[rusEfi] 1994 USDM Mazda Miata #11

Your chance to introduce yourself and your vehicle
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

1994 USDM Mazda Miata #11

Post by spags »

Hi everybody, I've started this thread to document work on my 1994 miata. This is more or less a continuation of this thread http://rusefi.com/forum/viewtopic.php?f=2&t=706 but in a proper spot.

Car: 1994 Mazda Miata, stock (engine anyways)
Objective: A good ecu for N/A applications.
Board: Frankenso r0.1 built by russian.

Status as of this writing: I've tested that injectors, spark and sensors are working. Russian has been very helpful during all of this. I've attached the log files to the first start attempts.

Things that are missing:
- Fan control wiring on the board. Will probably wire this out during the week.
- Battery light is on all the time (this might be nothing, waiting until the car runs to test)

Stay tuned
Attachments
MAIN_rfi_report_2014-11-09 14_32.csv
(341.84 KiB) Downloaded 906 times
MAIN_rfi_report_2014-11-09 14_41.csv
(429.63 KiB) Downloaded 891 times
MAIN_rfi_report_2014-11-09 14_52.csv
(164.29 KiB) Downloaded 905 times
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata

Post by AndreyB »

I've just fixed the bug with false 'pin not assigned' warning you have reported
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata

Post by AndreyB »

Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata

Post by AndreyB »

I've checked the logs but I do not see anything obviously wrong. Can you turn fuel off & record a video with a timing light? Do you want to charge the battery & maybe give it another try with a bit more fuel?

Also check the video of my cold start today just as a reference on how long I crank / how does it react etc: https://www.youtube.com/watch?v=UqBzuZEqcEo
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

Re: 1994 USDM Mazda Miata

Post by spags »

More testing with the board in the car.

There seems to be an issue with the ignition, the coils do not seem to be firing when the ignition switch is on the start position. It seems to fire when the engine is rotating and the switch run position, but not cranking. This happened under after any amount of cranking and on both coils.

Here are the videos:
http://www.youtube.com/playlist?list=PLyPvY2Eq9L9NoNxEBcUF-N8Uopm1YSfQ8
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata

Post by AndreyB »



Why does not it want to crank more, what is stopping the engine? You say started cranks forever if ECU is off. I guess that means something?

We need someone who knows something about cars :(
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Abricos
contributor
contributor
Posts: 849
Joined: Mon Aug 18, 2014 12:32 am
Location: Carteret, NJ 07008

Re: 1994 USDM Mazda Miata

Post by Abricos »

Hi ...
the car has anti-theft system ( alarm ??? )
because on video can be heard horn bep ........ bep bep ....... bep bep ...
?????
check the battery ...
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: 1994 USDM Mazda Miata

Post by kb1gtt »

Do you see the timing light blink when cranking? I didn't see it in the video, but crappy computer so perhaps it was there.

Could the timing be sparking too early, such that it fires and stops the engine? Can you easily disconnect the fuel injectors and see if it cranks more normal. AKA no fuel = no sparky.
Welcome to the friendlier side of internet crazy :)
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata

Post by AndreyB »

This does not make any sense whatsoever. So, here's what we have (please correct me if I am wrong) (everyone is welcome to see the whole playlist with new videos)

1) the car cranks perfectly without spark and fuel
2) the car cranks perfectly with spark only
3) the car only makes two cranks with spark and fuel
4) looks like the is no spark while cranking (!!!)
5) there is spark AFTER cranking (?!)

#5 could be the watch dog I guess - that would explain one spark on each coil after the engine has stopped.

(idea#1) I guess we need to test with fuel and without spark just to have a full set of tests?
(idea#2) We definitely need to run the sparkbench with a spare spark plug grounded on the engine head.

(idea#3) I guess it would not hurt to check if something is not soldered right? Who knows. I believe I did test it but a bad solder joint might be a simple explanation of all this weirdness which must have a simple explanation.
See http://rusefi.com/wiki/index.php?title=Manual:Hardware_Frankenso_board#Hi_side

two lower pins should have continuity to the MCU pins, one lower pin is GND, another one is unused.
two upper pins should be connected to the breakout area, another one is connected to +12, the last one is unused. obviously we are mostly interested in the two channels which are used to drive the coils - the other four channels are still unused.


Either the coils are inverted or something really simple.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
puff
contributor
contributor
Posts: 2961
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: 1994 USDM Mazda Miata

Post by puff »

if i were you i won't crank it for so long - have you checked the power wires? weren't they hot? %-)
might be interference. the test with timing light suggests that the outputs might be inversed.
the very first videos look like the anti-theft system is working or the outputs are messed up.

the digital log in dev console is way too slow to see the most interesting part. i'd hook up the separate logic analyzer to the discovery board to see what exactly is going on..
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

Re: 1994 USDM Mazda Miata

Post by spags »

Hi,

Idea #1 and #2 will be tested tomorrow.
Idea #3, no cold welds as far as I could tell on the ignition drivers. Traced everything down to the stm board pins and out to the breakout area. Ground works, one curious thing is I found continuity between the 12V_CONST line on the connector to Vdd _AND_ GND. No idea if this is normal. The LED's will tell us what is going on for good.

Also, while checking things, i noticed a loose strand of copper fall out of the board. Who knows that could of been it. Nothing seems damaged or smells funny.

Re: The car does have an alarm, but at this time I'm don't think the alarm is part of the problem.
1 - Not sure it that was the alarm. Zeb (the guy doing the cranking was fiddling with my steering wheel, momo....hmm....sexy)
2 - It didn't repeat itself with other cranking tries. No issues with the alarm the rest of the time on the stock ecu, this car is my daily driver (I have spares)
3 - This car ran on a DIYPNP for a few months without any issues with the alarm. Also we didn't notice any other subsystem affected by the alarm (fuel pump for example).

As for the cranking; I did check the battery wires after the 30 second crank, they where hot to the touch but not unbearable, In any case I agree cranking the car for that long is not a good idea. At that point I was trying to discard the car's electrical system as a suspect for the bad cranking while fueling.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata

Post by AndreyB »

Pin PE14 goes to high-side which goes to W7 on the board which is ECU pin 1G which is "Coil Pack Cylinder 1&4"

Pin PC7 goes to high-side which goes to W8 on the board which goes to ECU pin 1H which is "Coil Pack Cylinder 2&3"

This was upside down in the mazda_miata.cpp, thank you @ for pointing out the issues. Rumors are the car has started right away once the config was fixed :)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

Re: 1994 USDM Mazda Miata #11

Post by spags »

Image

It runs, as russian mentioned the final trick was inverting the pins. It started right up, while cold no less. The maps need work but where a great start. Got a nice solid idle. It smelled rich. What led me to the problem is that I ran sparkbench and got no spark on cyl. I heard one of the coils working and pulled out the plug for cyl 2. Lo and behold I got a spark. At that point it was a simple pin reassignment and it started right up. Should of done a sparkbench test from day one. Live and learn.

As for the spark not working while cranking. Maybe it was a wild goose chase or maybe that copper strand was the reason, no way to know :?

There was one casualty:I knocked the big capacitor next to the connector on it's side while disconnecting it from the car. No biggie, it can be easily soldered back.

Also found my car will idle very nicely using 300 as the PWM load. (set_idle_pwm 300)

So, whats next? I'm going to make separate posts for these items as work progresses but here is a short and possibly incomplete list:
Cluster tachometer has to work.
Wideband O2 sensor, this is a must for any aftermarket ECU, rusefi is no different. I have an opening on the ECU plug. Still debating how to do this.
Radiator fan needs to work, since I live in the outskirts of hell this is a must if I want to run this on my car.
Humming noise while the car runs, drives me bonkers. Maybe the PWM frequency on the ICV is not high enough and the solenoid jiggles? Funny thing, it doesn't hum nearly as loud when the engine is not running but in the ON position.
A/C, this will need some changes to the idle code.
World domination, that _IS_ what we are doing here, right.

Anyways, without further ado. Here are the videos.
http://www.youtube.com/playlist?list=PLyPvY2Eq9L9O7nKx8C8vgV7BlF9NxJjQU

I've attached the logs as well for your data munching pleasure.
Attachments
MAIN_rfi_report_2014-11-15 16_49.csv
(278.16 KiB) Downloaded 939 times
MAIN_rfi_report_2014-11-15 16_45.csv
(225.83 KiB) Downloaded 876 times
MAIN_rfi_report_2014-11-15 16_19.csv
(108.61 KiB) Downloaded 893 times
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata #11

Post by AndreyB »

Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata #11

Post by AndreyB »

spags wrote: Cluster tachometer has to work.
Hm.
As funny as it might sound, tach signal is generated by the coil pack - see BLK/WHT wire at the left-bottom corner of http://www.rusefi.com/wiki/images/2/2b/1994_miata_1.8_3.png

However this wire does go into the ECU and looks like it's R541 which is a 2K pull-up to +12v.

Step 1: open stock ECU case, with ignition ON measure voltage on pin 2I. Luckily for us that's the top row so it's easy to touch this wire where is says 2I on the oem PCB. If this in fact shows +12v then probably (on your own risk, sorry) you can proceed to
Step 2: add some pull-up resistor in the 2K range to +12 wire?

It would be fun if that would give you a working tach gauge.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

Re: 1994 USDM Mazda Miata #11

Post by spags »

After a bit of hunting for the right pin. I confirmed a healthy 11.55 volts on that line. Will rig something up tomorrow for testing.

Same for grounding 1L to test coolant fan relay activation.
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: 1994 USDM Mazda Miata #11

Post by kb1gtt »

I'm going to recommend you hesitate for a minute about the 2I pin, it probably has some significant voltage spikes. Let me look into this before you install the 2k ish resistor.
Welcome to the friendlier side of internet crazy :)
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

Re: 1994 USDM Mazda Miata #11

Post by spags »

Ok, holding off.

I know that the 94-mid95 models had a different arrangement with how the tachometer was fed. They use a signal directly from the coils. Trying to compare notes with other aftermarket ECU's to see how they set it up. Since it's something that happened during 1 model year it's relativlely rare.
deviator
Posts: 33
Joined: Wed Aug 27, 2014 1:25 pm

Re: 1994 USDM Mazda Miata #11

Post by deviator »

spags wrote:Humming noise while the car runs, drives me bonkers. Maybe the PWM frequency on the ICV is not high enough and the solenoid jiggles? Funny thing, it doesn't hum nearly as loud when the engine is not running but in the ON position.
Is it the same one I had? http://rusefi.com/forum/viewtopic.php?f=3&t=662&p=12363#p12312 This solved it for me.
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

Re: 1994 USDM Mazda Miata #11

Post by spags »

Confirmed... no noise with set_idle_pin off, and I have a diode.

My current theory is that the frequency of the PWM signal is to low and the solenoid jiggles a bit instead of holding. I saw this back when I had a MS. Have to dig up the discussion, but the issue was that.

Also, confirmed grounding 1L activates the coolant fan.

Still the tach signal is pending but i'm holding off until kb1gtt has a closer look.

EDIT: My PWM frequency theory for the idle control valve is totaly unsubstantiated, i don't know what frequency we are using and I can't remember what the stock ecu frequency was. Just hoping somebody that knows more than me discards it or makes it worthy of further inquiry.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata #11

Post by AndreyB »

spags wrote:Also, confirmed grounding 1L activates the coolant fan.
Perfect! Please add a wire between the corresponding breakout to E6 transistor breakout. Please check me, but looks like upper W12 hole to lower unnamed hole between W37 and W39
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: 1994 USDM Mazda Miata #11

Post by kb1gtt »

russian notes that there was some other web page or something that confirmed the 12V thing for tach was OK. That works for me, and I hear it works. So I understand the tach thing is resolved.
Welcome to the friendlier side of internet crazy :)
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

Re: 1994 USDM Mazda Miata #11

Post by spags »

Yeah, found a few pages for other EMS's. The clearest one was:
http://www.miataturbo.net/aem-59/ems-4-tach-setup-60653/

So I slapped a 2.2k ohm resistor I had laying around as a pullup to +12v and it works beautifully. Will change that to a 1.8K (don't have any 2k ones laying around) and see how it goes.

Still have to tidy up before calling it done.

That's 2 items off the list. Next up is finding what is causing that ICV hum and confirm that the a/c lines do what I think they do.

I've also ordered a case and a wideband so I can start actual tuning when I get those.
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

Re: 1994 USDM Mazda Miata #11

Post by spags »

Have an issue with fan control. Will have more time to hunt it down later but in the mean time i'm posting it here.

Issue: Fan is on all of the time. fanbench has no effect.
What I did:
I followed russian's instructions here: http://rusefi.com/forum/viewtopic.php?f=3&t=729&p=12375#p12375
This is a pic of the wiring (maybe it's wrong :? ): https://drive.google.com/file/d/0B-T5j-1DVQnaNUVuRG9ERmVKQUxwandkQnVhTGFvUW5jenRZ/view?usp=sharing
I'm running firmware revision 5396

I will probably have some time to tidy that up later tonight and test again.

Re: Tach
The tach reading is high by a about %5. Will see what running 1.8k does.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata #11

Post by AndreyB »

Tach: I bet nothing would change with a different value - it's a digital signal, it's either there or it is not. Also my own experience is that analog tach (OEM gauge) is off comparing with digital tach (aftermarket shift light)

coolant fan: let's see if that's a hardware or firmware issue
step a) update to latest firmware
step b) invoke tempinfo command and post result here

it's expected to show what the firmware thinks about the fan state.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

Re: 1994 USDM Mazda Miata #11

Post by spags »

Ok, tested. Everything seems to be working. The fan engagement temp was 75C and my engine was not cool enough. Eventualy, everything cooled down and I got to test engagement and disengagement.

Here is a sample output of the command:

Code: Select all

2014-11-16 20_50: PortHolder: Sending [sec!8!tempinfo]
2014-11-16 20_50: EngineState: confirmation_tempinfo:8
2014-11-16 20_50: EngineState: CLT v=0.92454 C=89.60513 R=612.51123 on channel 12
2014-11-16 20_50: EngineState: bias=2700.00000 A=0.00250 B=0.00000 C=-0.00000
2014-11-16 20_50: EngineState: @PC2
2014-11-16 20_50: EngineState: CLT R=612.51123 on channel 12@PC2
2014-11-16 20_50: CommandQueue: got valid conf! tempinfo
2014-11-16 20_50: EngineState: IAT v=1.26153 C=74.01892 R=911.11114 on channel 11
2014-11-16 20_50: EngineState: bias=2700.00000 A=0.00250 B=0.00000 C=-0.00000
2014-11-16 20_50: EngineState: @PC1
2014-11-16 20_50: EngineState: IAT R=911.11114 on channel 11@PC1
2014-11-16 20_50: EngineState: fan=Yes @ PE6
2014-11-16 20_50: EngineState: base cranking fuel 5.00000
I have to find some authoritative source for what the engagement temps are on the stock ecu. I'm sure they are higher.

So, confirmed, tach and coollant fan work.

Yay!
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: 1994 USDM Mazda Miata #11

Post by kb1gtt »

I believe 180F to 190F is a common temp for thermostats. I would think turn on at 210F and turn off at 195F would be a good starting point guess.
Welcome to the friendlier side of internet crazy :)
spags
contributor
contributor
Posts: 136
Joined: Mon Oct 13, 2014 1:04 am

Re: 1994 USDM Mazda Miata #11

Post by spags »

Yeah, those temperatures look reasonable, would like to make sure the calibration for the CLT sensor is sane (it looks that way?)

As for the humming sound, here's some info
http://forum.miata.net/vb/showthread.php?t=172015

I peeked at the code and the miata runs at 160 (Hz?); the stock ecu runs at 500 Hz and runs it through a low pass filter to make the wave form more friendly to the solenoid.
I think it would a good start to at least try and emulate that. Going to try this later this week to see how it goes.
Please stop me if i'm obviously going to break something!
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: 1994 USDM Mazda Miata #11

Post by kb1gtt »

I should also add a comment / request that the fan on and off set points be programmable via TS. I know at least one hill climb driver that wanted the fans to turn on at a lower temperature. I think he now just uses a toggle switch, and leave the fan running during the climb to prevent that issue again.
Welcome to the friendlier side of internet crazy :)
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1994 USDM Mazda Miata #11

Post by AndreyB »

kb1gtt wrote:I should also add a comment / request that the fan on and off set points be programmable via TS.
Already in the TS config :)

Code: Select all

	dialog = fanSetting, "Fan Settings"
		field = "Fan On Temperature",					FanONTemperature
		field = "Fan Off Temperature",					FanOffTemperature		
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Post Reply