Vvt Engine Swap Into Mk1 Complete...but Won't Start :(
 

OK I've posted this on the UK MX5Nutz forum but thought I'd also post it here as you guys have done more VVT swaps.:-

Ok so the VVT engine is in, all the wiring is done, (correctly I think), a few settings changed on my Megasquirt from its previous 1.6 configuration, but the engine won't start... ):

Basically it tries to turnover, I have fuel and spark on all cylinders, (defiately spark) but the timing is way out. I get occasional bangs out the exhaust or bangs through the intake if I have the throttle open...!!!

Here is what I have done so far with the install:-

I am using the VVT Crank and Cam sensors which have been wired into the original MK1 Cam sensor wiring like so. http://www.miataturb...ing_1-6_vvt-jpg However the wiring on my VVT sensors was Black/Blue, (Violet White or Grey/Blue for signal), and Black, (Not White/red like the diagram). I therefore wired the black wires as +12V positive to the wire/red on the 1.6 harness.

The original trigger plate on the crank pully was damaged, (one of the nubs on the plate was bent), so I replaced the crank pulley complete with a trigger plate which I took from a MK2 engine.

I am using the VVT coils which I have wired into the 1.6 harness by gutting the original ignitor and soldering the VVT wires to the pins like this:- http://www.miataturb...p_schematic-jpg Except I have only earthed the coils to the head and not the old ignitor as well and there is also no tach output from the VVT coils, (am using the ECU output to rev counter). I am definitely getting a spark from all 4.

I have re-wired my injector harness from batch to sequential. Pretty sure I have done this correctly and I have fuel going into the engine hence the odd bang when cranking.



I have made the following changes to my Megasquirt ecu from its original MK1 1.6 settings:-

Re-calibrated Throttle position sensor
Changed required fuel to 11.6.
Changed injection to sequential
Imported a new VE table base map from Reverant for a VVT engine.

All other ECU settings are as per 1.6 map, (AFR table, ignition table, cranking settings etc).

I have re-set the Trigger angle offset back to 0 from 2.6 that was used on the 1.6 engine.

Am I messing anything obvious?
Could the crank sensor have been knocked slightly out of position when the engine was removed/palleted/installed and if so how can I check?



Thanks

Chris

I have no clue about the wiring (since I never touched an NA and definitely not a 1.6), but have you verified that the timing plate is installed correctly (and not mirrored, if it was loose at any point)?
Reset the Crank Sensor gap to the thickness of a Credit Card ?

A composite log from TunerStudio?
This is what it should look like (but not overlapped of course)
https://www.miataturbo.net/attachmen...ine=1318954233

what does YOUR composite log look like?

what input circuits are you using?

Did not think to take a log :bang:

What do you mean by input circuits? (sorry for being thick)

Bloody car is at my father in laws, 1.5hr drive away. Will try and get back there as soon as possible. Will check gap on Crank sensor and try cranking the engine whilst recording a log this time and post the results.

what circuit is the cam sensor going into the MS CPU through?

what circuit is the crank sensor going into the MS CPU through?

Hi Braineack

I've spliced the 01 cam sensor and crank sensor into the old 4 wire 1.6 harness CAM sensor connector. Shared +12v and Neg for both sensors, then wired the cam/crank signals into the corresponding pins on the old connector. I've tried swapping the pins round at the connector in case I got the +/- wrong or the cam/crank pins the wrong way round but when I do I then get no spark or pop/bang so think I have it right. My MS2 is a Reverant built with his PNP harness that connects to the original ECU connector.

did you switch your tune for 99-05 sensors?

depending on what circuit reverant built your MS2 for, you might have issues with your cam sensor.

Where do I switch tune for sensors? I emailed Reverant to ask what changes I need to make to my ECU from my original MK1 1.6 settings but all I've been told to change so far from emails, or posts on forums etc is just what I listed above, (required fuel, sequential and a new VE table base map)? I sent my MS2 back to Reverant for him to add his VVT module to it and then when I got it back, refitted it to my mk1.6 and drove the car fine before I did the VVT engine swap.

ignition settings. you need to change for 4g63 to miata 99-05

OK, just checked my tune on my laptop which I have with me and my ignition settings are 4G63. Will change them to Miata 99/00. Unfortunaletly it will be a few days before I get back to my car to see if this make the difference.

Thanks again Braineack

Ok changed ignition settings to Miata 99/00 and the car runs :) !!!

Now have new problem. Injector No.2 is not firing. No signal?? Have swapped injectors to confirm its not a problem with the injector. Have checked and rechecked wiring. I had converted the injector harness to sequential as part of the engine swap. The wiring tests ok from pin 2 on my MS2 harness all the way to the no.2 injector connector. 12V also ok at No.2 injector connector. It's like the ECU is not sending the signal to injector No.2

EDIT- INJECTOR NO.2 IS FINE AND SEQUENTIAL IS WORKING - I HAD NOT CHANGED ALL SETTINGS ON MY MS2 AND THEN REBOOTED ECU

Chris

NiklasFalk - My log looks a little different. One of the crank trigger lines looks like a double line? Also I don't seem to have a full advance on my log?

Looking at my trigger wheel on my crank pulley, one of the nubs has white mark on it like its been knocked and the black paint scuffed off, (but it does not look bad enough to cause a problem?? ) checked gap at 1mm between all 4 nubs and sensor.

Can someone interpret my composite log for me to see if my ECU is receiving the correct crank signal?


Thanks

Chris

The full advance in that picture is added to show the location of the signals when then VVT is in full advance, there is always four crank signals between the cam nubs.
I choose that picture because it show all at once.

Thanks Niklas

Can you just check my composite log from the above link in my post to make sure the signal looks ok, (I think it is). Just so I know it's not a trigger wheel/crank sensor issue. In which case maybe my injector no. 2 not firing is down to the ECU driver for that injector as its the first time I've wired in no.2 as sequential, as previously I was just running from 1 and 3 for batch. If so I'll just re-wire back to batch and run the engine like that.

EDIT- INJECTOR NO.2 IS FINE AND SEQUENTIAL IS WORKING - I HAD NOT CHANGED ALL SETTINGS ON MY MS2 AND THEN REBOOTED ECU

The composite log looks to be with the cam nearly fully advanced. and it appears to be in sync.

I'm not experienced enough to see if the trigger disk is installed backwards using the log though.
A picture of the crank trigger area at TDC would answer that (one nub should just have passed the sensor then IIRC).

Cheers Niklas

Either way it looks like my crank sensor is not the cause of my injector 2 not firing.

Does anyone know how i can test that my MS2 ecu is sending a signal for the injector?

Can I run batch injection on my 01 VVT engine and if so what are the drawbacks if any?

Thanks

Chris

SEQ PW 4

It appears i'm an idiot as i did not select "additional drivers" on the sequential injection settings in tunerstudio. Hence injector 2 not firing. Injector no. 4 which is firing must be getting a short from somewhere else? Will go and try the above later tonight.

Thanks again guys for the help so far :) hopefully by the end of tonight i'll have s fully working VVT Mk1 Miata!

OK "additional drivers" made no difference, Injector 2 will not fire! Have check harness and circuit again only injectors 1,3,4 fire. So after hours trying to find the problem and 2 x 3hr round trips to get the car, I've wired back to batch and turned off sequential and all 4 injectors now fire. So we are staying batch!

EDIT - I DID NOT REBOOT ECU AFTER MAKING CHANGES WHICH IS WHY IT DID NOT WORK! NOW SORTED


Checked timing with a lamp, which was out by 10degrees, adjusted with the trigger offset, added a bit of fuel and it now idles smoothy at 14.0afr. Need to tidy up the wiring where I've re-wired back to batch, check for leaks etc and then start tuning the engine. Too late to make my trackday I'd booked this Friday but happy the car now runs well and the end is insight. :)

Chris

figure out the obvious fault in your wiring and go back to seq.

Thing is Braineack I can't find the fault! I've rewired to batch at the connector, (not redone the sequential harness), so there is was always a clear wire to injector 2, I tested and retested the circuit. For now I'm gonna stick to batch and get the car back on the road and out of my father-in-laws garage. Maybe in a few weeks or months I'll try sequential again. Just glad its all now running ok.

it could have been a bad INJ #4 driver/wiring inside the MS...is what I was getting at.

I think the problem is on the ECU side, but no idea how to really test/check it, (I don't fancy opening up my MS2 as I have no idea what I'm doing). Will stick with batch as I don't think I'm missing out on any power etc over sequential, just maybe a bit lower mpg? When I have more time in the future I may try sequential again, but for now I just need to get the car out of my father-in-laws garage and back on the road.

you're missing out on: power, mpg, idle quality, and overall smoothness.

In that case I will try sequential again, just not for a month or so. Just want to get back on the road now :skid:

Please tell me that you reset the power after changing the option.

Ok....... I'm not 100% sure on this. I did turn the ignition off fully on more than one occasion throughout the evening, but whether it was after the change and before I then went back to batch, I'm not sure? I still don't get why no.4 is firing as I can't find a short?

Next time I go back to work on the car I'll give the above one more go. I did not want to again remove the intake manifold etc to get the injector harness again so cut the wires at the harness plug and block connected back to batch, so I can easily re-connect to sequential as each iinjector still has a seperate signal wire attached. Once I decide to either leave it batch or if I get sequential working I'll again redo the wiring neater. My biggest issue is time. I need to get the car out of my Father in laws garage and don't really want to spend more time trying to get sequential to work.

So I got to work on the car today. For one more time I swapped the wiring back to sequential, changed MS settings and this time reset power and...... All 4 injectors working :inout: I did not know I needed to reboot the ecu after the changes and obviously did not the last time :facepalm: Still not sure why injector 4 was firing though before.

Anyway sequential appears to be working so now started tuning the VVT fuel base map from Reverant which I needed to add quite a bit of fuel to, more than I was expecting??

Now have VVT problem. Went for a test drive and above about 2.5krpm and approx 50% throttle the engine dies, rev counter drops, then restarts and repeats if I go above those settings. Pulled over and disconnected the VVT connector on the solenoid and car runs fine. It seems when the VVT tries to kick in the engine dies??? Its like the VVT tries to start but then I lose the crank signal yet they are not wired together? Is it just some kind of electrical fault and I need to check my wiring again!!!! My VVT solenoid is connected to my 12V on my injector harness and directly to the upgraded MS2 ecu. Should I choose a different +12v source for the VVT solenoid?

Are there any more settings in need to change on my MS2 or on the VVT module???

At the moment I'm wishing I'd stuck with my megasuirted 1.6 engine as it was running great before I decided to do the VVT swap.

Chris

What you describe falls in line with a complete ECU reset, like I had when the ground connection on the FIA switch was flawed.
The rev counter gets a constructed signal so it's not directly related to the crank signals.

The VVT OCV is working all the time so it's not idle below a certain Throttle or RPM (unless you have the VVTuner turned off).
I have however managed to burn a trace in my DIYPNP when shorting the OCV connector (by mistake). The VVTuner pulled enough power (fed through the DB15) to do so without damaging itself.
Since you a difference with the OCV disconnected this might be related (a ECU reset can be of many different reasons, low voltage being one, log of the repeatable incident?).

Hi Niklas

I did a log but realised I left it running for ages by mistake so will create a new log file later today. I can cause the event just by revving the engine. If I rev above say 3.5krpm I get what feels like an electrical cut, rev counter dies for a second, (which gets is signal from the ecu). There is also a clicking noise from behind the dash when it happens?

Tunerstudio log file: https://app.box.com/s/zna2rhg900atubwkty1x

With the VVT plugged in if I press the throttle say >30% car dies but starts as soon as I lift off. VVT unplugged, car is fine???

Log file shows me revving the car whilst parked.

What's "Lost sync reason: 31" ?

It's not an ECU reset at least.

Megasquirt-3 Sync loss and remedies
"31 Miata 99-00 - 2 cams not seen"
Assuming MS2 use the same codes as MS3... :dunno:

I have no idea Niklas. VVT soleniod upplugged car runs fine, plug it in and go above about 30% throttle = Lost sync 31??

I would triple check anything related to TPS, OCV and Cam sensor (wires connectors, solder, traces, ...).
Isolated the Signal Ground from the ECU ground (should only meet at one place in the ECU)?

v3.0 with expander?

The cam and crank triggers are going out of sync, at idle the VVT is pretty static - fully retarded - it's only as the OP accelerates the VVT advances and pushes the signals closer together until they cross.

I bet if you took a trigger log it'd show this.

I wonder if the crank trigger wheel is on backwards, or the cam pulley has been incorrectly fitted?

Hi Mr Slow :D

Composite ignition log https://app.box.com/s/hnlfxqs6qn655ahnmg5a

The original crankpully/trigger plate was damaged on this engine so I swapped it for one off the MK2 engine I had laying around. I don't remember removing the trigger plate from the pulley itself, (swapped them as a complete unit), for it to be on backwards, but the MK2 engine it came from was not using the crank sensor (it was using a CAS on the exhaust cam as it had previously been fitted to a MK1 car, so it could have been the wrong way round already and no one noticed?

How can I tell if the trigger plate is on backwards? The engine seems fine with the VVT solenoid unplugged. But when I first got it running and checked the base timing on the VVT engine I had to increase the ignition offset by 9degress using the Trigger wizard to get the timing marks to line up correctly.

Thanks again guys for your help.

A composite log with the OCV disconnected could be compared to the fully retarded one. If it differs something is mechanically different than stock.

If the cam is one tooth advanced (15 degrees) the VVT can only reach +12 - +56 (if stock zero is -3 and 44 degree range)
If one tooth retarded the range would be -18 - +26.

Since the CAM sensor reads the cam and not the wheel it should not move it out of sync if you don't ask it to do so. Ask the VVT to reach 0 and try again?
What VVT angle is reported with the OCV disconnected?

Hmm 9 degree offset to get one of the marks to align? Are you sure you have looked at the right one?
If you advance the trigger wheel 25 degrees it will be symmetrical.
https://www.miataturbo.net/attachmen...ine=1317773390

But if the sync error comes with Throttle angle and not RPM I would look into electronic gremlins.


Edit: looking a bit deeper into the last log the Lost Sync 31 come after the VVT Target goes above 25 degrees. The VVT Angle and VVT Duty are both zero in the log so i have no idea what's actually happening.

This is a much newer FW than what I'm using, but the pattern is visible in Excel.

Just clocked this.

With the MK2 crank and cam sensors it's unlikely your timing would be out by this much, there's little play in the sensor positioning.

Are you 100% sure the timing mark lit up by the strobe was the correct mark?

Are you 100% sure the timing mark actually marries with TDC (check with the dipstick in Cyl no 1).

Try setting the trigger angle back where it was and see if you loose sync again.

Unless the damper (which have the marks) have slipped some, which would fool anyone with a timing light.
The trigger wheel does not follow the damper and zero offset should be OK, and if you need precision you should measure on a dyno anyway (so the absolute reference is not that interesting).

Hi Guys

Just checked again with my timing light. With the marks lined up as per the below picture the angle offset required was 10.5 degrees with the megasquirt set to fixed timing of 10degrees. I set it back to "use table" and checked again it now lines up at 14degrees as per my ignition table.


https://www.miataturbo.net/attachmen...ine=1376929916

I have not checked however if the marks actually line up with TDC by physically put something in cylinder no.1 to see.

Here is a composite ignition log with the OCV unplugged: https://app.box.com/s/0nwgwh7bi97bjfm2pa2j

Show us the tooth logger. I had a similar problem, it was a bad sensor out of the Mazda box.

One tooth log for you Hustler https://app.box.com/s/o4i8acivnm5a5o6pk15e

:)

On a side note my VVT engine seems to run a bit warm? Seems to sit at 95/96 decrees C when driving/cruising, which means my radiator fan is also running, (it's set to come on at 95C on my Megasquirt). My 1.6 used to stay around 87-89C. I'm using the 01 temp sensors, (not my original MK1's) which I have spliced into the MK1 harness. Will the 01 VVT sensors give different readings to my old MK1 1.6 temp sensors? They will be a PITA to swap now, (I would have used them but did not see the removable grub screw on the back of the VVT engine for the temp gauge sensor until after I had fitted the engine, so I thought I needed to use the 3 pin 01 sensor instead).

I'll try setting the trigger offset back to zero, plug in my OCV and see what happens.

Quick update - followed Richy's suggestion and checked TDC with my dipstick in cylinder no.1. At the highest point the left mark on the pulley lines up with the T on the cover. Also at the sametime the nubs on the trigger wheel are in the correct place according to the picture Niklas posted. So the crank pulley/markings and trigger plate are correct but to get them to line up with the engine running means adjusting the angle offset by 10.5degrees!

What should I try next? This is all new to me and I don't really know what I should be doing, was hoping the engine swap would just happen without any issues?

I'm using DIYauto tunes VVT base map. Is it worth me modifying it by reducing the numbers in each cell by 10.5 degrees to match the offset I'm using?

Can you communicate with the VVTuner and read the duty cycle values from the software?

Hi Dimitris

I'm actually away from the car for a few days so can't check till after the weekend. I have been able to connect fine with the VVTtuner though through the new serial connector, so will check the OCV duty cycle when I get back.

Thanks

Chris

Duty cycle, warm idle. Limits are Min 25% max 80%. Should I have anything selected for Trigger Output? Its set to none (off)< I guess because on Reverants ECU it gets its reading direct from the ECU, not direct from the sensors.

https://www.miataturbo.net/attachmen...ine=1377512119

Can you try going to 2500rpm in neutral and watch the duty cycle when the problem occurs? Report said duty cycle.

Ok duty cycle is in the low 30's. the problem happens at either 3000rpm+ or above 30kpa.Duty cycle does not really change when problem happens. It's kind of like a signal crash/clash. Engine splutters/dies for a split second, Rev counter bounces down and back up, clicking noise from behind the dash, like I just lose electrics or something. OCV unplugged car runs absolutely fine?

Try to change the advance taget in the table to be max 24 degrees (or just zero, or 5,..), just to test if the log was telling anything (where it lost sync when the VVT target got above 25 in the msl). Start in the lower left corner and tab your way though the table (if you don't want to do off-line editing).

Just a question though, why was TS putting a VVT Target in the msl?
In my experience the VVTuner is very separate (and can log what it's doing as well).

If you can control the VVT angle to let's say 5 degrees all over revs and MAP, there should not be an electronic problem or software problem, but rather an mechanical problem with the location of the cam/crank nubs.
The ving at the back end of the VVT is just shrunk in place, no key at all. But if that's seriously wrong you would notice a strange Cam Advance in VVTuner with the OCV disconnected (-3 is normal IIRC).

Sorry if some of this have been mentioned/asked/answered before, disregard as needed :)

NiklasFalk - Any suggestions are gratefully received whether they have be mentioned before or not. :)


Set all my VVT table values to 5 and plugged back in the OCV and tried the car. Car works fine at all revs / map. I now need to increase the values until the problem occurs.

With the OCV disconnected the cam advance in VVT Tuner at idle is 3 (not -3), it's also 3 with the OCV plugged in? Seems to make little difference plugged in or not according to VVTuner?

Thanks again

Chris

Logs from the VVTuner when you do those tests might reveal something when it fails.

With an advance of 3 as zero I see no way that the adjustment range would explain the loss of sync. :dunno:

Or you can just run the poor thing with an advance below the level it craps out?

VVT Log https://app.box.com/s/n6bi124qk7okx857i536 With the target set at 17 degrees car is fine however at about 20/21 degrees I get the problem/signal crash. Even with the table set to a max of 17 I still get the odd crash when the VVT tuner sometimes overshoots the target, (pulled away from some traffic lights and cam angle jumped to 21 and problem occurred (electrical/spark cut? revs drop/bounce, car dies). Is there some way to reduce the occasional overshoot?

Any suggestions of things for me to try are gratefully received.

For now I can live with a an advance of 17, its better than nothing! I need to get my car tuned and ready for its trip to the Nurburgring and Spa in October.

Row 2580-2628 does not look healthy (15890rpm for one).
Could it be that the VVTuner is creating this problem and shorts the cam sensor when resetting or something (and MS is reporting lost sensor)?
It looks to be two seconds of messy data before a reset (or whatever happens).

Or is the MS the culprit?
https://www.miataturbo.net/megasquir...31-32-a-65472/

This is where I am stuck, I have no idea if its my Rev MS2 with built in VVT Tuner or a dodgy sensor. I've emailed Reverant to see if he has any other suggestions I can try to find the cause?

Ok Quick Update. Above 20degrees I get the problem/signal crash? So I have set the max target on my VVT table to 19 and need to play with my PID settings to prevent the overshoot I'm getting which can trigger the problem. Not ideal as I'd like to be able to try the settings recommended by DIYautotune , but at least a max of 19 advance has to be better than nothing?

What is the best way to tune PID? Which do I tune first? P, I or D?

Another thing that may be affecting my VVT is my oil? When I fitted the engine I used Millers Competition 10w-40 fully synth which is what I was using in my NA1.6 and seemed great on track. It is however very heavy for a 10w-40 and I've noticed my oil pressure gauge now seems very high. It seems I should be using a lighter grade for the VVT engine, so will try a high quality 5W-30, (Fuchs Titan Race Pro S instead). Car is used on track so I want something decent and not too low viscosity.

I'm now using the following P20, I10, D8 and a maximum target advance of 18degrees on my VVT table which is based on the original DIYautotune table but with 18 being the highest number. My PID probably needs tweaking further but at least I am no longer getting overshoot above 20+ degrees and the subsequent signal overlap/crash whatever it is???

Still not sure wether to just give up for now and pull the plug on the OCV and run without VVT :( I have a track day /road trip coming up so need to just get the car tuned and running well but running out of time and have no ideas on how to solve the VVT issue. Would it have made any difference if I had upgraded to MS3 rather than have a VVT Tuner wired into my MS2E case?