RyGuy

Hi Guys,

I was reluctant to start a thread over this but i've been sorting through this for over a week now and I've come up with nothing. I am hoping someone on the forum has encountered the same issue or can see something I've overlooked while doing a NB parallel install.

SO, the car!
1999 10AE Miata
Bone Stock Engine - No power mods
Hard Dog roll bar
Rebuilt shifter and Aluminum shift **** upgrade
KYB shocks
Center console delete - gas cap and trunk pull tabs relocated to drivers doorwell
AC compressor & accompanying components deleted

Turbo Project in Process!
DIYMS2 configured for fuel & ignition control, also controlling fuel pump and radiator fan
Stock ECU retained for IAC valve control, alternator voltage regulator, and tacho & CEL outputs
TD04 & stock wastegate
Kraken TD04 turbo manifold & downpipe
Stock Exhaust (for now)
Stock Injectors & fuel pump
NGK spark plugs and wires (stock coil packs)
LC-2 wideband

The backstory!
I've been working on the car for some time now doing other non-power modifications (roll bar, shifter, bodywork, etc.) and started the turbo adventure over the winter last year. I've built an MS2 from Diyautotune and configured the ignition input to run stock NB cam/crank sensor inputs as well as used the Joe Perez inverted spark output circuit. That being said, this is a parallel install with the stock ECU being retained to control the alternator, tach, CEL, and idle valve. I found out early on that there can be some issues sharing inputs with the stock ECU, namely the temperature bias resistors and the pullup resistors for the ignition input circuit. Once those were removed, the MS2 plays nicely with the stock ECU. The car starts (although roughly) but idles very high and while not oscillating, seems to change idle rpm frequently. For example on a cold start, the car will start and idle at ~1800-2200 rpm but will slowly increase as the car warms up reaching up to 2900rpm. I have set my ignition timing by locking the timing to 10BTDC even at the higher idle and adjusted to 5.5° in tunerstudio, however I don't want to start tuning until I can get the idle to settle down... the fueling, ignition, WUE & ASE tables I have are all based off a variety of startup maps I've found online. I've attached the msq file for reference. I should also add that I am currently running the stock IAT and narrowband sensor. While I have both the GM IAT and the LC-2 wideband controller on-hand, I'm trying to get the car up and running as stock with the MS2 before starting to swap additional components into the car so as not to confuse myself even more.

From what I can tell, the tachometer is not working (nor are any of CEL lights on the dash) and the alternator is not charging the battery (battery drains until dead, does not operate at ~14.6V). I believe how the stock IAC works is that it receives a constant 12V from the stock ECU (pin 3O) and the ECU pulls pin 3M to ground in order to modulate the pulse width signal. If that is correct, then if the stock ECU does nothing at all, the IAC would remain wide open which would create the extremely high idle problem I am seeing. Due to these issues, I believe the stock ECU is simply doing nothing at all as it does not seem to be performing any of its other functions either. I've verified continuity in all harness connections, and verified 12V at pin 1B so I know the ECU is getting power. I've swapped back to just the normal stock ECU and the car starts and runs as normal. I thought maybe removing the AC possibly had an effect as the stock ECU might believe it is running the AC compressor and therefore constantly opening the IAC to offset the load of the compressor but that can't be the case as the car runs with no AC and just the stock ECU without issue.

I've searched high and low through the forum and haven't found any threads or comments about this issue with a parallel install. I don't believe it's anything to do with the tune or the MS2, I think the stock ECU is not happy that i've taken some control away from it and that is somehow preventing the ECU from doing anything. I'm hoping someone who has been through a NB parallel install will have some insight or might see something obvious that i've overlooked or am unaware of. Any help is appreciated, I'm all ears!!

Thanks!! Please Enjoy Random build picture collage below...

DNMakinson

iTrader: (3)

Wire the wastegate open until you have bigger injectors. I suggest Flow Force 640.

I have no input on parallel install, only dismay.

DNM

RyGuy

You don't think the stock nb injectors can handle wastegate pressure on the TD04? From my research it sounded like they could support a smaller turbo. Either way the car is currently stock, while i have the turbo components, none have been installed atm. Im working on getting the car running and tuned with the ms2 before installing the turbo.

Haha everyone on the forum seems to loathe the parallel install and i can see why. However if done right it should be a beautiful thing with no real disadvantages other than the work upfront.

At this point i am tempted to just tear out the stock ECU, redo the harness, and build the alternator control and TIP120 iac circuits (i already built it I'm just not using it haha). However the engineer in me wants to understand what is causing the issue and to solve the problem!!

DNMakinson

iTrader: (3)

You understand correctly. I don't think stock injectors will support 7.5 psi boost. I have not proven it, because someone warned me as well. I went with Yellow Tops at first (will sell them to you for $120, one has broken connector tab, were cleaned and flow tested after I purchased them), but maxed them out as well at around 13psi, IIRC. Someone is selling FF NIB for $280 right now on the forum.

DNM

SpartanSV

You misunderstand how the IAC circuit works. If the stock ecu is doing nothing to control it then it isn't supplying the ground it needs and the valve stays closed. The stock ecu is still controlling the fuel pump right? Then we know the stock ecu is working.

What happens when you unplug the IAC valve when it's running?

Why didn't you attach a log?

deezums

iTrader: (3)

NB do not have 1:1 fuel pressure regulators, it's fixed to 60 PSI. At 10psi boost you shrink the injectors by 10psi. I think they are 265 at 3 bar, so at 60psi NB pressure minus 10psi of resistance means you are approx 280cc. By my math, that's like 130-150 wheel worth of injector. Way too thin of a margin, especially because NB usually peter out up top anyway on a stock pump.

If you swapped to return style with a 60psi 1:1 FPR referenced to the manifold I'm sure it would be close, but manageable. Maybe a slightly larger fuel pump. If you insist on running stock ECU you could even vent the FPR reference via solenoid to lock everything below 100kpa to stock fuel pressure.

EricBrute

Stock NB injectors should be good for 180-200 whp with the right setup. I was running stock NB red injectors with a Walbro 190 (non hp) on my M45 supercharged and they still got maxed out. You can see the duty cycle quickly reaches 80+ without reaching redline or peak boost ~7psi.

concealer404

iTrader: (3)

Incorrect use of word.

RyGuy

Gentlemen please! As of right now the injectors are not my concern. They will inevitably be replaced, as for now lets focus on the issue at hand!

The fuel pump is currently being controlled by the ms2 so no i do not have an indication that the ECU is working however, when unplugging the iac the engine immediately dies which leads me to believe the iac is wide open. So if the stock ecu pulls the iac to ground which opens the valve, then the valve must be constantly grounded - wide open. Would you agree? I'm trying to think of a way to probe the stock ecu, if the ecu is properly working i should be able to connect via obd2 correct?

Apologies for lack of logs, i am traveling currently and will attach when i return!

DNMakinson

iTrader: (3)

My experiences are more like Deezums’. But really, why play with special FPR and all that, when stock regulators, either NB non-return 60psi to tank OR NA return style 43 psi to Manifold will support all the flow the TD04 needs, and idle well.

Eventually, you will likely need a new fuel pump, though. Consider an OEM one. A new OEM will borderline support your max HP, and I have not found an aftermarket that was of equal quality. If your top HP goal is 200, it will be adequate.

EDIT: simultaneously posted. I’ll be quiet now.

SpartanSV

Yeah it sounds like that pin is being grounded. I would measure continuity to ground with the ignition off to confirm. Since there is no other sign of life from the factory ecu I'm thinking your installation is grounding it.

Obd2 should still work. No tach means the ecu either isn't working at all or isn't seeing the engine rotate. The check engine light should come on for a few seconds when the key is first turned on. Not having that is another sign the factory ecu is not powering on.

​​​​​​I think you should go through the harness you made and verify its correct.

RyGuy

Haha this is all good information regardless and i appreciate the input!

It is possible i have screwed up the harness although i have used the oem ecu diagram for the 99-00 ecu provided on trubokitty.com and verified several times, perhaps the diagram is wrong though i doubt it. I can post how I've wired my pinouts when i return.

I am currently sharing cam and crank sensor inputs with the stock ecu. Originally i had a 5v internal pullup on the crank and 12v pullup on the cam (or swapped?) But found the pullups between the stock ecu and the ms2 were conflicting. Removing the ms2 pullups solved the issue though i wonder if the stock ecu is still not seeing the cam and crank inputs correctly? If the stock ecu received no position sensor data, would it be grounding the iac, not regulating battery voltage, no dash output, etc? I suppose it would be in the same state as accesory position....

Even after the car has been running for some time, the initial startup check engine light and red battery light remain illuminated and never disappear from the dash.

SpartanSV

Your initial post said the CEL wasn't working. If the light is on them we know the factory ecu is on.

It's very likely the oem ecu is grounding the IAC pin when the key is on and the engine isn't rotating so the factory ecu not seeing crank signal seems likely. It also explains all of the other symptoms.

To troubleshoot further I would unplug the MS from your harness but leave the factory ecu plugged in with your adapter harness. I would then crank the engine and look for movement on the factory tach. If you have tach movement then we know the factory ecu has crank signal.

RyGuy

I thought the dash lights were supplied via a seperate controller in the dash and the CEL light indicator was lit when the ECU supplied a signal to the dash controller. I just assumed the CEL and all other lights on the dash were arbitrarily lit on startup by the dash controller regardless.

I did actually try cranking the engine without the ms2 and i did not get any movement on the tach. HOWEVER, even with just the stock ecu connected in full control, the tach does not display any movement during cranking... again i think the dash is fed by a seperate controller that won't register until there is power after the car is running.

SpartanSV

The ecu controls the check engine light directly and supplies signal for driving the tach. There is no separate dash controller. No tach signal means ecu doesn't see the crank sensor signal.

What are the chances you have your crank sensor wires flipped?

deezums

iTrader: (3)

The CEL is not controlled by the microcontroller in the cluster, at least on a 2002. One half of the bulb is directly tied to ignition, and the other switched ground trace runs out the harness. In fact, none of the lights are switched by the micro, except maybe the low fuel light in newer NBs.

You are trying all this without schematics? Brave man.

The tach in the cluster is fed by the ECU. You could use the MS2 to feed it, exceptionally easily. Trubokitty.com has the ECU pin that feeds the tach. Everything listed in your initial post but the alternator could be controlled by the MS2, and if I'm not mistaken a NA8 alternator would fix that easily if you are that opposed to the westfield board.

Are you trying to run fuel off the megasquirt 100% of the time? And you expect to be able to use the CEL off the NB ECU for what? Use the MS2 to make a shift light out of it and be done with it.

Seriously think you need to reconsider a parallel install...

RyGuy

Helloooo Gentlemen! I'm back, I have not yet succumbed to the icy grip of the viral plague. I hope everyone here is healthy!

Sorry for the delay, due to all the corona panic and starting a new job (in the middle of the corona panic - not fun), I had to set this aside for a time but i'm back now!

I've attached the missing logs that were requested. I apologize in advance, I only have the "lite" version of megasquirt so I can personally only view them in limited capacity - I didn't find them very helpful. They should be of idle during/after warm-up with a few throttle blips.

I've gone through my harness and verified the pinout against the pinout on trubokitty.com and traced the cam/crank sensor signal lines on the oem diagram so I know everything should be connected correctly!

For reference: I have cmp input on pin 2H, and ckp input on pin 2J for the stock ECU.

Sparton/Deezums, you guys are right. Both the check engine light and battery charging light are directly connected to the stock ECU. When the ECU is connected, both of these lights remain on indefinitely, even while the car is running. With the ECU completely disconnected, they do not illuminate (obviously). So now we know for sure the stock ECU is alive!! For whatever reason, it must not be registering the engine position and does not know that the vehicle is running - which explains the wide open idle valve, no charging battery, no tach, etc.

Quick background: per the megamanual instructions for 99-00 Miata, I wired the ckp signal into the VR input circuit and the cmp signal into the optoisolator circuit. I used a JimStim on 99-00 miata trigger simulator to set the trim pots.

I'm no electrical genius so don't laugh if this is a stupid question, but is it possible the trim pots on the VR input circuit are somehow affecting the stock ECU's ckp input considering the signals are shared? I know that the trim pots are only for conditioning the signal but perhaps I have them conditioned to a point where the stock ECU can no longer recognize the signal? Is that possible or am I just stupid?

I do have the oem schematics, I just glazed over that part I guess...

I don't necessarily care about the CEL light, I wanted to retain the stock ECU to control the alternator, idle valve, and the VICS system. I only wired up the diagnostic lights and egr/evap system because why not.

I will honestly just switch to full standalone, I just want to figure this out first. I know it's probably something incredibly simple that i've missed, I just don't know what...

M2Ken

One of the advantages of parallel install is ability to quickly switch back to OEM for emissions checks. I run parallel since Shiv suggested this many years ago and honestly never switched to full standalone. Granted my output is mild compared to most. And the only thing that gets funky is the MAF reading out of range and throwing a code.

Since you are parallel, what happens when you switch back to OEM ECU control. This is a couple of plugs for injectors and spark plug to stock coils. This tells you exactly what stock ECU is (or is not) doing. Start from known point, and then move forward. Right now you think stock ECU isn’t doing something; so eliminate that as a variable.

RyGuy

Do you mean just plug the stock ECU back in as full standalone? If that's what you mean then yes, that's the first thing i tried..

One thing to note, i am sharing both the 5v ref and signal grounds between both ECUs. I realize now that it makes no sense to have both ECUs tied to 5v ref line, regardless that shouldn't have an effect on the cmp/ckp signals.

Would sharing of the sensor signal grounds between both ECUs potentially create an issue?

deezums

iTrader: (3)

No, but running two 5V rails together can't be good. I guess it's probably alright since you shouldn't be driving past the current limit of either independently, but as you should share the signal ground and actual ground I'd only use one or the other, probably the stock ECU one. Could modify the megasquirt and run it off Vref, but might not have enough current for it, but would help ADC accuracy if it did. And since the megasquirt ties sgnd to actual ground internally, you're kinda stuck using it that way. Shouldn't matter on the triggers.

Do you have a scope? I figured the NB crank sensor was a VR with built in conditioning circuit that makes it behave like a hall sensor, even though we configure for a VR. Figure this is why we get away setting pot voltages so loosely. If it is a open collector ouput sharing should be alright, so long as you take care of the "parallel" pullup resistors as you have.

Check this out, a add-on board designed for MS2 and a miata, adds knock control. He is using a 74LVC as a buffer for the inputs. Kill it with fire?

https://github.com/drytoastman/MiataECU