Cold start fuel enrichment (Choke)
Discussion
Hi guys, My Chimaera 4.0HC starts ok when very cold and idles very roughly but when pressing the accel it nearly stalls as if there is no choke enabled. This improves as the engine warms up and if fine when at normal (70deg) temp. What should i check apart from the stepper? I've also noticed, which may be related, that my Rev gauge doesn't work most of the time but seems to come on/off at random. Also when i slow to stop at lights say, the engine revs at around 1800rpm for about 10 seconds before dropping to around 1000rpm and idling fine. Perhaps these are related as the ECU reads many sensors?
Thanks in advance
Thanks in advance
My Chimaera 500 is the same. I like to idle it for maybe 30 seconds to a minute before trying to move off dependant upon the time of year and outside temperature.
In my rush to move off early yesterday morning, to keep the noise down for the neighbours on a Sunday morning, I managed to stall it. It is definitely a 'rich cut'. I ended up with a flooded engine and generated far more noise getting it going again than I saved by trying to move off too quickly!!
In my rush to move off early yesterday morning, to keep the noise down for the neighbours on a Sunday morning, I managed to stall it. It is definitely a 'rich cut'. I ended up with a flooded engine and generated far more noise getting it going again than I saved by trying to move off too quickly!!
The fuel enrichment is set with the initial injector pulse width pulse for a fixed period (about 3 seconds) until a stable voltage can be taken from the AFM, at which point the pulse width is set to match the air flow, dependent on engine temperature and base mapping. The mixture may still be wrong at this point (normally they over fuel), and will remain so for about 20 seconds until the Lambda probes are hot enough to come into play and correct the mixture, and then it should drive OK(ish). If however the ECU does not get this feedback, it will never trim the mixture and just rely on the base mapping and you would never know otherwise most of the time. If you listen to the cold engine note, its quite marked as it trims the mixture, from the "on choke" sound to the normal idle so its worth listening for at that 20-30 second interval. Ive not had a chance to look at a car with this specific fault, but quite a few people seem to have the same issue, but without taking some readings on a cold start I would not want to diagnose further.
To get the car to idle correctly, the stepper has to be clean, the speed pulse from the TVR box of electronics under the dash has to be at 0 volts,(these can fail and keep producing a pulse when stopped) and the throttle pot base setting has to be low enough to let the ECU know its at idle.(first point to check apart from the stepper). Apart from this air leaks, or mis fueling, or ignition problems can generally make the idle unstable.
To get the car to idle correctly, the stepper has to be clean, the speed pulse from the TVR box of electronics under the dash has to be at 0 volts,(these can fail and keep producing a pulse when stopped) and the throttle pot base setting has to be low enough to let the ECU know its at idle.(first point to check apart from the stepper). Apart from this air leaks, or mis fueling, or ignition problems can generally make the idle unstable.
Ref. the rev counter
The rev gauge takes it's signal from the coil -ve via a one of a pair of resistors (black plastic rectangular jobbies) mounted in the engine loom near the coil, if the terminals are corroded you can start getting issues. It's worth noting the other resistor feeds pin 37 of the ECU and is basically what tells the ECU that the engine is rotating - if thats getting a poor signal too you could be having alsorts of wierd and wonderful issues.
Regards
Iain
The rev gauge takes it's signal from the coil -ve via a one of a pair of resistors (black plastic rectangular jobbies) mounted in the engine loom near the coil, if the terminals are corroded you can start getting issues. It's worth noting the other resistor feeds pin 37 of the ECU and is basically what tells the ECU that the engine is rotating - if thats getting a poor signal too you could be having alsorts of wierd and wonderful issues.
Regards
Iain
Its a plasic geen box about the size of a packet if fags, marked with something like Speedometer calibration unit. The one I pulled to bits was a late Chim, where the top of the dash came off to expose the loom and back of the instruments, but I gather this is not the same on all cars (ask Spend on this one). Before you rip it out, its worth checking the speed sense signal on the ECU is at 0 volts with the car stopped. A test meter may not be able to measure the speed signal acuratley, as the signal is "spikey" so it does not sit at 5 volts for long enough for a DVM to grab it, but you may get some indication. Idealy you need an osciloscope, or you might get away with an LED and 220 ohm resistor to ground as a test.The image here is approximate, due to the limitations of my data logger.
Edited by blitzracing on Tuesday 23 November 15:26
I did a very basic test the other day to see if I could improve the cold start characteristics.
Removing the connector from the coolant temperature sender (ECU one) I connected a spare sender.
With the spare sender in a mug of hot water I was effectively starting a cold engine that the ECU thought was fully warmed.
So what were the results?
CRANKING: Engine cranked for 4 to 5 seconds longer than normal before firing.
IDLE: Engine maintained smooth & reliable idling straight away.
THROTTLE RESPONSE: Unlike with a cold ECU temp sender it was possible to increase the revs on the cold engine without the engine cutting out.
The revs did occasionally dip a little, but it was significantly better.
Essentially the engine went from it's normal un-drivable state to a car that can be driven gently straight from cold without stalling.
So in summary the engine did take a little longer to start but when it did start it ran so much better than normal.
The next step is to test what effect the fuel temperature sender has on cold start by conducting a similar test.
After this its a case of running some off car tests on the senders, placing a sender in a beaker of water while increasing the temperature should give me the operating resistance range.
I can then use this data to fit a resistor to fool the ECU into thinking the engine is warmer than it is.
My simple initial tests seem to indicate this will more or less solve the cold start over fuelling and cutting out problem.
With the cam shaft lubricated by splash I doubt it sees much oil at idle especially when the oil is cold & thick.
So if you cant touch the throttle for a few minutes when the car is cold I am sure it isnt good for camshaft life.
More to follow.
Removing the connector from the coolant temperature sender (ECU one) I connected a spare sender.
With the spare sender in a mug of hot water I was effectively starting a cold engine that the ECU thought was fully warmed.
So what were the results?
CRANKING: Engine cranked for 4 to 5 seconds longer than normal before firing.
IDLE: Engine maintained smooth & reliable idling straight away.
THROTTLE RESPONSE: Unlike with a cold ECU temp sender it was possible to increase the revs on the cold engine without the engine cutting out.
The revs did occasionally dip a little, but it was significantly better.
Essentially the engine went from it's normal un-drivable state to a car that can be driven gently straight from cold without stalling.
So in summary the engine did take a little longer to start but when it did start it ran so much better than normal.
The next step is to test what effect the fuel temperature sender has on cold start by conducting a similar test.
After this its a case of running some off car tests on the senders, placing a sender in a beaker of water while increasing the temperature should give me the operating resistance range.
I can then use this data to fit a resistor to fool the ECU into thinking the engine is warmer than it is.
My simple initial tests seem to indicate this will more or less solve the cold start over fuelling and cutting out problem.
With the cam shaft lubricated by splash I doubt it sees much oil at idle especially when the oil is cold & thick.
So if you cant touch the throttle for a few minutes when the car is cold I am sure it isnt good for camshaft life.
More to follow.
I think messing around with the coolant sensor is a hard way of going about it. Why dont you just fit a variable resistor to start with, set it to match the sensors cold value, then start the car and then reduce the variable resistor until the engine runs nicely when cold. This will give you your basic setting, and with a bit of ohms law you can then work out how much you need to drop the sensors value by, by using a parallel resistor.
Now adding a variable resistor is a great idea! Should in theory be able to fine tune the cold starting when say outside temps get to or below a few degrees. Perhaps even run a cable inside and leave it connected........if any one tries this let me know. Best potential solution to this cold/over fuelling problem i've heard on this site.
blitzracing said:
I think messing around with the coolant sensor is a hard way of going about it. Why dont you just fit a variable resistor to start with, set it to match the sensors cold value, then start the car and then reduce the variable resistor until the engine runs nicely when cold. This will give you your basic setting, and with a bit of ohms law you can then work out how much you need to drop the sensors value by, by using a parallel resistor.
FIRSTLY: from what I have read here nobody has solved what seems to be quite a common and annoying condition.Actually I think I am being quite charitable by calling it a condition.
What it should be referred to is a fault.
Lets face it if I can set up a pair of SU carburetors on on MGB to provide better cold start performance (& I can & so can my granddad), why the hell should this Lucas fuel injection be so bad at it?
SECONDLY: All I am doing is a few basic tests to prove or disprove some logical theory, again it doesn't seem (from what I have read here) that anyone is making any genuine efforts to do the same.
Its all very well to discuss fitting a variable resistor but have you actually tried it yourself?.
A resistor may well end up being part of the solution but I was always taught to move forward in small proven steps rather than take ill-informed leaps of faith.
So far I certainly haven't seen any evidence of anything more than conjecture, speculation and guesswork regarding this fault.
That's why I thought some testing would help us all, and basic scientific testing always starts at the beginning and records the results in a systematic way.
If the objective here is to find a solution to the fault at least I have taken some practical steps to start us on the journey.
Lets see where it takes us, I would speculate it brings us a lot closer to a solution that armchair criticism.
Edited by Chimpafrolic on Sunday 12th December 20:59
Here here.
Even with the current Distributor etc surely a re map should work? I have mentioned that in other posts and been told that remapped chips are available (Mark Adams etc). However none of these seem to solve the problem. WHY NOT. Other posts tell me about waiting for Lambda sensors to warm up (30 seconds or so). WHY, most modern cars with full mapping don't need to. I understand that the current system cannot adjust timing unless the distributor is replaced with coil packs etc. The current system still replies on vacuum advance and weights. At low revs the vacuum surely is not high as the cam lift is high?
Sounds like the only real result is a FULLY electronic mapped system which, in theory, COMPLETELY solve the problem.
Now has anyone done this and does it FULLY solve all this nonsense?
Even with the current Distributor etc surely a re map should work? I have mentioned that in other posts and been told that remapped chips are available (Mark Adams etc). However none of these seem to solve the problem. WHY NOT. Other posts tell me about waiting for Lambda sensors to warm up (30 seconds or so). WHY, most modern cars with full mapping don't need to. I understand that the current system cannot adjust timing unless the distributor is replaced with coil packs etc. The current system still replies on vacuum advance and weights. At low revs the vacuum surely is not high as the cam lift is high?
Sounds like the only real result is a FULLY electronic mapped system which, in theory, COMPLETELY solve the problem.
Now has anyone done this and does it FULLY solve all this nonsense?
Has it occured to anyone that the reason this "problem" has had little air time on here (or anywhere else for that matter) is that there isn't actually a problem with the majority of cars?
My car is completely standard and whilst it probably doesn't run like a 2010 Lexus from cold I have had no problem driving it from cold, neither have plenty of other owners I know.
Maybe those with cold start/running problems have other normal running issues they are unaware of.
My car is completely standard and whilst it probably doesn't run like a 2010 Lexus from cold I have had no problem driving it from cold, neither have plenty of other owners I know.
Maybe those with cold start/running problems have other normal running issues they are unaware of.
Well it seems that all standard, normal and fully working Chims exhibit the over fuel cold start hesitation etc. Depends on one's perspective but as far as i'm concerned, it's a problem and a half baked design that TVR never resolved perhaps because they thought the cars were for warm sunny days and/or track days were the cars are warmed up. Not for everyday use where it might get cold. Much as it's a wonderful car (potentially), sadly it's yet another example of a great British idea not developed.
captaintact said:
Well it seems that all standard, normal and fully working Chims exhibit the over fuel cold start hesitation etc. Depends on one's perspective but as far as i'm concerned, it's a problem and a half baked design that TVR never resolved perhaps because they thought the cars were for warm sunny days and/or track days were the cars are warmed up. Not for everyday use where it might get cold. Much as it's a wonderful car (potentially), sadly it's yet another example of a great British idea not developed.
Complete rubbish.You obviously know a lot about the subject, what is the first thing Mark Adams does before remapping a car?
captaintact said:
Please enlighten me
He makes sure everything that is standard on the car is set up correctly and working properly before he does anything.When did you last change your plugs, plug caps, leads, d.cap, r.a ............ all of which can affect warm running to a lesser extent and cold running more.
Mine has full TVR history. Last full service in the summer. I do not know Mark Adams or have one of his chips. I am in no way rubbishing him so no need to be defensive. My comments are purely as a result of what i read on this forum.
I am not an engineer or mechanic. This is my first TVR however i have been driving for over 30 years and can correctly service many cars. I have a fair understanding of how things work even if i do not necessarily have the knowledge to make it work. However, i have not read anywhere from anyone who does not have the same problem. So are all of these cars running badly? It is the sort of car one can "tinker" with. Despite the blueprint design, the RV8 is fairly tolerant of poor tolerances. This may be why there is so much discrepancy between cars. As i understand it the current Lucas ECU management is "rough" and does it's job. Some have suggested twin SUs would be better. I might be inclined to agree to some extent.
However as you say "total rubbish" so what do i know?..............
I am not an engineer or mechanic. This is my first TVR however i have been driving for over 30 years and can correctly service many cars. I have a fair understanding of how things work even if i do not necessarily have the knowledge to make it work. However, i have not read anywhere from anyone who does not have the same problem. So are all of these cars running badly? It is the sort of car one can "tinker" with. Despite the blueprint design, the RV8 is fairly tolerant of poor tolerances. This may be why there is so much discrepancy between cars. As i understand it the current Lucas ECU management is "rough" and does it's job. Some have suggested twin SUs would be better. I might be inclined to agree to some extent.
However as you say "total rubbish" so what do i know?..............
Gassing Station | Chimaera | Top of Page | What's New | My Stuff