[FIXED] CSR / Duratec 2.3: Intermittent engine 'cough' under load

[vertew]

Chris - yes tks ... I'll be checking out the bad tooth count tomorrow (now I've worked out how to do it!).

Jonathan - it's certainly a strange one.  Current draw would be higher under max load with ancillaries working hard, but why near a particular rev range.  Chris could be on to something with the ECU data stream check, perhaps ECU is browning out.  It could even be something awful like a resonant frequency near 4,200 RPM which is vibrating a dodgy component somewhere, agh.  On the battery - the cranking was done after battery had been on the conditioner.  But battery shoiuldn't be doing anything when car is on the move, so not sure how this could lead to a negative voltage spike ...?

cheers.

[aerobod - near CYYC]

From a logging perspective, I would use automatic logging to collect all common parameters, then have a look at all logged parameters when the "cough" occurs. I think the voltage drop could be the cause or effect of the problem. If the ECU reset counter increments when the cough occurs, the ECU has effectively rebooted for a fraction of a second which can cause the cough, the voltage supply, grounding, ECU issue or disturbance from another source such as an intermittent short of a component could cause the voltage drop that leads to the reset.

If there is no ECU reset, the engine speed spikes may be caused by incorrect sensor input under certain conditions, for example the lambda sensor or TPS sensor - this should be evident on the values for those sensors. A voltage spike could cause rapid change in the injector voltage compensation value, causing the cough. I think the correlation of the cough "spikes" across the different logged parameters will tell a lot (all electrical parameters echoing the same spikes at exactly the same time would point to an electrical problem, spikes out of sync from a time perspective would point more to a non-electrical issue).

[vertew]

@aerobod - yep, I did collect the standard set of parameters, then removed some data lines from the pdf plot to simplify the view.  See updated pdf for a new 'page 5' that shows more data parameters.  'Ignition Advance', 'Duty Cycle', 'Fuel Injection Time' and 'TPS fuel + Trim' do all show glitches at the same time as RPM wobble (though some are small due to axis scaling).

@ChrisC - I tracked 'bad missing tooth count' and 'reset count' using easimap live view.  Once the engine had started, the values remained static throughout the 30min test drive (when misfires did occur).  So looks like there are no data corruption / reboot issues.  The values don't appear to be recorded in log data for some reason.

Next steps ...

I've uploaded the raw easimap data here - could I ask easimap experts to have a look.  If you spot anything not yet discussed, please let me know ... many thanks.

I've swapped out all four coil packs and the MFU box ... which made sod all difference, so at least they're ruled out.

I've now got a replacement alternator on order.  Decided to go for a new one rather than risk a bench test / repair not addressing a possible weird dynamic load issue.  I'll recondition the original and keep it as a spare.  Hopefully this will improve operating voltages and get rid of the dim / flickering ign light, then I'll go from there.  Battery remains a suspect.

[ChrisC]

Beats me, but I would want to know what the logged faults mean.   I think thats going to be a call to SBD, unless you can find them on their website (which I can't). 

It would be where I would look next. 

They start at 44s into the run, and are logged at 2m32s, and another fault at 3m12s. 

[aerobod - near CYYC]

I think it is a bad TPS or wiring to the TPS that is the problem. If you look at 9m21 the TPS voltage goes from zero throttle (4.62V) to about half throttle (2.33V) and back again to zero throttle in less than 100 milli-seconds. There are also other spiky changes in the TPS voltage in other areas, such as here:

Perhaps these are extremely rapid throttle blips, but it seems to have the characteristics of a noisy TPS input.

 

 

 

 

[revilla]

Those throttle blips do correspond to sudden ramp increases in RPM on the blue line though which suggests they may be genuine.

[aerobod - near CYYC]

It is difficult to tell without a wide band lambda reading, as the TPS voltage change would cause a throttle map change and consequent fueling change, which could cause a rev change.

[John Vine]

I think it is a bad TPS or wiring to the TPS that is the problem

That's interesting. It would be worth checking whether the TPS wiring is standard or the uprated race loom. (I'm assuming that this can be fitted to 2.3L engines as well as the 2.0L.)   IME, Duratecs have a nasty habit of fracturing TPS (and COP) wiring, probably as a result of high-frequency vibration.

JV

[vertew]

Hi and thanks again for your responses which have inspired me to take a much closer look at the easimap fault data (and this takes a bit of deciphering to say the least).  I've described this in a bit more detail below, but in summary:

• there do appear to be MAP and Oxygen (lambda) sensor errors that are definitely worth exploring.
• there is a battery fault flagged, but this only occurs during cranking.
• I'll keep the TPS in mind, but this does not seem to be giving an error - perhaps I'll focus on the erroring sensors first.

I'll add checking this lot out to my to-do list.  Any thoughts on this or detail below, please do let me know, esp if you spot anything wrong.

If nothing else, I'm getting a crash course in 7 engine diagnostics ...
cheers, Andrew.

More details ...

(1) Descriptive faults parameter - cycles though current (transient) faults

The descriptive fault parameter is giving faults:

• 'Manifold Pressure 1 Sensor' - worth checking out
• 'Throttle 2 Sensor' - think this is not fitted / 'normal'
• 'Unknown' x 2? - ????
• 'Baro Pressure Sensor' - think this is not fitted / 'normal'

(2) Current (transient) Faults A and B (ref spreadsheet)

These two parameters appear to be logged as decimal values.  To decode, convert to binary (16 bit), then the faults can be looked up from parameter fault list in easimap.  I created a spreadsheet here to decode mine, others may find this useful - just enter the decimal code to see the faults indicated, but note it's just a first attempt and may not be correct.

(Also - I'm not sure if I've for the list of faults for 'A' right - there are only 14 faults listed for a 16 bit code.  Are rhe two final bits unused?  But if so, easimap should say 'unused' in parameter info. Perhaps a call to SBD at some point.)

Taking the fault look-up list at face value for now, my decoded faults were:

Fault A list:

• 'Battery' - only happens while cranking.  Might be normal if trip is at 11V
• 'Steering A' - no idea what this is. Anyone?
• 'Oxygen2' and 'Oxygen 1' - Don't think 'Oxygen2' is used.  'Oxygen 1' could well be the lambda sensor
• 'Unknown' - there are the two missing bits that don't have a fault translation and one has been flagged?

Fault B list:

• 'Oil temp' and 'Oil pressure' - both flags seem to be triggered at ignition on / start, then go off for normal operation.
• 'Mapping knob' 1,2,3 - seem to be triggered at ignition on - probably normal
• Baro - is always on.  Think I recall reading a post from aerobod that baro sensor is not used on a 7?

(3) Logged faults A and B (ref spreadsheet)

These follow the same approach as current faults A and B, but each fault (bit) is persistent and (I presume) gets reset after a power cycle.

[vertew]

A few more things tried this week and the issues remain - yay.

I've changed the alternator, the MAP/T sensor and cleaned the lambda sensor.  I went out for a test drive each time ... long enough to find that the problem is not fixed.  A few more things I've noticed - see here for easimap engine parameters pic.

• I was concerned about a lambda (oxygen) ECU fault code.  This code clears after a few mins (when exhaust oxygen levels reache reach nominal).  So lambda doesn't seem to be an issue.  I also did a road test with lambda unplugged and got the same misfire.
• Manifold Temp looks fine.
• Manifold Pressure is frozen at what looks like a default value.  Is the CSR 260 not supposed to measure manifold pressure?  I tried easimap on a Duractec R400 (thanks Stuart - Herts club) and that is measuring manifold pressure.  Could be a wild goose chase, but looks odd.  Any ideas?

Looking back at previous suggestions ...

• aerobod - thanks v much for spotting the TPS spikes.  The TPS trace looks generally fairly smooth - even when the 'coughs' do happen - the spikes could be my throttle blips on down-changes?  Definitrely worth investigating though.  I'll try a test not doing any throttle blips and see is spikes are still there.
• John Vine - the CSR has the flxible TPS 'race' loom aleady fitted as OEM (not to say it might be faulty).

Coming back to the basics ...

On near/full acceleration only: (a) there are occasional engine coughs, (b) the ignition light flickers dimly and © easimap graphs show a substantial negative spike often (but not always) aligned or near the engine cough.

Could be wrong, but the problem still seems electrical to me.  Other than points above, my next steps will probably be a closer look at the ball of wool that is the loom, starting with things that could cause ignition light flicker - would be great to fix this, even if it turns out to be unrelated.  I also need to find out whether the CSR should be monitoring manifold pressure.

An occasional misfire is not the end of the world and I could just live with it til it fails properly ... but it is annoying especially in a car that's all about the driving experience.  I did get a great run out last weekend to meet up with the Herts club and the CSR was fantastic on the way (except for a few engine coughs).

Many thanks for continued interest and as usual - any more ideas very welcome.

tks, Andrew.

[Jonathan Kay]

Could be wrong, but the problem still seems electrical to me.

starting with things that could cause ignition light flicker

I've changed the alternator

Have you had a chance to measure those voltages again:

• At rest
• Minimum during cranking
• At 1,000 rpm increments?

I suspect there isn't a single fault, but it would be good to know which bits are solid.

Jonathan

[John Vine]

Is the CSR 260 not supposed to measure manifold pressure?

It won't be able to measure Manifold Absolute Pressure (that is, the partial vacuum in the manifold) because it's fitted with roller barrels and there's no manifold.

Did the R400D you tried have the standard plenum and throttle butterfly?  If so, it will have a MAP sensor to measure the vacuum.  But with RBs, that sensor is moved to the air-filter back box, where it becomes a combined air temperature and barometric pressure sensor.  At least, that's how my R400D with RBs is set up, and I imagine the CSR260 is the same.  When I run Easimap on mine, the baro pressure remains a constant 1.040 (and flagged red, suggesting an "error" condition).

I'm following this saga with great interest as I'm very keen to hear what the problem turns out to be!

JV

[Geoff Brown]

My R400D had the same problem for a while more so when traveling at night with lights on. With light throttle load OK but pushing on caused a miss fire/cough which was alleviated by easing off the loud pedal/changing down a gear & decelerating for a while. Cruising on steady throttle was also affected at times.

Cure: changed the battery !  It appears that when the engine was under a lot of throttle load & even more with the lights on the ECU was being starved of its minimum voltage to operate properly due to a poor performing battery.

Ever since the battery change a year ago, no problems.

The alternator was checked & pushing out 13-13.5 V at tick over.

 

[Jonathan Kay]

Very interesting, Geoff.

I'm suspicious of something like that (see post#9 and #14), but I'd like to know what takes the extra current intermittently/ under load. You didn't find anything else faulty... ?

Jonathan

[jimh]

Have you considered taking the car to the Two Steves and getting them to remap it on their rolling road? I would have thought they would also be able to cure your issues.

I had my CSR260 done by them and it made quite a difference to the drivability of the car.

[Geoff Brown]

JK.

Nothing else was faulty. At the time new spark plugs fitted, TPS in fine fettle also Lambda Sensor.

All other plugs & sensors checked & the only thing I found was light corrosion on all of the pins for the individual coils which I carefully removed.

Obviously the engine when crated was somewhere cold & damp or maybe in a submarine ?!

Battery change was the easiest go to fix after I tested it........yet another carp Banner & only after three years in use.

[Jonathan Kay]

Thanks, Geoff: that was going to be interesting with either answer!

Jonathan

[vertew]

Re. Jonathan's point on voltages and Geoff on battery issues ...

I've not had chance to measure voltages (via easimap) again after the new alternator.  I'll do that at the weekend.  

As I understand it in most cars: all the battery should do is start the car.  Once running, the alternator takes care of power.  If the battery is helping, there's a problem.  The battery warning light comes on when the battery has a higher voltage than the alternator supply and current is flowing from the battery.

But - Geoff's problems were fixed by a new battery and I do have a flickering battery warning light as well as a 'battery' ECU fault code on cranking.  I did buy a new Banner battery soon after I bought the car in Sep 2015.  The car had the issue then and it didn't go away though.  Anyway - may as well settle it - so new battery arriving shortly!  

Another option I'm thinking about is that my flickering battery light could be accurate - total power requirements under a stress condition (full acceleration) are more than the (new) 40A alternator can supply.  This would mean that something wants too much power.  Faulty fuel pump? Easimap data doesn't particularly point at a fuel issue.  Voltage doesn't look like it drops off with a 'cough' and pressure looks stable ...

Re. John Vine's point on MAP

I think you've nailed the 'manifold pressure' question.  Stuart's R400D did have the full manifold plenum so MAP was 'supposed' to be measuring a partial vacuum.  As you say, mine has throttle bodies, so the MAP is 'supposed' to measure atmospheric pressure.  The 'baro' sensor is flagged red because it's not fitted and not needed - the airbox MAP does it.  So can probably cross that off the list.

Re. Jimh on remap

My CSR has had the '2 steves' remap - difference was night and day and 99.9% of the time it's smooth as silk - thanks Steve Greenald!  This isn't a map problem, it's a nasty little fault.

Please - let it not be a loom issue ...........................

 

[sjmmarsh]

I too am interested in the outcome. I have had intermittent misfires in the past, but these have always cured themselves after a while and some tinkering with the wiring. Check there is no chafing of the coil pack wires under the carbon cam cover, particularly around the hole opposite the heater. 

The thing that caught my attention was the cough occurring at the same RPM. Since I got my csr 10 years ago, I have had a very intermittent problem under hard acceleration - it feels like the engine hits the limiter when the rpm is around 4000. I have learned that the fix is to momentarily back off the throttle and reapply it - the car then drives through the limiter with no problem. 9 times out of 10 it occurs when I am overtaking on the road and so my priority is to get the power back rather than diagnose it. It only happens a handful of times over the year and I haven't successfully recreated the problem when I want to demonstrate it. 

My suspicion is that one of the electrical connections is being afffected by a resonant frequency. 

Steve

[Jonathan Kay]

Two questions about yours, please, Steve:

1. Any suggestions that the charging system is otherwise marginal?
2. Or that the system voltage is dropping when the misfire occurs (as with vertew's voltage measurements or flickering warning light)?

Thanks

Jonathan

[vertew]

Hi Steve - great to have your CSR input.  That doesn't sound a million away from the issue I've got.  It does feel a bit like a rev limiter (as power is reduced) and is often (not always) around 4,000 - 4,200 RPM.  It's not fully like a rev limiter in that I can keep the throttle pinned and after a second or so, it'll clear.  If a 'bad one' there might be two or (once) three of these hiccups that might last 1-3s as engine speed goes from say 3,000 to 6,000 RPM.

This has been the case for the 3yrs I've had the car - the reason I'm chasing it down now is that it's been getting slowly worse.  It's got to the point now where it's reasaonbly repeatable, so at least I can use diagnostics tools to catch it happening.

I've had the resonance thought too which would point to loom / connectors / chafing.  I visually checked the where the coil wiring exits the carbon cover a few days ago, looked ok, but a rubber grommet was out of position which could well be a chafe spot.  I'll take another look though.

As Jonathan suggests - could I ask you + Jimh (and any other CSR owners who might be reading) can you check if you get the flickering igntion light on full power?  It's very dim / not easy to spot, so you'll need to be out a night.  Try full power acceleration from 3,000 to 6,000 RPM and see if you notice the ignition 'spark' light on the speedo coming on.

I've bought a cheap-ish USB ossciloscope which will work with my laptop - I'll see if I can detect any flatlining signals in key wiring when the 'cough' happens ...

tks, Andrew.

[elie boone]

Maybe too far fetched, but if it's gradually worse, can it be your alternator multi V belt slipping due to vibration ? and this could make the alt. brushes to float ?

[John Vine]

I've bought a cheap-ish USB ossciloscope

Could you provide a link, by any chance?  I'm looking for something similar myself.

JV

[vertew]

Hi John, I'm no expert here, but I've ordered one of the picotech 2000 series scopes here.  I went for the 2205A (25MHz), but the 2204A (10MHz) might do for the low speed world of auto-electrics.  For info, take a look at this dedicated vehicle diagnostics scope and guide here.  Seriously expensive, but very cool.

I may cancel my order though as there is some news - see next post ...

[vertew]

I got a PM from Mic who suggested that the issue was probably the immobiliser and to un-mount it and leave it hanging down, then try the car.

I did that and (as well as fixing a really dodgy immobiliser earth) and the 'cough' seems to be gone!  I then fixed a more permanent 'vibration damped' mounting and a more secure earth, went for a longer run and ... no 'cough'.

For the record, I captured a few pics and an explanation of what I did here.

So it seems that some of the hunches were correct - it did come down to wiring issues and vibration / resonance.  4,000RPM probably caused a resonant frequency in the bulkhead area that either disrupted the immobiliser internally, or the poor earth connection.  I've not yet been out at night, so yet to check whether the flickering ignition light has also gone. Also, I can't say I fully understand exactly what was going on, as I'm not sure what the immobiliser is doing.  But given that it's able to immobilise the car, any issues with this are probably going to be noticeable.

I'm fairly sure I've got a CC standard immobiliser (the CSRs were factory-build), but generally, the standard of immobiliser fitting is often not up to the standard of other electrics.  Others might want to check their immobiliser installations.

There's a small possibility that the problem might still materialise again, but I think I can mark this thread fixed.  I can hardly believe it myself.

MASSIVE thanks to Mic for taking the trouble to post.  I don't know how long it would have taken me to work that out by myself, possibly never.  This has made a huge difference to the driveability and my enjoyment of the car.

Huge thanks to all of you who offered advice.  I've learned a load more about my 7, including the MBE CAN interface and Easimap.  Just goes to show how well the blatchat community works and that even problems can be an interesting part of the whole experience (ok I, might not have said that earlier).

I can now get back to more mundane things like fixing a leaky rad and actually driving the thing.

Hoping I don't have to make a new post that starts 'you know that misfire' ...
Andrew.