1994 Miata MPG project B3 STEALTH MK II

Sorry.... I sort of left this topic hanging.

As you know, the car ran pretty bad when it snorkeled full strength hot air. The air/fuel mixture was all over the charts but mostly the car was running alarmingly lean. ... or so I thought.

The hot air upset the volumetric efficiency in a negative way, so the Megasquirt was set up to automatically adjust the air/fuel and try to maintain the target ratio. This is when I discovered Megasquirt and the wideband controller were reporting two different air/fuel ratios.

After several evenings of investigation, part of the mystery was solved. Apparently the spark plug wires were throwing off a bunch of EMF and the stray energy was being absorbed by the air intake temperature sensor harness. This "noise" was causing havoc with the megasquirt.

So this hot air experiment more or less got put on the back burner while the megasquirt gets sorted out. The megasquirt still has an offset that I can calibrate out, however there is also a slight signal drift that has me baffled. I think I may have broken my Megasquirt or my harness has some major issues..

The car still runs fine and the Megasquirt continues to more or less "work" but it wont lock in on the target air/fuel ratios in my setup tables. Something is definitely broken because the gas mileage is hovering in the 40ish neighborhood. While the MPG is awesome for most cars, it sucks for this car.

Trying to determine if the wideband or the megasquirt was at fault is nearly impossible with the engine running. Eventually I had to do a bit of rewiring so i could safely inject a known voltage and read it back on both devices. When I injected a steady 2.4 volts DC into the system, both the wideband and the megasquirt reported 14.60 AFR. This is encouraging . The problem only presents itself when the engine is running. Automotive electrical systems are known to be "noisy" and I'm picking up a lot of ignition noise on the battery circuits. I added a bunch of capacitors here and there to sort of filter the junk but the problem remains.

I typically troubleshoot a problem until it's fixed.... however this time I said berkeley it. I'll figure it out on the dyno. The dyno electrical system is simple and clean. The dyno is still a few weeks away from being 100% functional, but preliminary findings indicate the megasquirt is still acting goofy.

So... this project is still going 110% full speed ahead, but um unless you stand back and look at the big picture... it doesn't seem like much is going on. I'm relying on the dyno to sort out zillions of minor issues that have cropped up and new issues that are going to crop up. The car plus the dyno is a massive project but I am so close to getting results.

Eventually the hot air project will make its way back to the front line.... if it works on the dyno. Snorkeling 100% hot air ain't going to work, I can tell you that much. A compromise has to be made.... We will find out soon enough.

My highspeed internet modem is on the fritzes. I only get internet in five min chunks before I have to kick the modem off line. I'm going to try and upload this post, but the formatting may be a bit off.

Oscilloscope was picking up a classic ignition "flyback" signal from with the probe attached to nothing. That is a significant amount of energy .

Some more electrical noise on the 12V DC circuits. I don't recall if this is after the ignition was fixed.

El'cheapo spark plug wires were fabricated to try and reduce the EMF under the hood. These temporary wires worked but the problem was not 100% solved.

a bit of re-wiring and I was able to inject 2.4 VDC into the megasquirt and the wideband gauge.

Megasquirt reports the correct AFR with the injected voltage.

Gauge agrees with megasquirt ... as long as the engine is not running.

Stay tuned!

High speed internet in five minuets chunks sucks. I should have the new modem next week. First world problem.....

in reply to Mtompson,

Yeah, my 4.10 Torsen ain't doing me any good at the race track. Apparently I need to address the issue of 63 hp before I can see a benefit from the torsen. I tried to trade it off for a 3.90 a few times but no luck. The car is almost off the road for the season, maybe I'll try and trade the torsen off again.

In reply to SFO,

The Miata is currently tuned to make full power up to 4200 rpm. The AFR table is written so the margins are power friendly and cruise is on the lean side.

The whole hot air experiment was flawed because it was all or nothing. The dyno should give me a better idea if the hot air has any chance of working in light cruise.

At some, point a I will find a way to screw a turbo onto the engine. Right now, I don't know E36 M3 about turbo's.

Does the wideband display vs tunerstudio afr discrepancy get worse the longer the car runs? Or is it just always a bit off?

Swank Force One wrote: Does the wideband display vs tunerstudio afr discrepancy get worse the longer the car runs? Or is it just always a bit off?

Good question...

After tweaking Tuner Studio AFR calibration both the megasquirt and the wideband agree for x-amount of time. Seem everytime I connect the laptop and check, the numbers are off. I'm not sure how long it takes for the calibration to drift.... never thought to track that. That's a good idea.

Any updates?

Not much going on with the B3 Miata. I stopped driving it in early December because of Michigan road salt. This car hails from Arizona and is remarkably clean for its age.

The little B3 engine normally doesn't work hard to propel the Miata down the road and as a result, not much heat is available to keep the cabin warm. Over at the Festiva camp heater issues are a common complaint. This car is running a custom thermostat to get the engine to run a little hotter, but Mother Nature has a way of sucking all the heat away. I really need to move to Texas or somewhere that's warmer in the winter.

Anyway, as most folks are aware, the B3 Miata project is linked to the low buck engine dyno project.... Both projects are insanely complex because I like a challenge.

I have discovered the B3 Miata will not run on a tune that was developed on the low buck dyno.... If I had any hair, I would be pulling it out... ARGGG!

I can get the Miata to start, but the air/fuel ratio is in the 15:1 range and gets leaner as the engine warms up. The volumetric efficiency table is not compatible between the two engines. I can enable autotune in tunerstudio to correct the VE table, but when I switch the Megasquirt back to the dyno, the dyno engine runs way rich.

Both engines are nearly identical...except the dyno has a .010 shaved head and the B3 Miata has a .020 shaved head. Shaving the head bumps the compression and also changes the cam timing slightly. I don't think this should be an issue.... at least not a big enough issue to cause the difference in VE tables.

The other difference between the engines is the ignition system. The Miata has a custom coilpack type ignition and the dyno uses a distributor. Unfortunately I cannot put a distributor on the Miata because of space limitations. Perhaps the custom coil pack ignition is too weak?

Although this is a major dilemma, it is encouraging. The dyno engine proves that something is wrong with the B3 Miata setup.

I have several plans in the works, To begin with, I'm going to transfer the custom coil pack ignition over to the dyno. It will be interesting to see if the coilpack ignition is causing the problem.

I also picked up supplies to build a complete standalone harness for the MegaSquirt. The B3 Miata currently has an altered harness made to run a Festive ECU. A short adapter mates the Miata/Festivea harness to the megasquirt. This could be where the cluster berkeley is taking place.

The B3 Miata uses a real Miata CAS combined with a custom circuit board to trigger a Saturn coil pack. Electronically this arrangement ought to mimic a conventional distributor. The megasquirt has no idea that it is dealing with a different ignition when it is transfer it from the dyno to the Miata. The custom circuit board sorts it all out and megasquirt thinks its dealing with a distributor.

A schematic of the custom ignition adapter. I'm actually going to replace this with an Arduino nano sometime soon.

what the custom board really looks like.

Dyno engine has a Festive distributor because I was to lazy to build a coilpack ignition. I'm goint to transfer the Miata coilpack to the dyno for a test in the near future.

I was surprised that you can drive coil packs with a 2N2222. The Miata coil packs must have an integral driver.

erohslc wrote: I was surprised that you can drive coil packs with a 2N2222. The Miata coil packs must have an integral driver.

Good catch!,

Yes, both the Miata and the Saturn coil packs I have used have integral triggers. The 2N2222 is just buffering the signals.

Just a quick update,

The Miata is out of hibernation and has returned to daily driver status. Moar updates soon...

So.... this happened. Don't judge me

Ok, I really like the CorkSport adjustable cam sprocket, but I also want a matching sprocket for the dyno.....so after quick search of ebay I found a matching set of adjustable sprockets of questionable quality.....

A lot going on in this picture, so pay attention.. or not. (sometime I just look at the pictures too)
Anyway, I'm fix'n to install a timing belt and cam sprocket. I used TunerStudio to lock the ignition timing at baseline. I'm using the ignition timing as a reference for cam timing... sort of like dropping bread crumbs. If all goes well, the ignition timing will remain the same, if it changes... then I'll know the cam timing is off.

Zoom Zoom......

Keep up the good work! I love this thread and your dyno one!

Thanks!

More dyno stuff and Miata results coming soon.

It's been 50,000 miles since the car was transformed into the illest of the high MPG econoboxes. berkeley them hippies with the hybrids. This car rocks it at the gas pump and its edgy new look is a hit with the chicks. Let me hear a well deserved berkeley yeah!

Vinyl decals aside, this season starts off with a few other superficial mods and a fresh Megasquirt II. The previous MSII is being retired to the dyno project. In its place, a brand new MegaSquirt II with a 3.57 mainboard. Yeah, its a bit of coin to drop for essentially a trivial upgrade, but I called in some markers and in the end the MSII was very budget friendly. I'll take another berkeley yeah!

And with that, let's look at some pictures.....

Have you considered any aero for added mpgs. Like a partially blocked rad opening, may help with heat also, saw a clear one made with lexan with some holes drilled in it that looked clean.

Maybe a smooth bottom and some tire spats also, or some lowering. Tire spats made a noticable difference on the vette.

Come on Im not working on my turd so I need to live vicariously through you guys.

In reply to Opti:

Well alright, I'm willing to do subtle aero mods, I did remove the radio antenna after all. I'm usually extremely conservative in the appearance of what I drive so wheel spats ain't gonna to fly, however the radiator opening is do-able.

I think a block off panel with automated slats would be fairly easy to build. Dual door lock actuators would allow half or full ventilation.

I'll do a cardboard mockup and see what I can get away with.

edit.. Oh I forgot to mention the car was lowered 1 inch a few years ago. Ironically, according to the internet lowering a miata increases the drag coefficient. That don't seem right but the internet is never wrong. Anyway I added a semi stylish front chin spoiler thingy to help with whatever I berkeleyed up. Apparently lowering a miata and adding a chin spoiler is what the internet recommends.... Why cant things be easy.

In reply to fujioko: There are a bunch of cars out there will grill shutters. Maybe find one of them to work around.

See if you can find a crashed 1.0l Fiesta.

In reply to alfadriver:

Good point! I have also heard that the Chevy Cruze uses shutters. I'll take a look at the neighbor's car this afternoon.

What's the avg mpg been this season?

In reply to 2002maniac:
Good question!, apparently I have been pre-occupied with shooting cool photos and haven't provided any data.

Whelp, long story short, we are looking at 40-45 MPG. This is seemingly unchanged from last year. I cant really nail down the average because too many changes are being done but I did see 45 MPG twice this spring.

While the apparent no change may seem disappointing it is encouraging because the cam timing is still more or less stock. As I recall, last summer when I pulled the adjustable CorkSport timing cog and replaced it with an OEM cog the gas mileage dropped to the high 30's low 40's. The reason the OEM sprocket was installed last year is so I could gather baseline data for use on the Dyno.

The dyno has been gathering dust while I sort out the Miata, Anyway, I'm almost ready to capture another baseline profile on the Miata and feed it into the dyno simulation runtime computer. This will be the summer of results!

For your aero, here's some useful CFD.

https://hanchagroup.wordpress.com/2013/01/29/splitter-or-air-dam-which-design-is-best/

In reply to Keith Tanner:

Aesthetically, I would choose the lip spoiler which was modeled at .32- that closed front dam looks not so good.

That, plus address the turbulence under the car, and I bet that would be a pretty significant improvement over the stock configuration. Would make grill shutters more useful, too.

Keith Tanner wrote: For your aero, here's some useful CFD. https://hanchagroup.wordpress.com/2013/01/29/splitter-or-air-dam-which-design-is-best/

Very interesting. I am surprised by the differences between stock & lowered. I expected a small decrease in lift, but the increase in drag was pretty substantial.

Thanks Keith!

Looks like some radical changes need to be made for optimal CD. Perhaps at some point I'll go full cardboard and duct tape and see what happens.

alfadriver wrote: In reply to Keith Tanner: Aesthetically, I would choose the lip spoiler which was modeled at .32- that closed front dam looks not so good. That, plus address the turbulence under the car, and I bet that would be a pretty significant improvement over the stock configuration. Would make grill shutters more useful, too.

Apparently I agree with the aesthetic aspect I've got a flat undertray as well. Can't say I've really noticed a change in drag as I can punch through the air pretty hard, but I definitely notice the front grip.

I meant under car tire spats, like this, but you know better.