[rusEfi] 2004 Mazda MX5 1.8 VVT #50

What engine are you torturing? Is it inside some car?
User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: 2004 Mazda MX5 1.8 VVT

Post by kb1gtt » Tue Apr 16, 2019 10:13 pm

How about started fluid to check spark. Once it starts on started fluid you then start with fuel.
Welcome to the friendlier side of internet crazy :)

User avatar
russian
Site Admin
Posts: 9779
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: 2004 Mazda MX5 1.8 VVT

Post by russian » Wed Apr 17, 2019 12:06 am

Crazy Striker wrote:
Tue Apr 16, 2019 7:38 pm
https://youtu.be/yva_zOCNAzU
RPM reads steady around 200-210 RPM during cranking
do not see your laptop. console would have muc more details. do not see engine sniffer showing coil control.
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Wed Apr 17, 2019 5:20 am

I will produce cleaner feedback today. May have a bottle left in da House for my carbureted Renault, I will try this too. Thanks

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Wed Apr 17, 2019 3:44 pm

I have tried with reverse sensors for trigger and the results are very encouraging. Car seemed to catch a bit but soon stopped. RPM sould have rise to 500-700 rpm during this phase.
I was unable to register a succesful attempt but i've made a screen of a cranking in engine sniffer :
2019-04-17_15_13_33rpm_0_maf_0.0.png
2019-04-17_15_13_33rpm_0_maf_0.0.png (68.84 KiB) Viewed 892 times
I still have pops coming off the exhaust.

Should I try to set offset with a timing light ? Isn't the offset and all preprogrammed in the NB2 preset ?

I've also made a video of tunerstudio while cranking :


My battery ran a bit low because of succesive cranking attempts. I'm charging it and will try again maybe in the evening.

Thanks

Gwendal

Abricos
contributor
contributor
Posts: 727
Joined: Mon Aug 18, 2014 12:32 am

Re: 2004 Mazda MX5 1.8 VVT

Post by Abricos » Wed Apr 17, 2019 3:48 pm

8 dwell I think it's tomuch and battery is deed ...
https://rusefi.com/docs/html/

is a general agreement to measure all angles in crank shaft angles. In a four stroke engine, a full cycle consists of two revolutions of the crank shaft, so all the angles are running between 0 and 720 degrees.

Ignition timing is a great example of a process which highlights the need of a hybrid approach to event scheduling. The most important part of controlling ignition is firing up the spark at the right moment - so, for this job we need 'angle-based' timing, for example we would need to fire up the spark at 700 degrees. Before we can fire up the spark at 700 degrees, we need to charge the ignition coil, for example this dwell time is 4ms - that means we need to turn on the coil at '4 ms before 700 degrees'. Let's assume that the engine is current at 600 RPM - that means 360 degrees would take 100ms so 4ms is 14.4 degrees at current RPM which means we need to start charting the coil at 685.6 degrees.
Last edited by Abricos on Wed Apr 17, 2019 3:53 pm, edited 1 time in total.

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Wed Apr 17, 2019 3:50 pm

Actual battery voltage was more close to 11.8 when mesured directly to it. I spent two days cranking on it, I think it's normal it is so low.
I've tried with 6.0ms and 8.0ms of dwell aslo but more or less the same result

Abricos
contributor
contributor
Posts: 727
Joined: Mon Aug 18, 2014 12:32 am

Re: 2004 Mazda MX5 1.8 VVT

Post by Abricos » Wed Apr 17, 2019 4:08 pm

I think for start engine you need fixed ignition angel ...
So if you change your dwell ignition angel is chenged to ...

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Wed Apr 17, 2019 4:12 pm

I tried with ignition angle of 10 and 15 degrees both with 6 and 8 ms of dwell, and with injector PW from 3 to 6, but no luck so far.
Is the 50% IAC PWM generic ? Car seemed more willing to start when blipping the throttle open

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: 2004 Mazda MX5 1.8 VVT

Post by kb1gtt » Wed Apr 17, 2019 11:41 pm

Does your timing light indicate the timing is correct? Aka could your offset be different than the expected for some reason?
Welcome to the friendlier side of internet crazy :)

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Thu Apr 18, 2019 11:35 am

Yes !!!!
It finally started. 2 things were needed, inverting primary input and setting VVT trigger to falling edge instead of rising edge, as it is not possible to invert VVT input in software.
The only issue I have is the car idling at 4500rpm, but at least it seems trigger is no longer a concern. I'm able to review up to 6000rpm but the car is running a bit rich.
Need to know why my IAC valve do not seem to regulate at all, it seems that it is stuck open when running.

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: 2004 Mazda MX5 1.8 VVT

Post by kb1gtt » Thu Apr 18, 2019 11:46 am

NASA knows video or it did not happen.
Welcome to the friendlier side of internet crazy :)

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Thu Apr 18, 2019 11:59 am

I'm working on the IAC circuit to verify if there are any shorts and then will film it 😉
Looks like the IAC is stuck open at 100%, which can enhance my starting issues

User avatar
russian
Site Admin
Posts: 9779
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: 2004 Mazda MX5 1.8 VVT

Post by russian » Thu Apr 18, 2019 12:06 pm

idle bench test button should click IAC on non running engine. fo you have flyback? is it mouted with virrect polarity?
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Thu Apr 18, 2019 12:08 pm

I have installed a flyback but I'm starting to wonder if it is correctly installed. Could not hear anything with the IAC test bench...
I tested the wires and they seem OK. Maybe the driver is dead ?
Attachments
IMG_20190418_140746.jpg
IMG_20190418_140746.jpg (3.99 MiB) Viewed 855 times

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Thu Apr 18, 2019 12:21 pm

Stupid me, I was powering the IAC through main relay output... No wonder why it isn't working. Will swap it in TS and see if it works

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Thu Apr 18, 2019 2:27 pm

I have a little question. On 2003+ miatas the ECU should be powered by the ignition +12 (4S).
I didn't connect the 4S and the ECU seem to be powered by 4AF at all time, like in the 2002 diagrams.
Is this because it is EU spec or something ?

Abricos
contributor
contributor
Posts: 727
Joined: Mon Aug 18, 2014 12:32 am

Re: 2004 Mazda MX5 1.8 VVT

Post by Abricos » Thu Apr 18, 2019 2:41 pm

Check RusEFI wiki maybe you see wiring diagram...
Or alldata or mitchell on demand ...
Last edited by Abricos on Thu Apr 18, 2019 2:44 pm, edited 1 time in total.

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Thu Apr 18, 2019 2:44 pm

I wondered about it because of the wiring diagrams I found on this wiki.
The 2003 and 2002 pages seemed reversed for me. I can try to check on my own car but what is sure is that 4S isn't connected and the car is running 🤔

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT

Post by Crazy Striker » Thu Apr 18, 2019 5:53 pm

Here is a video of it running on RUSEFI :D
https://youtu.be/O0SYoIAgDAk
Still have some issues with the decode at warm startup but once it catches it works well, except for the IAC. It still wide open so the video you see is with the IAC unplugged and me governing the engine
My Chinese MAP sensor just begin to have a weird behavior. It shows approximately half of the value it should show. It wasn't doing it before... But with appropriate values the engine is running but quite rich except on flooring where AFR are close to 14.7
I never saw my engine rev so happily 😯
When on stock MAF it seemed so sluggish :lol:
I no need to sort out this IAC issue...

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT #50

Post by Crazy Striker » Fri Apr 19, 2019 3:12 pm

So quick recap of what works and what not :

Working :
- ECU power
- Trigger
- Ignition
- Injectors
- MAP
- Wideband
- IAT
- CLT
- TPS
- VBatt
- Warm cranking
- Running engine (up to 6500 RPM)

Not working :
- Cold start
- IAC and Idle

Not tested yet :
-VVT

TODO :
Solve the IAC issue. I will unsolder and swap Main relay and IAC. I will check the flyback diodes.
Tidy up the harness especially for trigger signal as I still get no reading at first when hot until the ECU catch. For now the signal comes to the board via loose connectors.
Build a nice metal box for the ECU.

I keep you updated 😉

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT #50

Post by Crazy Striker » Mon Apr 22, 2019 4:09 pm

I solved the IAC problem today. It was just due to a component in the wrong oientation ( the MOSFET).
Now it starts (only wtested when warm) and it runs.
I've made my first rolling test but the tune is a bit too rich on idle (I mean way too rich, like >10 AFR), therefore it it perfect at WOT at 7K RPM. I was in high density traffic and the car stopped. Had to revert to stock ECU in a hurry and the car had a strange behavior after. During 3 minutes, it worked like if you pressed on and off on the injectors but was able to maintain idle.
I now have an issue with fuel pump powering. I had to use the jumper inside engine compartment to run the fuel pump or it wont work.
I start thinking it can be related with the immobilizer. Have you remove it when you ran the VVT Miata ?

Have a nice day

Gwendal

Abricos
contributor
contributor
Posts: 727
Joined: Mon Aug 18, 2014 12:32 am

Re: 2004 Mazda MX5 1.8 VVT #50

Post by Abricos » Mon Apr 22, 2019 4:59 pm

It's no really good thing to spin engine up to 7 rpm if ECU is not tuning ...it's really easy to kill motor ...

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT #50

Post by Crazy Striker » Mon Apr 22, 2019 5:04 pm

I got to that point pretty progressively while checking for AFR and detonation. I think the tune is taken form the same car and is kind of conservative.
The only tuning issues were at load load and low RPM where the map is way off

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT #50

Post by Crazy Striker » Tue Apr 23, 2019 2:59 pm

Fuel pump issue solved, it was a bad solder joint between PE4 and the 20 Ohm resistor for the mosfet charge.

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT #50

Post by Crazy Striker » Tue Apr 23, 2019 6:35 pm

After loosing the sensor ground and reconnecting it, I'm still facing my MAP sensor issue. I ordered a MXP4250AP from RS this evening and I should get it by Thursday.
I got my first coldstart this afternoon, but the engine was quite strange because of the problem stated above.
Is the frankenso board already wired for that sensor in the proto area or should I make the circuit by myself ?
Thanks

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: 2004 Mazda MX5 1.8 VVT #50

Post by kb1gtt » Tue Apr 23, 2019 10:19 pm

Proto area has 5V and GND. You might manage to land those pins. The signal wire would need a little jumper wire to connect it to the analog input. Also might be worth mention, that you can use a GM MAP or similar and connect that to the harness connector. This then does not require soldered jumpers. As well you hose is much shorter.
Welcome to the friendlier side of internet crazy :)

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT #50

Post by Crazy Striker » Tue Apr 23, 2019 10:59 pm

I bought a GM style knockoff and it failed pretty quickly, I now prefer using a branded sensor, and considering how much I paid for it I hope it would be a real one :roll:
I soldered crank input directly but still have some misses during cranking, only when engine is hot and cranking faster.
I got the signal, my fuel pump starts (I can tell that because my wideband gauge is wired into fuel pump power) and the while still cranking it goes off and then on and eventually it catches after 2 or 3 on/off cycles.
When running I can't feel any misfire by the way
I may try to setup the sensor under the hood close to the manifold, but I would probably run into some supply decoupling issues. Maybe with a small proto board and a few capacitors it would work...
I keep you updated on that as soon as it arrives

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT #50

Post by Crazy Striker » Fri Apr 26, 2019 10:15 am

So I just received my MPX4250 and I have the same issue with the map value. It is down of the the real value with a coefficient of 1.77. I may have messed with the voltage divider on the MAP inputs. I will try switching inputs to see if it solves the issue

Abricos
contributor
contributor
Posts: 727
Joined: Mon Aug 18, 2014 12:32 am

Re: 2004 Mazda MX5 1.8 VVT #50

Post by Abricos » Fri Apr 26, 2019 12:39 pm

Crazy Striker wrote:
Fri Apr 26, 2019 10:15 am
So I just received my MPX4250 and I have the same issue with the map value. It is down of the the real value with a coefficient of 1.77. I may have messed with the voltage divider on the MAP inputs. I will try switching inputs to see if it solves the issue
In datasheet mpx4250 we have 3 option of pinout ... If one time you connect wrong it's 99.6% sensor is deed ... And you supply 5volts !!!
Screenshot_20190426-083612__01.jpg
Screenshot_20190426-083612__01.jpg (322.56 KiB) Viewed 721 times

Crazy Striker
kit assembler
Posts: 118
Joined: Wed Aug 08, 2018 6:05 pm
Location: Brest, Britanny, France
Soldering skill: yes
Coding skill?: yes

Re: 2004 Mazda MX5 1.8 VVT #50

Post by Crazy Striker » Thu May 02, 2019 2:46 pm

Hi all !
I wasn't particularly active last week because of my schedule but I'm back.
Just redid all the solders on the MAP input and as I thought it was a cold koint fucking all up.
Now with an accurate MAP reading I will be able to run on rusefi hopefully on a daily basis and let the autotune do his job to correct my fuel map.

So finally both sensors were OK...

I keep you updated
Gwendal
Last edited by Crazy Striker on Sat May 04, 2019 10:43 am, edited 1 time in total.

Post Reply