Avon CR500 tyres
20/7/2004 
Saturday July 31st was the Bristol Pegasus Motor Club, Castle Combe Track Day.  This was my first outing on a circuit with my aeroscreen and new sticky tyres (Avon CR500’s) & the car was transformed – now I’m asking why didn’t I make the change earlier?   The grip from the CR500's was amazing, towards the end of the day I felt my neck aching from the cornering!

This is a *very* long story, but to recap some of the basics are:  
* The car is a 2.0L Zetec engined Westfield on Jenvey TBs 18 months and 1500 miles old (when the problem occurred) built with all new parts from Westfield.
* The ECU is an MBE 956E with software version 956AA260.

ECU case image: 
The problem first happened in February 2004.  The symptoms were that the car died after about 15-20 minutes running.  After about 30 minutes or so it would re-start and the entire cycle would repeat.  After checking for loose wires, water in the fuel, broken connections, I contacted Westfield several times who each time suggested replacing new items ... the coil, plugs and ECU sensors.  So for the next 3 months of investigations and many many calls to Westfield I replaced the coil, plugs, leads, fuel filter & all the ECU sensors but the problem still persisted.  I had regularly contacted the Westfield help line but with little joy. I really felt left high and dry.   Through my own investigations  and with tremendous support from fellow WSCC members I'd got as far as knowing it was overheating IC's inside the ECU which resulted in the misfire & the engine cutting out because they were scalding hot and showed scorch marks inside the unit.  The unit would also come back to life for another 15 minutes or so when I doused it with electronics cooler spray.  The problem was this: although it was overheating IC's resulting in the shut down, what was the cause of them overheating?  It could be faulty IC's themselves - but two failing together - unlikely?  It is more likely to be a fault elsewhere causing too high a current to be drawn from the IC's thus resulting in them overheating. It had been suggested to me that maybe there was EMI on the crank position sensor to ECU lead.  I pared back a little of the lead and earthed the foil sheathing with crocodile clips & wire to the chassis.   No improvement.  I'd also discovered that the overheating IC's were the coil-on amplifiers and were made by SGS-Thompson part no VB027.  I got detailed spec sheets on these IC's and discovered that they had a thermal overload protection so when they exceeded their rated temperature range they would shut down.  With a thermal probe I found that the temperature on mine after 20 minutes (or when the engine died) was way over the rated max so it was correctly shutting down.  But why were the IC's overheating?  Through various conversations with WSCC folk and electronics people I  had discovered that too high a coil-on time (also known as dwell) would result in too high a current being drawn from the IC's thus stressing them.  I checked the coil-on time with EasiMap on my ECU and this was fine - at the recommended 3ms.  So what now??  I was getting nowhere fast. I mailed Westfield the details of my findings & spoke to them again explaining it was probably an ECU issue but they seemed totally uninterested in this electronics problem & batted me to and fro with MBE.  MBE initially told me they don't sell directly to the end user only via dealers, and as Westfield were MBE's customer they would deal with them.  Westfield wouldn't facilitate the communication with MBE and eventually said they couldn't help me with the problem.  Despair.  I finally obtained help from a WSCC electronics and ECU wizz Adrian Clinton-Watkins (and my brother in law) both of whom know loads about micro-electronics.  

Here's the findings .....we traced it all through with an oscilloscope, digital multi-meter and EasiMap and found the following.

ECU internals image:
Firstly the two IC's to the left (the suspect ones) control two coil halves (IC8 and IC9) the fuelling is controlled by IC10.
The Crank Position Sensor (CPS) input to the unit at idle is clean and sensible and measuring the frequency tied up with the RPM. Checking the input and output of the power drivers showed good signals coming into the units and appropriate switching transforms coming out of the units.  No EMI was found on any loom wire. 
However we noticed that the two coil power drivers were getting VERY hot at less than 950 rpm. The Tick over is 900 rpm at present. We discovered that if you were touching the units as a measure of heat that quickly at 900 RPM they became far too hot to touch, whereas as soon as you revved the engine to around 1200 rpm and above they quickly cooled back down to a reasonable (but still very warm) temperature.  So using the scope we measured the switching period in ms on the TTL side of the power driver. We could see that at 900 rpm the coil on time was in excess of 12ms whereas quickly above 1100 RPM this would drop to typically 6/7 ms. in addition above around 1500 rpm the reducing coil on time would be steady at 3ms (i.e. the value reported by EasiMap in the chipfile).
So it appeared that the dwell angle / time was incorrect in the ECU or at least far too large at low RPM to avoid over driving the coil and creating excess heat within the power driver. This overheating was the cause of the thermal shutdown of the unit.
However we checked the dwell time (fixed) and the dwell time map and made changes to these with both Easimap 5 and Easimap 4.14d and in both cases this only actually effected the dwell time as recorded on the scope above 1500 rpm. Still the dwell below 1500 rpm seemed static to 12ms at tick over.
EasiMap trace:  
We also noticed that the EasiMap reported bad injector duty cycle only occurred at lower RPM and suspect this is probably processor errors / firmware issue pertaining to low RPM or possible too high current drain from the high coil on time. Checking the duty cycle at this time showed correct injector firing on the output side of the power driver for the injectors.
As another check we increased the tick over to above 1000 - 1100 rpm and the unit ran happily for a long time with it still getting quite warm, but not too hot to touch, and now not failing.

So our conclusion is that there appears to be something odd (firmware) that is causing the unit to use a large coil on time at lower RPM. I suspect that this has always been within this ECU but that in fitting a second throttle return spring has only recently set the tickover low enough to trigger the problem.

At this point we contacted SBD and sent our findings to them.  SBD Motorsport were incredibly helpful.   As a customer of theirs they bent over backwards to help me.  I explained to them that we had "masked" the problem by setting a high tickover and mounting the ECU in a heat sink.  I'd explained that I wanted to find out the following:

Can the ECU be repaired?
Should I just have the power drivers replaced?
Can the board be replaced? 

SBD came back within a day and explained to me that there was a "known" problem with the 956E firmware which was related to coil on times.  They said that the 956E ECU's had been superseded about 3 or so years ago.  Also, a year or so before that the final release of the 956E board was software version 463.  The one on mine supplied to me in mid 2001 was software version 260.  Furthermore they explained that the 260 version was replaced several years before I was supplied mine brand new from Westfield.  With this in mind SBD very kindly re-contacted MBE who very courteously offered to replace the main board with a new version having the final release 463 firmware.  Within a week the replacement MBE unit was back and everything worked perfectly.

So during this time I'd spent almost 4 months off the road, missed a couple of track days that I'd been booked onto and spent a couple of hundred pounds on parts that I didn't need.  Also I felt let down in that having spent almost £18,000 with Westfield over the build that I would have expected a greater level of interest and service.  SBD, with whom I'd spent a few hundred pounds with were superb.  MBE finally were very helpful and apologetic for the run around I'd had.  To say I was unimpressed with Westfield's care and attitude is an understatement.  

24/10/2002
I'd e-mailed Mark Walker at Westfield to ask about the twitchy handling.  He called me back & said that maybe as I was running the whole thing in as recommended with the dampers set on zero clicks things may improve when there was some damping set in.  He recommended starting with 6 clicks on the rear dampers and 5 on the fronts.  Over the weekend I changed the settings as recommended and re-checked the tyre pressures.  However with the bad weather & family commitments I didn't get the opportunity to drive the car to test things out.

30/09/2002
Replies from my WSCC posting on the twitchy handling:
1.    I was still getting this with mine even after the suspension set-up, so last week I whipped the pads out, smeared the back with a bit of copper slip (the back of the pads were worn where the pistons were pressing against them), and popped them back in.
Don't know why, but this seems to have totally cured the brake unevenness, and they are now much less prone to locking up at the front early!    (They're the 4 pot "upgrade" brakes.)

30/09/2002
Update after almost 150 sunny miles over the weekend.
Having assumed that paying for the suspension to be set-up by Westfield that everything would be ok I hadn't checked damper setting and tyre pressures before Thursday & Fridays evening outings.  So on Saturday morning I got round to checking both.
Tyre pressures (cold) were wildly out.  Front L 35psi R 32psi.  Rear L 25psi R 32psi.  I set them to 20psi all round.
Dampers were a bit out too  Rears were both zero clicks, front L 2 clicks, R 4 clicks.  Set the fronts to 4 clicks and left the rears at zero.
Front and rear ARBs were on their softest settings so I was happy with that & left them alone.
Next outing of 30 ish miles felt much improved.  At higher speeds - say above 60 snapping off the throttle still induced a little twitch left, but nothing like before.  Heavy power on gave a similar small twitch right.  As I'm trying to "gently" run the engine & everything in for 500 - 1000 miles I don't intend to keep testing that too often.
The brakes were starting to bed in nicely also.  Front right still locks up just before left but it's not too worrying.  Back discs hardly get warm & haven’t even cleaned up yet.
Got back home & re-checked tyres, then re-set front dampers to two clicks & tried another 30 miler again.  Pretty much as before - slight left twitch on power off, right twitch on big power on.  Also slight misfire in 2nd gear at 4000RPM.
Finally - as the manual says for bedding in, I set the fronts to zero clicks, left the rears at zero clicks & resolved to not fiddle until it's run in.
Now it's got about 200 miles on the clock it's feeling more stable all the time.  

27/09/2002
Yesterday I got over the last hurdle & got my new registration number and road tax.  Right back from the DVLA I had the plates made up at 4:30pm and went out for a late afternoon drive clocking up 70 ish miles.   Went out in 20 mile or so stints with my sons on some reasonably quiet B roads.  I did notice the car felt a bit twitchy especially when backing off the throttle. Admitted everything is brand new & so is bedding in - is this normal?
It twitches slightly left when backing sharply off the throttle in a dead straight line - noticed it above 60mph - not really noticeable below.  So far I've only checked for the brakes binding, if anything the front right locks just before the left so I don't think it's that.
The suspension set up was done by the Westfield factory so I've assumed it's ok? Any thoughts?
The set up was done by Westfield before I got it SVA'd and on the road.  I took it to them for a final once over.  For sure nothing will have settled down yet so maybe I'll just watch it for 500 - 1000 miles & let everything bed in then check it all again then.

Dunnell Exhaust Manifold & silencer
14/9/2003
Over the weekend I fitted the new manifold.  The old Westfield one came off considerably easier than it went one.  I trial fitted the silencer to the end of the down pipes but it wouldn't go straight on.  With lots of filing at the end of the manifold together with heating the silencer inlet with a blow torch I finally got the silencer to fit over the new system. I took it all off again & trial fitted the manifold to the head.  It went on very easily, no fiddling around trying to get it through the hole in the fibre glass.  I marked up the position for the lambda boss & took the lot to my local garage to be welded on - this cost a mere £10.  Fitting the exhaust and silencer were then dead easy and the lambda sensor lined up perfectly with the original hole in the bodywork (picture 2nd from right).  I finally made up an exhaust mounting to hold the back of the silencer in place and that was it!

10/9/2003
Dunnell manifold arrives
Dunnell Zetec manifold arrived at last.  I spoke to Paul Dunnell the week before and agreed that they would not weld in a lambda sensor boss as it would be too awkward to determine exactly where it should fit. They supplied me with a lambda boss which I need to get welded in the correct place.   When I fit the manifold I should offer it up to the head, mark the position where I want the boss fitted & take it somewhere to be welded up.

15/5/2003
Dunnell silencer arrives
The new Dunnell silencer arrived.  As the new silencer inlet pipe has a bore of 2.5 inches & the Westfield manifold outlet is just under 2.0 inches I knew I'd have to bodge an adapter.  At the weekend I went to Merlin Motor Sport at Castle Combe & bought an stepped sleeve and a couple of exhaust clamps.  With much hacking of metal on the sleeve, and liberal amounts of exhaust cement I managed to get the two to join - without leaks - although it's most definitely not a permanent job!. 

Jan 2003
I've decided on the Dunnell exhaust manifold and silencer.  The silencer is a stainless steel, repackable one with a 2.5 inch bore inlet.  The manifold and down pipes are larger bore than the standard Westfield ones & more importantly they have been tuned for the Zetec so don't sap any power.  My order is in, and just as soon as the metal fabricators have started supplying Dunnell with the new systems I'll have one.
Haven't done much driving in January of February, the roads aren't dry enough.  Went to the WSCC Bristol & Bath area meet on February 20th (in my Audi).  Lots of events have been arranged for the summer months :)

12/11/2002
I took another days leave off and went up to Mech repairs to have the new injectors fitted and the ECU mapped.  They flow tested the new and old injectors and found the new ones gave an increase of  60% in flow rate.  They also ultrasonically cleaned the new ones to ensure they flowed at the same rate.  Fitting them and re-mapping the ECU took about 4  hours.  The power comes on brilliantly - with the Piper cam it runs strongly up to 5000 RPM and then just goes wild from 5000 to 7500 - with the soft cut out at 7600 RPM.  It reminds me of a motorbike with a sharp power band.  Driving home I looked for quiet bits of road just to have a quick blast.  Great fun!.

Reply from the WSCC regarding ECU mapping:
2.    Try Steve Richards Website here. You will find a rolling road register. I am surprised that the engine is set up lean, as I would imagine this would be bad for a brand new engine. Westfield set mine up so rich that it was getting strangled at high revs. I took it to pro-tune (North London) and they leaned it up and got much more BHP and torque out of it. Mine is a factory built 1.8 Zetec so they didn't have to do the SVA on it. I would get it re-mapped sharpish once you pass the emissions

My response:
I think that could be happening to mine too - won't rev above 5000.  I phoned Westfield to ask about this and confirm the warranty thing.  They said it must get re-mapped - warranty would still be ok.  It has the 1.8 Zetec map but the car is 2.0 Zetec + hot cam.  It's cutting out at 5000 RPM - they reckon it's fuelling so a re-map is essential.  Looking at the rolling roads on Steve Richards list, then   Mech Repairs in Cheltenham seem to be the nearest MBE outfit.
Has anybody any experience of Mech Repairs?

Perspex Wind Deflectors
12/10/2002

07/10/2002
Perspex wind deflectors arrived by Parcel Force from John Moore.  Instructions look very straight forward, so that's this weekends job lined up.