Joe Perez

iTrader: (8)

Since I first mentioned that I’d removed the AFM on my '92 back in December I’ve gotten a few p-mails asking about the setup, and I’ve promised those folks that a writeup is coming. My apologies for the delay, insert generic excuse here.

At any rate, it's finally done. I've been running with the AFM removed for about a month now, and it's going quite well. I've given this its own web page due to the graphics-intensive nature. You can find it here:
http://miatajoe.50megs.com/index.html

Perhaps this will finally earn me the coveted Sticky Thread for which I have longed O' these many months. (Hint to moderators )

Edit: Geocities sucks. Site moved to new host.


UPDATE: musanovic has now done a version of this for the NB (which may work with the 1.8 NA) here: http://www.rpaperformance.com/miatamaf.html

F20turbo

I cant view that outlook site....

cjernigan

iTrader: (13)

Even though i don't have an NA that was a hell of a write up.
+1 for making this a sticky.

I had to copy and paste the url into my browser, clicking brought up an outlook site as well.

hrk

iTrader: (6)

Link asks for username and password.
Copy/pasting works.

hrk

VRTSid

heres the clean link
http://www.geocities.com/thraddox/AFM/index.html

Atlanta93LE

iTrader: (5)

I hope all of those things under "about the author" aren't true...

Joe Perez

iTrader: (8)

Dammit. Now I'm getting a "site exceeded data transfer limit" error...

I gotta find a new web host. Hopefully this will be back online shortly.

Braineack

iTrader: (62)

He has been knifed in a London subway, tear-gassed by the Michigan State Police, narrowly avoided a terrorist bomb at a theater in Tel-Aviv, and was once denied service at a Howard Johnson's restaurant.

you boner! :gay:

Joe Perez

iTrader: (8)

Ok, sorry about all that. I've temporarily moved the site to a new host:
http://miatajoe.50megs.com/index.html

Tonight I'll try to migrate it to a permanent server. Damn Geocities...

VRTSid

hey joe I made a mirror, if you want I can leave it there forever or until you find a home for it.

http://jazz.sdsc.edu/miata/afm%20index.html

RicanmiataRacer

wow you kick *** joe perez, thanks for the read, doess this go for the 1.8 miata as well?

Ben

iTrader: (33)

holy crap dude that's a lot of work. there's absolutely no way I would have the time/energy/patience to do that.

What are your driving impressions?

Joe Perez

iTrader: (8)

The basic principles are the same, however it would require an entirely different map. For one thing, the hotwire sensor that the 1.8s use is non-inverted, meaning its voltage output rises with increasing airflow. This is in contrast to the 1.6 flapper-style AFM where the output decreases with rising airflow.

More importantly however, the airflow vs. MAP/RPM characteristics for the 1.8 are going to be a lot different. Not only is the engine larger, but it has different cam profiles, a different head design, etc. I suspect that you'd actually need different maps for the different 1.8 head styles- one for the '94 to '97, one for the 99+, one for the variable intake cam...

However, if anybody has a 1.8 Miata they'd care to loan me for a few weeks, I'd be happy to generate the data for it.

Joe Perez

iTrader: (8)

That's why I did it for you guys.

Well, I was a bit discouraged by the initial hesitation problem, but that seems to have been cleared up by the Accel map. I suspect if I had a real TPS I could have dealt with it more elegantly, but such is life.

I have not really noticed any serious change in driveability, which I consider to be a good thing. One benefit is that I'm now able to run my BOV direct to atmosphere, through a K&N filter of course.

My biggest gripe with the AFM was that after I'd installed by BOV the fuel pump relay would click off and on a few times every time I shifted due to the flapper banging back and forth. It did this even after I plumbed it for recirculated, which puzzled the hell out of me. Fortunately this didn't cause any serious problems, but being toggled every single time I shift can't be good for the relay contacts. Switching a big DC load like the fuel pump causes relay contacts to degrade, which is the typical cause of failure for these devices.

sbrian2

iTrader: (12)

Awesome job. Glad you got it sorted out. I spent a little time with it but gave up soon after aquiring the TEC3. Get a auto TPS, you will thank yourself. Using that with the accel enrichment maps made my car perform very well, no tip in or stumbles.

Falk

iTrader: (1)

Hello Joe,
great write up!!

2 points:

1: whats the reason why the 2 bar sensor dont work correct?
From the data, it should deliver same results.

2: The hotwire 1,8 Airflow Meter ( is ít still a AFM?)
dont restrict the inlet that much and need not a lot space.

I remember, with the Ultimate its easy? possible to use a 1,8 AFM
on a 1,6 system with 1,6 ECU. Is that correct?


Falk

Joe Perez

iTrader: (8)

For one thing, the output of the 2 bar sensor was a lot noisier. Individual intake pulses were showing up as huge spikes in the output. It's almost like the Greddy sensor has some built-in dampening to ensure that it produces a smooth output.

Secondly, the EMU software limits you to dealing with a voltage range from 0.5 to 4.5 volts, which is what the Greddy sensor produces. The 2-bar MAP sensor operates from 0 to 5 volts, so it's off the scale at very high vacuum and at very high boost.

The 1.8 is technically a MAF (Mass Airflow Sensor) and you're right, it's a much better design. I probably wouldn't have bothered with this whole project if I had a 1.8 ECU.

I've heard people say that this is true, and there is a drop-down for it in the vehicle parameters tab, but I don't see the BP or B6(NB) sensors listed, so I don't know if Greddy has actually coded the data for it.

You could do it via the same method- namely connecting it to the AFM input and selecting AFM(v) in the Airflow Map, but since you couldn't log both sensors back-to-back that way I'm not sure how you'd generate the basemap.

It might also be possible to use the Analog Input-Output wires for this and just do a 1-d map, but again, I'm not sure how you'd do the initial logging without some external equipment. I can't seem to enable Analog Input on the logger screen.

Chris Swearingen

iTrader: (1)

I got a chance to log Joe's AF Output map against my RX-7 AFM. Looks really close.


The dyno also showed about 1 pound of boost improvement. The only thing I can attribute that too is the change of AFM to the rx-7 and larger filter. Can't wait to get rid of the AFM completely.
Log file attached

Ben

iTrader: (33)

Seem to have my initial "issues" cleared up. I guess my car pulls a little less vac at idle than yours?

I set the first column to 650 rpm and raised the voltage output in the -12.3, -10.1, and -8 cells in that column... Seems to be running well, pretty nice idle, and right now I'm running 0 injector in that area, but I may end up running a little correction due to my big 550s.

Chris Swearingen

iTrader: (1)

Joe,

Thanks for this mod!!!!!! It has produced somewhere between 1.5 and 2 PSI FREE BOOST on my M45 blower.
--Chris
You can read all about it here.