgrade
Sector111 gPan3+
BAB26-12 Lightweight battery

Things I've addressed / refreshed in the last 2-3k miles attempting to get rid of my issue:
Adjusted throttle cable & stop
New spark plugs
New air filter element
New coils + dielectric grease
New Pre Cat o2 sensor (generic Bosch 15733 wired in w/ OEM plug)
New Post Cat o2 sensor (generic Bosch 15733 wired in w/ OEM plug)
Verified o2 sensors are attached to their correct wire harness location AND wire colors match (see attached image)
Cleaned ground(s) on driver side bulkhead
Cleaned ground on head for ignition system
Hunted around for obvious exhaust leaks but found none except at the Larini narrow exhaust clamp - replaced with a wider one

Things I've tried AFTER initial posting - to no avail:
New MAF & cleaned connector pins
New TPS & cleaned connector pins
New IAC & cleaned connector pins
New TB gasket
New PCV valve
brake booster vacuum nipple at intake manifold was checked and reinstalled w/ teflon paste
New intake hose t-clamps
Clean/inspected throttle body thoroughly during IAC replacement
Clean ECU pins on both connectors
Inspected/Replaced all engine compartment fuses except for the 3.5A (didn't have any on hand)
Checked and double checked for vacuum leaks - no variance in idle w/ judicious application of propane & maf cleaner.
Cleaned / Flow tested injectors - all were good - no change
Disconnect / Isolate brake booster - no change
Disconnect / Isolate evaporative emissions system - no change
Reverted back to good OEM lead acid battery - no change
Smoke tested intake post MAF w/ the stock intake tube - no leaks seen
Smoke tested exhaust up the tail pipe - only leak was very minor at the exhaust clamp POST secondary o2 sensor - no effect
Unplugged primary pre-cat o2 sensor - STFT went to 0% and car ran normally
Swapped o2 sensor connectors - car warmed up but once cat was at temp car was hunting towards a stall. Reverted - original symptoms returned.
Called Lotus dealer and all they found on my car's software was the VIN update - uncertain if this could be to blame (bad software load?)

Ideas? I've heard fuel can be a cause but I'm not sure if that would manifest itself in surging... Anything I can / should test there?
 

Attachments

See less See more
1
#36 ·
I think I should be clear the surging issue is not really reflective in RPM's - it's in the seat of your pants or heard through engine tone changing. So perhaps this term is wrong. It's not so much apparent with engine speed as it's just ping ponging the fuel... so it's as if someone is softly feathering the accelerator pedal making you go mad. Likewise it does not seem to change oscillation based on engine speed.

You can see it in the data under idle but it's far worse 2000-5000 rpms. o2 sensor reading just oscillates as does the STFT.

Past few days I've tried to no avail:
reverted to OEM lead acid battery
isolate evap system
isolate brake booster system
isolated vacuum source from intake manifold
replace tps
replace iac
replace maf
replace pcv
replace throttle body gasket
replace clamps on the throttle body to airbox hose to quality band clamps
felt around for exhaust leaks more
sprayed more maf cleaner around the engine bay

I'll try unplugging the primary o2 sensor and see if it enters closed loop. Just so we're clear the entire project started with the o2 sensor being the culprit and thus they have been replaced AND verified to be correctly connected. So far everything points to the sensor reading correctly as it's mirroring the post cat o2 sensors prior to the catalyst reaching operational temprature. This tells me either I have TWO incorrect sensors, or even more unlikely - two broken sensors to start with then maybe two wrong or defective sensors? Again, two new sensors, behavior persisted. So really the question is - where's the leak?

" you could try logging the ecu commanded injector pulse length (mode 0x22, pid 0x205). this parameter includes more than just the fuel trim influence on fueling." - I don't have fastworks... unless I'm missing something I can't get that via my pretty basic obdII reader (BlueDriver).
 
#37 ·
Disconnected the primary o2 sensor and the audible and digital artifacts disappeared. The STFT went steady and of course the primary o2 sensor was silent. So we can narrow it down to that loop and interactions with it...

My real question is - does this rule out the fuel pump / regulator / filter in the tank? That has been a source of possible issues in the back of my mind, but if the car can seemingly keep fuel steady flying blind on a safe fat fuel map - I'd say with the exception of edge cases the tank contents can be ruled out? Also I've had no issues at the track (Grattan) with fuel starvation or with daily driving conditions (aka fuel baffle / glue in the filter issue). Just this ever present partial throttle scenario slowly getting worse.
 
#38 · (Edited)
Disconnected the primary o2 sensor and the audible and digital artifacts disappeared...does this rule out the fuel pump / regulator / filter in the tank?
I agree with your reasoning. More-so, this corroborates your evaluation of the injectors; if any part of the fuel delivery system was unsteady, then this would still manifest as misbehavior in open-loop fueling.


If I understand the short term fuel trim algorithm correctly, I do not believe that it has any 'learning compensation' capability. It is mostly the sum of values from lookup tables based on engine speed and engine load. It seeks to modulate the upstream oxygen sensor voltage between two threshold levels. Thus, I do not expect an ECU reflash to affect it's behavior unless the calibration is being updated.
 
#42 ·
I went ahead and cobbled together a smoke test machine. The internet standard: e-cig wire + paint can + miner oil + air source. I checked post MAF just in case it was messy... oh, and it was. I included the ribbed intake tube. No smoke found after almost an hour turning the smoke on/off and looking over the entire engine. I also tested the vacuum lines (one way check valve removed) near the intake. No smoke found. Tested the vacuum reservoir - no smoke. I repeated this for the 'hot side' by directly hooking up to the tail pipe. No smoke except very small wiffs at the new exhaust band clamp - but AFTER the post cat o2 sensor - so not at fault. All in all that was my only 'find'. Was half hoping I'd see smoke around the header / heat shield... but no dice. The hunt continues for the phantom leak!

For kicks I swapped the o2 sensor plugs, because who the hell knows... as expected the cat came up to temp and the car hunted to the verge of stall. I turned the car off and reverted the o2 sensor plugs to their verified correct stock install - original symptoms returned, but car idled fine. So we can put the common 'swapped o2 sensor plugs' to bed for good.

What's the best way to find out my ECU software version? I called the dealer and the only records they have of updating my ECU were the standard VIN update for the CEL after your battery dies. Uncertain if that was a discrete patch or if that was an entire software flash... When I leave my door open, press the odometer reset button, and turn the ignition to full on (re calibrate gauge cluster) I see a date that's January 2004 and a v1.0 (roughly) - this isn't related to my ECU software is it?

I'm going to call a local shop on Monday that everyone likes and see if they can fit me in before my Lotus visit 7/16 @ 8AM (Auto Europe, Birmingham, MI)... half tempted to buy the ECU that's on the for sale section at $350 since the dealer is $150~ to just look at my car...plus driving there and back... so I figure 6-10hrs door to door.
 
#43 ·
As far as I know, the numbers in the gauge cluster are not related to the ECU software.

The VIN can be rewritten without an ECU flash, but there has also been an ECU code update that allows the P0630 to be evaluated and set when the VIN is lost. Thus, this could have been or have not been a full code reflash.

You can figure out which calibration ID is on the car with OBD Mode 0x9 PID 0x4. The calibration validation number is OBD mode 0x9 PID 0x6 (this is calculated by the ECU each time it boots with a convolution over 0x6000 bytes at address 0x70000), so it will show if the calibration is corrupted.

OBD mode 0x22 PIDs 0x20e-0x211 will return ASCII characters that indicate which bootloader version is on the car: (16 ASCII chars, ex. "C121E0001F K0303")

OBD mode 0x22 PIDs 0x21c-0x21f report the calibration ID another way: Cal ID in ASCII (ex. "B121E0002H ")

OBD mode 0x22 PIDs 0x221-0x224 return the make, model and year that the ECU code is intended to run on (I think this is just supposed to identify the car): (ex. "Lotus Elise MY05")

More specialized tools are needed in order to read the ECU code and calibration directly.

I don't know what numbers are usual responses to some of these queries since I've never asked my car for these all of these PIDs.

I've had good experiences getting paint protective film replaced Auto Europe, as well as a tour of a new Evora 400. I hope you have a good experience too.
 
#44 · (Edited)
I got a new 'known good' ECU in the mail someone on here was using for emissions testing on their 2005. The partNo on it was slightly different: C121E0001F, the original ECU was B121E0002H - cautious at this point I almost didn't install it. After a little googling it appeared both are acceptable so I disconnected the battery and threw it in, every so gently.

The new ECU seemed to run richer - which isn't shocking given how they seem to 'learn' over time. After first impatiently not letting it idle for 15min I got a CEL going down the street: too rich. Duh. AFTER letting it properly letting it idle for 15min the CEL did not persist and the car ran much better (aka not bogging as much). I went out for a test drive - original problem persisted. Boo. So it's not the physical ECU - still can't rule out software but I find that unlikely as this ECU is allegedly out of a car that easily passes emissions and was known to run well.

Data from the car being at operational tempratures - starting from stop light and going 30-35mph you can easily see the STFT and (new w/ ECU) LTFT adjusting:
https://drive.google.com/file/d/1XWWMehF33TynlTbR4iTCFk2Eh7CTIjXK/view?usp=sharing

Scratching my head what the source of the problem could be now.... fuel pump / fuel pressure regulator in the tank? wiring harness?
 
#45 ·
Update:
Exhaustively checked continuity between ECU and primary o2 sensor. NO FAULTS detected. All wires correctly connected and not grounding out or broken connections. Doesn't rule out an intermittent fault as I cannot easily test with the car running. I also did not exhaustively check every pin on the same wiring loom fore connectivity issues with neighboring wires (hopefully I don't regret this in a year...we all know these decisions...).

After talking w/ Andrew @ BoE I didn't have much faith it'd work - but I paid a visit to Auto Europe in Birmingham MI. They were not exactly enthusiastic to solve my problem. To be fair on the rough SE MI roads the problem was hard to detect. So I re-created the problem, showed the data, talked w/ the technician. Long story short - I spent 8hrs in a waiting room and all they did was re-set my ECU (reverted back to original ECU after testing the new one I swapped in). Didn't appear they attempted to trouble shoot the issue at all beyond reading the write up I gave them. In the end they agreed to reprogram/reset the ECU. Unfortunately according to the tech my ECU was up to date, and he could not reload the software (ehhh - really?), only reset the ecu's memory. So 8hrs waiting and $260 later I tested the now re-set car - immediately the problem was present. Before leaving I was assured it was fine and the car was meant to be driven all out, not at partial throttle... right.

Not phased I proceeded to dig a little deeper into the Toyota service manuals. The procedure for testing the fueling system is to prime it and verify the pressure is held. If the pressure drops it basically tells you to inspect all things fuel. As all lotus owners know - the pressure did drop. So I caved and decided to attempt every 2005 Elise owner's nightmare: fuel pump and basket bits (pressure regulator, sock, and filter). Unlikely as it seemed there's still a chance it was failing partially due to a clog, electrical issue, etc. A week later the nightmare is over and the new parts are in. I let the car idle for a while after I adjusted the base idle speed (had to remove all the cables to get to the shear panel). Problem persisted... ouch. So what else could be leaking? Well the entire system by design could have leaks. I don't smell fuel - but having had longer / hard starts before I do suspect my "flow tested and cleaned" injectors... but I feel that's just throwing more money at the problem with no conclusive reason to suspect the injectors. What are the chances an injector could fail at partial throttle and instantly be fine at WOT or at high speed cam? There's a chance, but it seems fairly low... probably as low as the fuel pump being to blame in this scenario...

Shot off several emails to Lotus and Phil @ BoE. Curious what inputs are used for the short term fuel trim (STFT)... Does anyone here definitively know?

My problem is very specifically boxed: any scenario where the STFT is adjusting
30s after warm up
partial throttle very detectable
disappears at wide open throttle
disappears when the high speed cam lobes engage

Problem does not exist: any scenario where the STFT is flat lined
All sensors read normal - including the o2 sensor which then behaves in a more normal fashion in a steady state..
 
#47 ·
My current attempts to fix:
Replaced/swapped the injectors with known good new ones from Monkey Wrench Racing. Just in case my "cleaning/flow tester" was a sham. With the exception of the new injectors being louder, nothing changed. Last but not least I went ahead and swapped the bosch o2 sensor for a denso one. Again, nothing changed.

Out of the blue I ended up getting a call from Auto Europe saying they would like a chance to look my car again and feel they CAN fix it. Given their level of optimism and desire to get another crack at it - I agreed. They're supposed to call me back next week and arrange pick up of the car so I don't have to drive the hour plus to hurry up and wait.

While I wait for my second visit to Auto Europe I went ahead and scheduled up my third and fourth track days of the year. My problem seem to be getting any worse and i'll be within AAA towing distance of home... so I figured "nothing ventured, nothing gained" for enjoying my car on the track. At the very least if it does strand me - I'll at least have a solid lead.
 
#49 ·
Bueller?

Another complication in diagnosing this behavior is that the ignition advance reported over the OBD port is not necessarily the value in use by the ECU. At low engine speed/partial throttle the ECU smooths changes in ignition advance to something like 1-2 degree changes per 50 milliseconds. The value in use is always approaching the value reported over OBD, but these values are not necessarily the same.

Also, if you'd like to talk about the STFT control strategy I can walk you through the ECU code or through a simplified explanation: https://www.lotustalk.com/forums/f171/notes-about-2005-elise-fueling-timing-control-463664/
 
#51 ·
It appears I’m beginning to have the same issues. My car has slowly started surging/bogging almost like a minor fuel cut at light load transitions. Full throttle or steady state throttle is typically ok. Hoping this problem was eventually solved?
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top