rusefi.com

Oh I should also say the curve I'm using is essentially min(voltage - 1.2, 0) ^ 3 * 2.5, which matches the value I found on the forum. I have been changing the start parameter (1.2v) to match the bottom of my sensor (~0.7v). with this set to 1v it tops out at 160kg/h so that's very close to what you are saying.

We do not have calibration for this MAF sensor do we? Being the first 95 mazda you have to start with MAP sensor and WBO AFR sensor to get scientific MAF calibration for everyone else?

@Andrey B is correct. We do not have a verified calibration curve (transfer function) for this MAF. As near as I can tell, at idle your air-flow should be approx 7-10kg/h. This is what I get with the NA6. Your MAF calibration seems to be reporting 2.0 kg/h or so, according to your log. The NA6 calibration curve starts at around 7kg/h. So you seem to be off by a factor of approx. 4.

The curve you found on the internet seems to be off. At this point, we need to come up with a calibration curve for the NA8 MAF. While it's possible to do so from scratch, it would be much, much, much faster and more efficient to get the car running with Speed-density, record MAF readings and calculate the transfer function from there. You will also definitely need a wideband O2 sensor. Sorry, but at this point, getting a transfer function without using the MAP sensor and a wideband would require hours and hours of trial and error. And even then you would end up with something quite inferior.

Point taken. I will purchase a wideband and install the vacuum hose.

I noticed that the MAF transfer function on AndreyB's NB2 tune https://rusefi.com/online/view.php?msq=28 would put my idle in the right ballpark, so I used that. I also realized that my car always dies aproximately two seconds after startup, when the fuel unexpectedly drops, and I figured out that this is because of the cranking enrichment period. So I increased that time to 4 seconds, and was able to run the car longer than before. Increasing the enrichment period to 10s yielded in this log: https://rusefi.com/online/view.php?log=43 I was able to rev it up somewhat with the throttle which I thought might help me keep it alive, but it still died about the same time the cranking enrichment went away. It looks like the MAF may have the same callibration as an NB2, which makes a certain amount of sense.

What do you guys think? Is this how rough you might expect an un-tuned car to be before it has warmed up on the base tune (ie if it was one of the well known cars like an NA6 or an NB2)? Or should I avoid playing with it until I have a wideband to look at?

Without the wideband and MAP you are effectively blind.

Cars run just fun while very not tuned. The fact that yours shut down tells me that it's either _extremely_ off tune or something else is wrong.

While we are blind we are not very effective.

I realized that was a pretty silly question, since I can warm up the engine myself. I warmed up the engine using the stock ECU, and tried again after richening up the VE table between 150 and 120 for safety. I set the timing at 10deg just to be safe, and got this log: https://rusefi.com/online/view.php?log=44

It definitely wasn't pretty, and I still had to use the throttle to keep it running, but ran for quite a few seconds and even charged the battery. The O2 sensor even seems to have started working. I'm not certain if I have it configured correctly (it's left on a custom preset that was in the base tune), but it's showing a value, and it suggests that the car is rich (which I would hope that it is with 150% fuel.) Hurray!

It's not all good news however, since there was a loud buzzing noise that sounded like electrical interference (not rpm dependent) At first I thought it was coming from my stereo, but I verified that that was off. It sounded like maybe 200 hertz sawtooth or maybe square wave, and slowly built in volume after a few (maybe 20) seconds of running. Could this be from a PWM circuit somewhere? Could it have been overdriving the IAC or some other PWM driven control?

Also does anyone know how to output the tachometer signal for this car?

dang, didn't see your reply earlier... Is there some way to get notified when people respond?

EDIT: found the subscribe to thread function.

The buzzing you hear most likely comes from the IAC valve. This is controlled by PWM. We may have to change the frequency. Nothing to worry about for now.
I am going to assume that you are not seeing RPMs on the tach and this is why you asked about making the tach work.
The tach should just work out of the box. The tach signal on the NA6 and NA8 Miata is generated by the coils (igniter) and not the ECU. The only thing the ECU provides is a pullup resistor to 12V. For this to work, you need R55 which enables the pullup installed on the baord. If your board is missing R55, it would be the most logical explanation as to why you don't see the tach working.

When you say you see data from the O2 sensor, is that the stock narrowband sensor, or an aftermarket wideband?

Thanks for the explanation.

- Tach. yes I am not seeing revs on the tachometer in the dashboard, but I do see them in Tuner studio. I'll look into this, but I believe the NA6 dash may be different to the NA8. I will check R55, and if it's not that, do some research and report back.

- O2 sensor. Yes this is the stock narrow band. I have purchased a wideband, but it is still shipping. I know that the stock O2 sensor is basically useless for my current needs, but if I wanted to get accurate values from it after the engine has been tuned, I should switch to the Narrow profile for the EGO sensor in Tuner Studio, right?

mapaul wrote: ↑

Tue Oct 06, 2020 7:38 pm

but if I wanted to get accurate values from it after the engine has been tuned, I should switch to the Narrow profile for the EGO sensor in Tuner Studio, right?

We have very little experience trying to get anything useful from a narrowband. In 2020 in the US it's just not an effective use of your stamina.

mapaul wrote: ↑

Tue Oct 06, 2020 7:38 pm

but if I wanted to get accurate values from it

Narrowband sensors, by definition, cannot do anything relating to "accurate values". The only way they're usable to is to slowly nudge fueling up/down to bounce either side of stoichiometric (14.7-ish AFR for plain E0 gasoline). If you want to target anything other than perfectly stoichiometric, they're completely useless.

My goal is to continue to be able to switch back to the stock ECU if I go on a trip or if rusefi is acting up for some type of driving. So I don't want to make changes that prevent the stock ECU from working well I also have no desire to add an ugly led lit gauge to my otherwise stock interior.

I am fully aware that the narrowband O2 sensor is essentially useless except during very specific circumstances, but my understanding is that many people in the mega squirt world tune with a wideband, but run with a narrowband. I had been hoping to do something similar, which I assume would require that rusefi is correctly interpreting the values from the stock sensor during the very narrow window where it has data. That is all I meant by "accurate data"

While I've never met anybody who actually removed the wideband to run a Narrowband, it may be feasible.
However, most widebands can emulate narrowbands and would therefore not throw an error code, if installed and set up correctly. So you can have your cake and eat it too.

Hi Mapaul!

Any progress?

All the parts have now arrived. I'm just trying to figure out how to install everything. This is my first time trying to run something through a firewall, and this car has AC and everything else stock, so there aren't a lot of holes to choose from. Hopefully I will get it installed this weekend.