Page 1 of 9

red 2003 VVT test mule Miata NB2 #27

Posted: Mon Oct 03, 2016 6:01 pm
by AndreyB
I've decided to unify my test fleet, please welcome this automatic 1.8 VVT 2003 Miata.

First step would be locating the OBD-II port and fixing check engine, the price considering nice interior was good to ignore the check engine light.

https://svn.code.sf.net/p/rusefi/code/trunk/firmware/config/engines/mazda_miata_vvt.cpp
http://rusefi.com/wiki/index.php?title=Vehicle:Mazda_Miata_2003

JimStim reference https://rusefi.com/forum/viewtopic.php?p=40438#p40438

Tune status:
+ stats reliably
+ stars reliably without throttle application
- 80C engine idles close to 14.7 with manual idle control
- smooth warm-up with manual idle control
- smooth warm-up with automatic idle

Re: 2003 VVT test mule Miata

Posted: Mon Oct 03, 2016 7:08 pm
by stefanst
Awesome!
OBDII Port is right above the fuse panel.

Re: 2003 VVT test mule Miata

Posted: Mon Oct 03, 2016 8:55 pm
by kb1gtt
Sweet. Is this waiting for that connector that I'm nearly done with?

Is that "drive by wire" or "drive by the seat of your pants" :)

Re: 2003 VVT test mule Miata

Posted: Mon Oct 03, 2016 9:01 pm
by AndreyB
First it is waiting for a spot in the garage (depends on Neon rolling out either dead or alive, hopefully alive by the end of this week), after that yes the connector would help :) In the mean time I will have a chance to find the ECU, it must be somewhere...

Simple throttle cable, just with a VVT on the intake side.

Re: 2003 VVT test mule Miata

Posted: Mon Oct 03, 2016 10:28 pm
by AndreyB
codes

Re: 2003 VVT test mule Miata

Posted: Mon Oct 03, 2016 11:20 pm
by kb1gtt
Sounds like an exhaust leak or an intake leak for the P0420. I've checked for intake leaks by using a vacuum to lightly pressurize the intake, then listen for whistling noises. Often people will do this with a "smoke" device. Another way to look for intake leaks on natural aspirated engines is to spray the engine with starter fluid. It will typically increase revs when you find the leak, as it sucks in the starter fluid.

Does the retarded CAM return after you clear the codes? Also when was it's timing belt last done. You had that issue with the Neon. Perhaps they just had the belt replaced and perhaps the mechanic couldn't get it properly adjusted.

Re: 2003 VVT test mule Miata

Posted: Mon Oct 03, 2016 11:42 pm
by AndreyB
Looks like p0012 pretty much means "vvt system is not happy, either o-rings are gone or just too much crap in the oil passages". Looks like first step is trying Rotella T6 high in detergents oil (dual rated but the larger label says "heavy duty diesel") to wash it out, if not swapping some rings like http://miataroadster.com/miataroadster/miataroadster-vvt-system-viton-seal-rebuild-kits/mrvvtrk/i-1964289.aspx

Re: 2003 VVT test mule Miata

Posted: Mon Oct 03, 2016 11:48 pm
by kb1gtt

Re: 2003 VVT test mule Miata

Posted: Tue Oct 04, 2016 4:26 am
by stefanst
On the NB1 the ECU is under the steering column in the driver-side footwell. I believe the same is true for the NB2.b

Re: 2003 VVT test mule Miata

Posted: Wed Oct 05, 2016 3:04 am
by stefanst
Are you going to convert to manual before going all rusefi on its' behind, or are you keeping it auto? I don't think the tranny control needs much from the ECU, so I *guess* you'd be OK keeping it auto.

Re: 2003 VVT test mule Miata

Posted: Wed Oct 05, 2016 3:19 am
by AndreyB
I want to at least attempt rusEfi before manual conversion. But before that I need to fix VVT and ideally pass inspection :)

Re: 2003 VVT test mule Miata

Posted: Thu Oct 06, 2016 4:05 pm
by ItaloManyak
Check the oil pressure and o-rings if it is good then
Try to clean the valve or replace it

Re: 2003 VVT test mule Miata

Posted: Fri Oct 21, 2016 7:23 pm
by stefanst
I'm going to add my '03 (manual trans) Miata with VVT rusefi-conversion to this thread if that's OK, since there's going to be a lot of overlap.
Updated the wiki on NB2 (2001 to 2005) Miatas: http://rusefi.com/wiki/index.php?title=Vehicle:Mazda_Miata_2003
Will also add suggested wiring including WBO2 and MAP sensor.

Re: 2003 VVT test mule Miata

Posted: Fri Oct 28, 2016 7:07 pm
by stefanst
Just to make sure I got this right: The only way that the ECU know what angle the VVT is advanced by, is by comparing the CAM sensor signal to the CRANK sensor signal - correct?

Re: 2003 VVT test mule Miata

Posted: Fri Oct 28, 2016 7:52 pm
by AndreyB
stefanst wrote:The only way that the ECU know what angle the VVT is advanced by, is by comparing the CAM sensor signal to the CRANK sensor signal - correct?
that's my understanding of this, and that's the kind of logic I have for the toyota vtti.

Re: 2003 VVT test mule Miata

Posted: Sun Oct 30, 2016 2:39 am
by Abricos
:)

Image
Image
Image
Image
Image
Image

Re: 2003 VVT test mule Miata

Posted: Sun Oct 30, 2016 4:01 am
by kb1gtt
Gotta love 3D printers.

Re: 2003 VVT test mule Miata

Posted: Sun Oct 30, 2016 11:38 pm
by AndreyB
Really confusing +12 running / +12 constant / main relay, got couple of wires wrong, diagram is a labyrinth, got tired added an unneeded wire which made power supply side really funny, nearly fried some sensors but somehow these survived, the ex-Neon board had a non functioning 5K resistor which was not used on Neon but was important on Miata... Thank you @ for doing half of the work!

But now I have both cam and crank, four injectors, fuel pump and main relay. Open question trigger shape but that should be available online.

Re: 2003 VVT test mule Miata

Posted: Mon Oct 31, 2016 12:49 am
by stefanst
Eating burritos with @ and his wonderful wife is a real treat :) Thanks for lunch!

On the VVT trigger shape, I found the following on Miataturbo.net:
Image
So it's the same as NB1 without VVT, the CAM signal just moves relative to the CRANK signal. Should be OK to run just using NB1 setup.

Re: 2003 VVT test mule Miata

Posted: Wed Nov 02, 2016 10:15 am
by Abricos
Image
Image
Image
Image
Image

Re: 2003 VVT test mule Miata

Posted: Thu Nov 03, 2016 12:45 am
by AndreyB
stefanst wrote:On the VVT trigger shape
just realized that we had an NB1 running - see http://rusefi.com/forum/viewtopic.php?f=3&t=467&start=33 see http://rusefi.com/wiki/index.php?title=Manual:Software:Trigger#Mazda_Miata_NB

Re: 2003 VVT test mule Miata

Posted: Thu Nov 03, 2016 1:22 am
by stefanst
Earlier this year we also had my NB1 running http://rusefi.com/forum/viewtopic.php?f=3&t=987 :)

Re: 2003 VVT test mule Miata

Posted: Fri Nov 11, 2016 4:05 am
by stefanst
Managed to simulate Miata NB2 VVT cam/crank signals with an arduino. rusefi accepts the signal and reads rpm. I can also see the crank signal moving relative to the cam signal, but am not sure how to use Miata VVT with rusefi in Tunerstudio. Will play some more tomorrow.
It only gets up to 3670rpm, because arduino is soooo slow, but it should be good enough!

Re: 2003 VVT test mule Miata

Posted: Fri Nov 11, 2016 5:09 am
by AndreyB
Unfortunately no VVT code for Miata yet. This signal is too complex for the current implementation - current implementation only works with one teeth cam sensor (we have three) and distinct crank signal (we only have 180 degree precision, not 360 degree because the wheel is symmetrical)

I think I have a plan how to implement this trigger, I just need to sit and do it.

Re: 2003 VVT test mule Miata

Posted: Mon Nov 14, 2016 3:15 am
by AndreyB
I've got some useful UK parts but UK-specific wiring diagram seems impossible to find :(

http://www.mx5nutz.com/forum/index.php?showtopic=324257

Re: 2003 VVT test mule Miata

Posted: Tue Nov 15, 2016 1:05 am
by AndreyB

Re: 2003 VVT test mule Miata

Posted: Tue Nov 15, 2016 3:11 am
by AndreyB
I think VVT Miata trigger support is ready - version 20161114

Crank is primary trigger - shaft position is based on that channel. Cam sensor is only used to detect intake cam position and engine phase (that's especially important with this symmetrical shape of crankshaft pattern). Special VVT mode is needed since three teeth, CAM gap ratio setting is temporary - once we confirm the numbers on a running engine these would be hard-coded.

Re: 2003 VVT test mule Miata

Posted: Wed Nov 16, 2016 12:53 am
by stefanst
Success!

Running the new VVT code with a stimulator we get VVT-angle feedback. We can sweep approx. 50 deg, which correctly mirrors what the simulator does. There is one issue though: The angle feedback from rusefi goes from 382deg to 432deg.
MS3 reports -5deg to 44deg.

Re: 2003 VVT test mule Miata

Posted: Wed Nov 16, 2016 1:13 am
by AndreyB
Great news :)
I wonder what is the meaning of -5deg to 44, is it in relation to where cam is on NB1 maybe?

Re: 2003 VVT test mule Miata

Posted: Wed Nov 16, 2016 1:16 am
by stefanst
My simulator has more freedom than the actual actuator on the intake cam. I believe the intake cam has a total range of 38 deg. I assume that MS3 just sets "full retard" at 0deg and "full advance" then at 38deg or so.