K series Cold Start

[Geoff Martin]

Cant get my 1.8K to start on the button anymore. Always needs a touch of throttle. If it turns over long enough it may initially catch, but not enough to start fully and dies. As mentioned before, with slight throttle, it starts fine.

I had to change the air idle control as it started to hunt during idle, this fixed that particular issue, but this problem was evident before hand. Have reset TPS but still no joy.  

Any ideas please

Thanks

[SM25T]

Looked at plugs .... gap ? Is it EU2 with distributor and HT leads ? .... check for corrosion or looseness in all HT leads into cap.

[Geoff Martin]

No Ian, EU3 no distributor. Haven't checked plugs though. Seems obvious and Ive totally overlooked that. Maybe a solution without getting too deep. 

 

 

[revilla]

When it starts, do you need to hold the throttle open for a couple of seconds before it will idle on its own?

[john milner]

Do a search on IACV. You probably either need to clean it or replace it.

 

 

[revilla]

The reason I asked my earlier question is that this sounds like a problem I have been involved in investigating in some detail. In a nutshell it seems as though the performance of the electrical system on an EU3 K Series is marginal during cranking the starter motor. There is a large transient spike in current drawn by the starter motor for a small fraction of a second and this causes a momentary drop in supply voltage as seen at the ECU, immobiliser etc. It seems as though when everything is standard and healthy, the whole system just about "gets away with it", but only by a small margin. Anything which degrades the supply voltage in any way may be enough to cause the electronics to misbehave during cranking. Whilst investigating this on a friend's car I cam across many many instances of this happening; symptoms were most commonly one of the following:

ECU Affected - The ECU thinks the ignition has been turned off when the supply voltage drops and this triggers the normal behaviour which occurs when you switch the ignition off; the IACV stepper motor is wound fully closed and then opened by a predetermined number of steps to the correct position for the next start. Symptoms: For a few seconds after operating the starter motor the IACV is not sufficiently open to allow the engine to idle without throttle; the engine needs a little throttle to keep it going just until the IACV recycle has completed (this sounds like your problem).

Immobiliser Affected - The immobiliser has been seen to re-arm itself or to behave as though it were armed when the starter motor is operated. In some cases this actually results in the immobiliser light coming back on again, in some cases the immobiliser just seems to get confused and fail to synchronise with the ECU. Symptoms: The engine dies a couple of seconds after starting.

The three main causes which we found were:

i) Brise Starter Motors - I was able to show that initial current inrush and therefore the transient voltage drop was larger with a Brise starter motor than with the standard Magneton. This does seem to be a common factor in a lot of cases. We had one Brise motor which, when tried on three different engines, caused exactly the symptoms you report on each, where each starter correctly using a standard starter motor. Note that the starter motor cranked the engine strongly, the engine was not failing to start due to the performance of the starter motor but due to the transient voltage spike it was producing affecting the ECU as described above.

ii) Low or Weak Batteries - I had symptoms identical to yours after leaving a Sat Nav turned on which partially discharged the battery. Again, speaking to people these symptoms have been seen numerous times with low or weak batteries. Once again the starter motor was still able to spin the engine over rapidly but the reduced performance of the battery meant that there was a lower margin of safety against the transient voltage drop which again tripped the electronics.

iii) Poor Wiring or FIA Switches - Voltage lost when high starter currents pass through poor or under-spec wiring again reduces the margin of safety. Resistance in the contacts of an FIA cutoff switch can lead to significant voltage drops. The presence of an FIA switch also introduces two other relevant effect. Firstly, the battery wiring is typically a lot longer and so thicker cables may be required. Secondly, in order to allow the FIA switch to disconnect ALL electrics, the take-off point for the engine and chassis wiring looms is typically at the starter or somewhere else downstream of the all of the extra wiring and switch, meaning that any voltage drop in the high current cables is seen directly by the ECU (which is not the case where these are taken directly off the battery).

The threads below are relevant. The first one is my detailed investigation of the problem on a friend's car with pretty much identical symptoms to yours.

http://www.lotus7.club/forum/techtalk/brise-starter-current-draw-transients
http://www.lotus7.club/forum/k-series-start-needs-throttle-brise-issue
http://www.lotus7.club/forum/techtalk/k-series-starts-then-dies-immediately
http://www.lotus7.club/forum/techtalk/trouble-starting

Regards,

Andrew

[Geoff Martin]

1: John, IACV has been renewed.

2: Andrew, your first question is a good one, yes I do have to make it idle on the throttle for a couple of secs before it will cope on its own.

3: Battery has been discharged before, maybe a couple of times. Subsequently topped up, recharged and now on a conditioner. Do I trust it? - probably not.                                                                                          I had a problem with my previous EU2 K series when the battery died. Replaced it, and a few little gremlins went away.  

4: I dont have any FIA switches, so a thorough look at wiring and earths is in order.

5: Immobiliser seems to work fine. Not sure what starter motor I have, suspect its standard, not a Brise item. I shall check. 

Thank you for the responses.

[CharlesElliott]

I suggest you measure the battery voltage before and during cranking and see what it says. A decent multimeter (or better a memory scope!) will be required. This should help determine whether it is a voltage drop problem or not. 

[Jonathan Kay]

That's what I'd usually suggest... but was this the bug that needed an oscilloscope to spot the transient dip?

Jonathan

[revilla]

Yes it took an oscilloscope to really understand the problem as the deep part of the transient only lasted of the order of 50 milliseconds, however now we know what happens it is clear that the "healthy" electrical system is actually quite marginal and any drop in the baseline voltage during cranking is likely to mean that the normal level of transient may drop the voltage low enough to trigger the problem. So a good check of the battery and electrical system health through the standard measurements you mention with a multimeter is still very much in order in the first instance.

[Jonathan Kay]

Suggested threshold using a multimeter, Andrew?

Would doing the same starting test while jumped to another car running at, say, 3,000 rpm be another useful approach?

Jonathan

[revilla]

All of the oscilloscope traces I show in the thread in the first link were taken directly from the engine loom (I actually unplugged the lambda sensor and read the voltage from Pin 4 of the connector, as this is connected directly to the supply pins of the ECU within the loom). These therefore take account of voltage drops through the ignition switch, loom, fuses etc. which were not insignificant. Voltages shown in my traces will therefore be lower than read directly at the battery and it's difficult to know what would be a sensible lower limit.

What I do know is that, on an otherwise electrically healthy car, the transient voltage drop caused by a Brise starter (to 7.2V) was about 2V larger than the transient voltage drop caused by a Magneton starter, and this was enough to trigger the problem every time.

What voltage do you normally see when asking people to take readings during cranking? I have just been and taken some measurements off my test engine now. These numbers are therefore off a new-ish battery that has been kept on a conditioner with a brand new loom and a healthy Magenton starter:

Before Cranking, Voltage at Battery 12.35V, Voltage in Loom 12.35V
During Cranking, Voltage at Battery 9.95V, Voltage in Loom 9.60V

These numbers seem in pretty good agreement with the oscilloscope traces I recorded before.

(12.35V at rest does suggest that the battery is only 75% charged, however I suspect this is more likely to be an inaccuracy in my multimeter - however it does agree with what I found on the car with the oscilloscope on a different battery and I have no reason to suspect either of them so not quite sure what to say about it!)

Roughly speaking I would guess there's an absolute maximum of 2V margin for error with these (as the Brise starter with a 2V larger voltage drop always triggered the problem, although modeling it properly would be a good deal more complicated than that), although I suspect the figure is quite a bit lower. If the battery voltage is getting down to 9V whilst cranking I would be worried. Taking measurements directly from the loom using the technique I described would give a better indication of the voltage that the ECU is seeing during cranking and would show up if the voltage was dropping due to corroded joints in the loom, aged fuses etc. etc. which would not show up as a drop at the battery.

If the problem is a weak battery then jump starting with a GOOD set of jump leads from a GOOD battery may well get it starting properly; if it does, it's a good indication that the battery is the problem (however if you jump it from a running vehicle you will raise the voltage enough to mean that it could fix the problem even if it is voltage drops in the loom as you will be raising the voltage at the battery by a good volt at least if it is charging strongly).

[Jonathan Kay]

What I do know is that, on an otherwise electrically healthy car, the transient voltage drop caused by a Brise starter (to 7.2V) was about 2V larger than the transient voltage drop caused by a Brise starter, and this was enough to trigger the problem every time.

Thanks. 

But is there a negative missing? (Must be a pun somewhere there... )

Jonathan

[revilla]

The second "Brise" should have read "Magneton" ... thanks for pointing out that it made no sense! I have fixed it now.

As an aside, I think I have discovered how to get paragraph breaks on Android! When I tried to edit my previous post on the phone, each paragraph was displayed with < p > (without the spaces) before it and < / p > (without the spaces) after it - and when I posted back I left them in the paragraph breaks were preserved. If you see two paragraphs in tbis posted, created from scratch on Android, it works. If not, it doesn't!

[revilla]

It works! (If a bit geeky and tedious).

[Geoff Martin]

Revilla you were bang on with the voltage drop when cranking.

Battery measured 12.6v and when engine was cranked over voltage dropped down to 8v.

I have now changed battery to a Odessey PC680, the problem is now resolved. The fuel pump now sounds like is supercharged and it cranks over much better. (Needed to change starter cable and battery earth cable lengths as the terminals are at the back of the battery)  

All in all very pleased.

Many thanks to all that gave advice.

Geoff

[Jonathan Kay]

Glad it's fixed...

... and who is going to explain why it started with a smidgeon of throttle?

Jonathan

[revilla]

@Geoff: Glad to hear you are all sorted  

@Jonathan: I know you always recommend basic voltage checks before and during cranking and here is one more case that shows the benefits of the standard basic measurements. On a "modern" car the engine electronics can become upset a long way before the traditional symptoms of a flat battery appear; in this case the engine was still cranking nicely.

So the IACV is actually a tiny little throttle controlled by the ECU. It is in parallel with the main throttle butterfly, allowing air into the intake when the main throttle is fully closed. It is small enough that when the main throttle is opened it becomes irrelevant but when the main throttle is closed it gives the ECU fine long term control over the idle speed (in the short term the ECU controls the idle speed by rapid changes in ignition timing but trims it long term using the IACV). The amount of IACV opening required will depend on engine temperature, air temperature, loads from air conditioning (in a Rover!) etc. When the main throttle is in the closed position, the butterfly valve is not in fact completely closed; it lets a little air in still and provides nearly enough air for the engine to idle, and the IACV then provides the last bit needed.

When the voltage dropped to 8V during cranking, the ECU reacted as though the ignition had been turned off. It reset the IACV to a "best guess" position ready for the next start. The IACV is a stepper motor rather than a servo so the ECU has no way of measuring it's absolute position, it can just tell it "one more step open" or "one more step closed". In order to get to the correct position for the next start it therefore winds it closed by a number of steps equal to its full travel (which guarantees that it is stuck up against the fully closed position) then winds it open again a fixed number of steps.

If the main throttle is closed and the IACV then fully closes, then engine will stall as there is not enough air admitted to allow it to idle. Once the IACV reset procedure has completed though, everything returns to normal and the engine will idle happily. In the meantime it takes a bit of throttle to admit air past the closed IACV to keep the engine running.

 

[Jonathan Kay]

Thanks. Excellent. I keep thinking I've followed that, but then it goes again, which is weirdly analogous. More coffee needed.

Jonathan