LocostBuilders - powered by XMB

R1 Locost misfiring

Sanzomat - 17/4/21 at 08:57 PM

Hi All,
Been reading a few posts on here for a year or so. Back in Sept I bought a Locost R1 that had been laid up for over 5 years. I bought it sight unseen during isolation but at a price that I figured gave me headroom to sort out the inevitable issues.

I've done a lot of work on the suspension (separate story!) but my issue now that I've tried running it is the thing is misfiring dreadfully.

It is a 4XV lump and it has been converted at some point to run 5PW throttle bodies/injection. The spacing is different so the previous owner has altered the spacing of the TBs to suit the 4XV inlet ports and it seems the size of the inlet stubs are also smaller on the 4XV so some eccentric ali spacer rings have been added to each inlet stub with the 5PW rubber connector then installed over these. Seems a bit of a dogs dinner but I guess it must have worked at some point.

Before I start messing with the megasquirt mapping (again, assuming it worked before it was laid up) I'm ticking off the things that are likely to have suffered during storage.

Symptoms - Starts okay but idle isn't great and runs rich and fouls the plugs after idling for a while.
If you rev it it fluffs/hunts/stutters but will pick up and rev. With no load, in neutral it'll rev all the way but AFR is going really rich and popping back and making some black smoke at times. With load it won't go past 9000rpm and leans out. It pulls kind of okay from around 7k to 9k though. Rich below 7k and lean at 9k, too lean to go any further, like hitting a rev limiter.

I've cleaned the injectors and they seem to be firing okay and flowing about right.
Fuel pump has good pressure. Haven't tested the pressure regulator though. New fuel.
I've changed the spark plugs and coil packs (had an R1 bike in the past, 5PW, and it was very sensitive to old plugs), no real improvement.

I'm currently suspecting air leaks from where the 5PW TBs have been mated to the 4XV inlet stubs so my next task will be re-sealing these (looks like plenty of clear silicone type gasket goo was used to seal them previously and it has broken down a bit)

Do any of you guys have experience of 5PW TB's mated to 4XV and can it be made to work? Is this a common thing or a bad idea in the first place??

Any other thoughts?
Cheers, Dave

adithorp - 17/4/21 at 10:01 PM

Not heard of anyone converting them. Sounds like you've covered everything I'd have normally suggested and I'd be looking at the tb to stub seals.

CosKev3 - 18/4/21 at 09:11 AM

Are they deffo 5PW ITBs?
I thought the heads were the same as the carbed R1.
5PW ones are the strange ITBs that still look like a carb with that float thing on them?

sdh2903 - 18/4/21 at 09:19 AM

We're the coil packs new or used that you swapped in? The 4xv I had was rough running on the 2 sets of coil packs that came with the engine. New ones cured it.

Sanzomat - 18/4/21 at 09:50 AM

quote:
Originally posted by CosKev3
Are they deffo 5PW ITBs?
I thought the heads were the same as the carbed R1.
5PW ones are the strange ITBs that still look like a carb with that float thing on them?

Definitely 5PW TBs. I've looked on a few bike R1 forums that confirm there is a different spacing and different size to the inlet stubs on the head between 4XV and 5PW. I've got the original carbs in a box of bits that came with the car, a few bits have been robbed off the carbs though otherwise I would have tried swapping them back on and seeing how it runs on carbs. The diameters of the carb outlets where they mate to the head inlet stubs are definitely smaller than the 5PW TBs. I've also got a full 5PW TB assembly straight from a bike so with the spacing fixed by the fuel rail etc and the 5PW ones on the car have clearly been altered as the fuel rail is in two parts and is blue anodised versus the plain ali of the original one piece fuel rail. The actual TBs are the same though, as you say with the same vacuum sliders like carbs but no float chambers/jets.

Sanzomat - 18/4/21 at 09:53 AM

quote:
Originally posted by sdh2903
We're the coil packs new or used that you swapped in? The 4xv I had was rough running on the 2 sets of coil packs that came with the engine. New ones cured it.

The replacement coil packs are new ones. Made no difference at all.

perksy - 18/4/21 at 10:00 PM

TPS been checked?

Checked for air leaks with carb cleaner or similar?

[Edited on 18/4/21 by perksy]

Sanzomat - 19/4/21 at 07:09 AM

quote:
Originally posted by perksy
TPS been checked?

Checked for air leaks with carb cleaner or similar?

[Edited on 18/4/21 by perksy]

I plugged the laptop into the megasquirt and the TPS dial moves smoothly from zero to 100 as the pedal is pressed down so seems fine.

Spraying carb cleaner around the inlets during idle didn't seem to change anything but the idle was pulsing a bit anyway.

I'm going to give everything a good dose of sealant next and try that.

ragindave - 19/4/21 at 07:53 AM

Are they balanced correctly worth checking.

blue2cv - 19/4/21 at 05:48 PM

Surely if its making black smoke its running rich

Sanzomat - 19/4/21 at 08:18 PM

quote:
Originally posted by blue2cv
Surely if its making black smoke its running rich

Yep - its rich up to around 7k rpm, well, the AFR is all over the place but generally rich. There seems to be a sweetish spot between 7 and 8k where the AFR is around 14 and then above 8k it leans out so that by 9k it the AFR is around 18 and it won't rev any higher. It has a wide band lambda with a digital AFR read out so I can see this (assuming it is reading right as I expect the lambda has gotten quite sooty)

My theory (possibly half baked) of what is happening is air is leaking in at the inlet joints. The Megasquirt is seeing the effect of this from the lambda and is increasing the fuel (assuming it can do this??) but if the air coming in is inconsistent then the fuelling adjustments can't keep up in real time so the fuelling is never right. At around 7k the air leaking in is matching what the injectors can provide at full flow so it comes on song. As the revs rise further and more air is sucked in, the injectors are already at full flow so can't give any more fuel so it goes lean.

What my feeble brain can't work out is how the amount of air sucked in through a leak can be anywhere as much as would come in through wide open throttles so then that theory breaks down. Or is it the location of the rush of incoming air, further down the port from where the injectors squirt, that is messing up the high end???

Anyway, my brain is hurting!

Sanzomat - 8/5/21 at 12:21 PM

Thanks for all the responses. A bit of an update. I resealed the throttle body to head joints and no noticeable difference. I ran a log through a warm up cycle and that has given me some really useful information. The coolant temperature on the log graph was showing at -18F and other than a few odd spikes it stayed there throughout. The coolant temp gauge on the dash went up as expected and the fan came on at the right temperature but the sensor the ECU sees stayed at-18F. As such the warm up enrichment stayed at max. Removing all warm up enrichment made it run better straight away but the lambda AFR readings are still all over the place so I'm guessing all that rich running has contaminated the O2 sensor, maybe permanently.

I'm now hunting around trying to find the temp sensor. I can only find one, near the thermostat housing, and that is driving the gauge and doesn't seem to be connected to the Megasquirt. I think I'll have to locate the pin-out and follow it from the box to see where it goes!

Anyway. definitely some progress.

[Edited on 8/5/21 by Sanzomat]

adithorp - 8/5/21 at 06:07 PM

The O2 sensor won't have appreciated that. The problem with that is the carbon crud build up on it "insulates" from the gas and it and it see's weak irrespective of the true mixture and will continue to richen the fueling.
If you take it out and get a blow lamp (or even gas cooker flame) on it and burn it off it can overcome that. Be careful not to get it too hot; cherry red at most.

The R1 only has one temp sensor (near the stat housing) so if your megasquirt is seeing a different one it must by a modification)addition that's been done. The other possibility is the MS is spliced into the standard one and theres bad connection somewhere in the loom causing a high reading.

[Edited on 8/5/21 by adithorp]

Sanzomat - 27/6/21 at 11:49 AM

Coming back to this as still not sorted. Update on where I've got to:
Having added a second coolant temperature sensor and connected that direct to the megasquirt coolant temperature pin, after a bit of calibration the CLT as seen by the MS2 was pretty close to what the dash gauge was showing throughout a warm up cycle so I reinstated the warm up enrichment. Then was starting fine (other than still having a few odd spikes/misses) and warming up fine and the no-load AFR's seemed okay. With no load it would rev all the way but the AFR readings went to very lean at higher revs.

No way of testing it under load as no tax/MOT/insurance (track day car only)

I thought at this point I'd solved it but the fuel mapping still looked all wrong and I assumed this was why the AFR's went lean at higher revs.

Thought about getting it road legal and building a fuel map over a few drives/logs/adjustments. Decided a rolling road session might work out cheaper than tax, insurance and MOT so booked a session.

On the rollers the ignition spikes/drop outs were much more obvious under a bit of load and were preventing any useful fuel mapping. The guy suggested just trying to get the top end (open loop area) mapped but giving it some beans showed up that the clutch was slipping (see separate thread).

I sourced and installed a second diaphragm spring in the clutch as advised by the kind and knowledgeable folks on here and this seems to have sorted the clutch. I traced and checked every wired in the ignition system as well as checking the resistance in all the sensors and couldn't find anything wrong.

I took the car back to the rolling road hoping to at least get a top-end map sorted but the ignition spikes/drop out persisted and the engine wouldn't run under load above 7700rpm at all. It put out 110bhp at 7700 though which is at least a bit reassuring as comparing that to other 4XV/5PW dyno graphs that is quite healthy.

Whilst going through all the wires I did find that the MS2 had four wires coming out of what would normally be spare pins on a fuel only set up and those wires had colours that perfectly matched the colours in a Yamaha 5PW ignition coil sub harness for the 4 coil trigger wires. These wires ran out to the plugs for a 5PW coil sub loom near the head left empty and tied back. I also found a 5PW coil sub harness buried in the box of bits that came with the car but no pencil coils.

A reminder - the car has a 4XV engine with converted 5PW injection throttle bodies. It came to me with the ignition running off the standard 4XV ignition module with two coils, each with two HT leads running off to the plug caps and the fuelling running through the MS2 set to fuel only.

This discovery got me thinking that maybe the previous owner was part way through a project to put the ignition through the megasquirt too. It seemed that all that was missing was the coil-on-plug pencil coils that the 5PW uses. I sourced a set and plugged in the sub harness and replaced the HT leads/caps with the pencil coils. The MS2 timing input was coming from on of the old coil primaries so a reconnected that direct to the crank sensor signal wire.

As the 4XV has no cam sensor (the 5PW must have as there is a connector on the coil sub harness right in the middle) the coil-on-plugs would have to run wasted spark as there would be no sensor telling the ECU which of the 2 revolutions that make up the 4 stroke cycle each cylinder was on. The dwell time for these coils is sufficiently short for this not to matter.

In tunerstudio ignition settings there was a drop down for YZF1000 in the trigger settings so I thought it should be fine with the standard trigger wheel (3 short, one long teeth). Setting it to YZF1000 populated a number of the other boxes so I guessed it knew what they should be. I set it to COP wasted spark though instead of straight coil on plug.

I tried the starter hoping it would fire into life and be ready for a final go on the rollers but now it won't start at all so something isn't right.

My thoughts are:
1) Can the MS2 read the trigger wheel? I stupidly forget to set a log going and didn't check for the revs being shown whilst cranking so this may be the problem.
2) Can the coils be fired straight from the MS2 or do they need external igniters? I need to check for a spark.
3) If the MS2 is getting a signal from the crank sensor and the coils are firing, is the timing right.

Unfortunately the car is in the yard of the rolling road and I only had a limited amount of time down there when I fitted the coils. I want to make my next visit count and hopefully get it back on the rollers before I bring the car back home so I want to be armed with as much as possible for my next trip down there so would be grateful for any suggestions!! sorry for the long and complicated post but thought it best to include as much info as possible.

I have a track day booked at Combe on 5th July - at the moment it is looking like it might be in the GTM!

CosKev3 - 27/6/21 at 09:17 PM

When I converted my 4xv to MS and injection I used a Ford wasted coilpack and fitted a 36-1 crank pick up in place of the standard Yamaha one.
Ran like a dream!

Sanzomat - 28/6/21 at 08:12 AM

quote:
Originally posted by CosKev3
When I converted my 4xv to MS and injection I used a Ford wasted coilpack and fitted a 36-1 crank pick up in place of the standard Yamaha one.
Ran like a dream!

Can you remember where you got your 36-1 wheel from? Was it a direct swap for the 4XV one or did it need fettling/fabrication? Did you keep the same pickup?
Which Ford coilpack did you use?

Might have to go down your route!

CosKev3 - 28/6/21 at 08:33 AM

quote:
Originally posted by Sanzomat

quote:
Originally posted by CosKev3
When I converted my 4xv to MS and injection I used a Ford wasted coilpack and fitted a 36-1 crank pick up in place of the standard Yamaha one.
Ran like a dream!

Can you remember where you got your 36-1 wheel from? Was it a direct swap for the 4XV one or did it need fettling/fabrication? Did you keep the same pickup?
Which Ford coilpack did you use?

Might have to go down your route!

Trigger wheels I think the website was called.
Had to file out the centre to match the R1 crank.

Coilpack is any wasted spark one off a ford,mondeo/focus etc.

Injection was set to batch fire.

Standard R1 CPS

CosKev3 - 28/6/21 at 08:36 AM

http://trigger-wheels.com/store/contents/en-uk/p7.html

Sanzomat - 28/6/21 at 11:24 AM

Thanks for that. It seems the standard wheel has on OD of 63mm and there was a 36-1 wheel on ebay with a 63mm OD for 15.99 delivered so I've ordered one. As you say, the centre hole will need altering and a keyhole shape formed.

Did you match the keyhole slot so the missing tooth was at the sensor when at the TDC marks lined up on the standard wheel? Start of gap or end of gap?

[Edited on 28/6/21 by Sanzomat]

Sanzomat - 28/6/21 at 08:57 PM

quote:
Originally posted by CosKev3

Coilpack is any wasted spark one off a ford,mondeo/focus etc.

Can I just check, these Ford wasted spark coilpacks, do they need to be fired via an EDIS module or are they controlled direct form the MS2? The coil packs have a 3 way connector - centre being +12v (from fuel pump relay) and the outer two are either direct from the MS2 or from an EDIS module?

If direct from MS2, does the MS2 need to be modded so that these are high current or do the coil packs contain their own "igniters" so the signal wire is just a logic current?

40inches - 28/6/21 at 09:56 PM

quote:
Originally posted by Sanzomat

quote:
Originally posted by CosKev3

Coilpack is any wasted spark one off a ford,mondeo/focus etc.

Can I just check, these Ford wasted spark coilpacks, do they need to be fired via an EDIS module or are they controlled direct form the MS2? The coil packs have a 3 way connector - centre being +12v (from fuel pump relay) and the outer two are either direct from the MS2 or from an EDIS module?

If direct from MS2, does the MS2 need to be modded so that these are high current or do the coil packs contain their own "igniters" so the signal wire is just a logic current?

The MS2 fires my Ford coil pack directly.

CosKev3 - 29/6/21 at 08:03 AM

quote:
Originally posted by Sanzomat
Thanks for that. It seems the standard wheel has on OD of 63mm and there was a 36-1 wheel on ebay with a 63mm OD for 15.99 delivered so I've ordered one. As you say, the centre hole will need altering and a keyhole shape formed.

Did you match the keyhole slot so the missing tooth was at the sensor when at the TDC marks lined up on the standard wheel? Start of gap or end of gap?

[Edited on 28/6/21 by Sanzomat]

Pretty sure I set it up like this;

EDIS-4: missing tooth is exactly nine teeth (90�) ahead of the VR sensor,

But was a few years ago so can't remember for sure,Google it plenty of info about

Sanzomat - 30/6/21 at 04:34 PM

Okay, I'm really going to show my ignorance now. I was assuming that changing over the ignition from the Yamaha module to the MS2 was just going to be a case of moving the timing trigger input from the input to the Yamaha module to the input of the MS2 and putting the coil primaries onto the MS2 outputs rather than the Yamaha module.

Whilst running fuel only the timing input to the MS2 was one of the coil primaries (fed from the Yamaha ignition module).

When I previously moved over the MS2 timing input from the coil primary to the crank sensor output I assumed that the lack of RPM being shown in tunerstudio whilst cranking was due to the Yamaha trigger wheel being incompatible.

I've now fitted a 36-1 wheel and still no RPM when cranking so the MS clearly can't just swap from a coil input to a VR input. I was thinking it would just need the software settings changed but it seems there are different circuits to be installed depending on which used.

This is now getting a bit beyond me...

Sanzomat - 24/7/21 at 01:00 PM

Updating this thread. A friend installed a VR decoder and wired in the ignition outs on the MS2 V2.2 board for me. Still didn't work and it seems the MS2 processor has now died, possibly on the operating table.

I now have a Speeduino ECU set up with VR decoder and all the right inputs and outputs. I've gone for a Polo coil pack with built in igniters. Wired it all in including stripping out the old Yamaha ignition module, set up the timing with the 36-1 wheel. It fired up first go and seems to be working nicely now. I can only test it with no load though as not currently road legal and I can't get a rolling road slot before the track day I have booked for 2nd Aug!

I've loaded in the fuel map from the previous RR session which seemed to be good up to 8k where the ignition died off and have extrapolated hopefully sensible figures above that. Have put in a basic ignition map.

I've appealed for any other maps to reference against in the EFI section of the forum.

Sanzomat - 18/8/21 at 07:09 PM

Back for an update. Thanks for all the input/suggestions by the way.

I abandoned the track day I had booked for 2nd August as the wideband controller also gave up the ghost so with an iffy map at best and no way of reading the AFR it was too risky to run the track day. I've now put in a new wideband controller and was thinking about MOTing/taxing/insuring it and trying to build a map on the road but in the end I decided the sensible thing would be to invest in rolling road time. Most of the rolling roads are booked out weeks ahead at the moment. Many of those that I rung won't touch Megasquirt/Speeduino or indeed kit cars in general, especially BEC so a bit limited in choice. A friend in my local kit car club suggested one of his contacts who does a lot of work with Scooby rally cars and books time on someone else's RR and they managed to book me a half day slot which was today. The rolling road was at Badger5 (a VAG tuning specialist) and is actually pretty local to me. Not the cheapest but you get what you pay for and their set up was pretty trick. After 4 hours on the rollers I've now got what seems to be a great map and a peak power of 147.5 which I'm happy with from a 4XV motor with 5PW throttle bodies. The standard injectors actually reached 100% flow a bit after the peak meaning it wouldn't go all the way to the red line but as this was well past peak power, other than top speed this isn't the end of the world - automatic engine protection!

Got a track day at Combe next Tuesday so really looking forward to it now.

Sanzomat - 1/9/21 at 09:03 AM

Castle Combe 24th August
Well, the Locost (mostly) worked well on the track day last week. Clutch cable popped off the linkage due to pedal backstop working loose and giving too much slack, gear change linkage went wobbly at the pivot, earth wire connection to fuel pump went bad, all of which easy to fix in the paddock and arguably what a shake down is for! 10 minutes before the flag dropped at the end of the day the left front wing stay broke but by then the car was going well and I was starting to get the hang of driving it. It is very different from the GTM Spyder! Definitely quicker. I took 3 seconds off my best ever lap in the GTM and for sure there is much more to come as my head was telling me to brake far too early based on the limits of the GTM!