The FASTiva

[Native]

^^^ Yep - I am one happy monkey! The thing was fun before, but wow, even betterer now! The engine keeps up what pull it has way higher into the rev range, and it's not just a matter of keep it floored and steer now - the throttle does stuff! And that toe-out really made it feel much lighter and willing to turn, good call. But, for me, the best part is the intake sound - both inside and outside it sounds downright mean!

[Loren]

Another FASTiva Friday. Philip has been assisting me with some tuning efforts. Made some hardware changes to the car today, but nothing fun. Added a cigarette lighter plug so that I could plug in a power inverter and plug in my laptop (battery doesn't last much more than an hour) for more extended tuning sessions. And added a "shift buzzer" to yell at us and remind us that what we're doing it pointless and we should think about shifting, dammit!

Lastly, we tinkered with the distributor timing. What I noticed today was that the MS was never allowing timing to go below 19.8 degrees. After checking and rechecking everything, I came to the conclusion that it wasn't doing it simply because IT CAN'T. Turns out, I was right. We had set a really high "trigger offset" angle to compensate for the fact that we had advanced the distributor timing when we were running the stock ECU. At the time, we wanted to retain quick & easy backward compatibility, so we didn't touch the distributor. Just adjusted the MS to match.

Backing the distributor off so that it's right in the middle of its adjustment range rather than almost fully advanced allowed us to use a trigger angle of 0 degrees, and bingo! Now the MS can command and get ignition timing all the way down to 0 degrees. Why do we want that? Well, really low timing advance makes the car easier to start, for one. And if we ever settle on a proper idle configuration, we'll want timing closer to 10 degrees. Right now, we're idling (high... close to 1200 rpm) with 16 degrees. The cool thing is that we now have CONTROL over it and can adjust from there. Whereas before it was locked in at about 20, and could only increase from there.

Subtle stuff. Doesn't make the car any faster. Just makes me feel better about tuning it!

If you really want to geek out on MS idle stuff, read on:

Finally gave up on closed loop idle today. Why? Because, while it works great and can provide idle quality that is better than stock... without a throttle position sensor (yeah, we don't have one, maybe we'll add one at some point) it's really difficult to tell the MS WHEN to idle. I thought I could manage it, and almost can. But, what happens is that because we don't have TPS (MS doesn't know if we've got 5 or 10% throttle going) AND we don't have a vehicle speed sensor (MS doesn't know that we're cruising 30 mph), the conditions that initiate idle look A LOT like the conditions at cruise! So, puttering around, in a low-load cruise, the MS would kick into idle mode and immediate take away all of your fuel! Not good.

So, until we get around to installing a TPS, I've gone to an old school "warm up valve" configuration. Same hardware, just telling it to open the idle control valve xx% at xxx degrees coolant temp and accepting whatever RPM that delivers instead of setting an RPM target and letting the MS magically keep it there. This is not as refined, and requires us to run a much higher idle speed (so that the engine doesn't stall when you turn the headlights on and the cooling fan kicks on), but it works. It's fine for the FASTiva. Because race car.

[Loren]

The car performed well Sunday. I think it might be pinging a little bit. Probably should pull some timing advance. (no knock sensor, playing by ear)

We've decided to use the FASTiva for the FROC challenge. That means the driver seat needs to be adjustable. (it's been fixed since we put the race seat in) Joe Brannon made us an offer we couldn't refuse on a set of seat sliders... that project is under way.

Not much else going on. Shopping around for our next transmission. Not an upgrade... just staying ahead of the inevitable wear.

[Native]

Shopping around for our next transmission. Not an upgrade... just staying ahead of the inevitable wear.

It whines all right, especially at light throttle cruise/coast.

[Loren]

Just collected the car from Joe Brannon. He revised the seat mounting to include a set of seat sliders (that he donated -- Thanks, Joe!). Also made the seat back brace adjustable and improved the seatbelt mounting. Very nice installation!

We'll be taking it all apart to paint tomorrow. Might need to raise the steering column a tiny bit because the seat is now slightly higher.

On the ECU tuning side of things, I've figured out how to wire up the stock "throttle position switch" to behave sort of like the MS expects a throttle position sensor to be. We won't have "resolution", but we'll at least have 0% 50% and 100% throttle position. That's good enough for being able to tell the ECU when to actually go into IDLE mode. Might get that done tomorrow, as well.

[Loren]

Ben and I spent some time on the car Saturday, I did some tuning this evening, and had some "tall guys" try the seat last night.

We got the new seat brackets that Joe fabbed for us painted and reinstalled. Also got the stock Throttle Position Switch wired into the ECU as a "throttle position sensor". So now the ECU at least knows when you're off the gas so that it can idle when it should. Got that all dialed in pretty good today. Still need to adjust the throttle stop and TPS a little bit. But, it's got a beautiful idle at 950 now.

We're definitely going to have to adjust the steering column up some to clear the tall guy legs. The seat barely goes back far enough for them. (when Joe and I were putting it all together, we said "no way anybody will need to go back farther than that"... oops)

[Loren]

Got the steering column spacers trimmed yesterday, I think I found a happy medium that will work for those long legged people, but still feels right for the rest of us. No complaints from drivers today.

We put 36 runs on the FASTiva today! The idle tuning still needs some work, but the car is performing beautifully otherwise.

[Loren]

Put a couple more hours into the FASTiva with Ben today. We were focusing largely on getting the idle more stable and reliable.

Quickly learned that the biggest reason why the idle went all funky at the last autocross was because the adjustment of the throttle position switch was just a tiny bit off so that it didn't always return to idle mode. We got that sorted out and made a couple of other minor adjustments to idle tuning.

The other significant change we made today was to rewire the engine cooling fan. The factory setup has it controlled by a simple thermostat switch mounted in the head. It reaches a certain temperature, triggers the fan relay to turn on the fan. Problem with that method is that the ECU doesn't "know" when the fan comes on, so it's hard for it to compensate. (the factory ECU probably has a better time of it because it is using a MAF sensor with gives it an immediate and accurate response to changes in load... plus they have a lot more experience tuning than we do!) We wired it to the Megasquirt so that it has full control of the fan. When the ECU sees the tempurature that WE select, it turns on the fan, and when it turns on the fan, it increases the idle speed just a little bit and the engine doesn't miss a beat.

So, once again... didn't make the car any faster. But, it's behaving quite nicely.

Gave the tires a look. They took a beating at the Dog Track. (36 runs) But, we're going to wait until after the Dec 3 event to rotate them. That will put fresh tire edges up front for the Challenge.

[twistedwankel]

Gave the tires a look. They took a beating at the Dog Track. (36 runs) But, we're going to wait until after the Dec 3 event to rotate them. That will put fresh tire edges up front for the Challenge

Every time I drive my car the tires rotate. I wish I didn't have to replace my clutch assemblt tho'. Shit. The parts are a grand with shipping. God bless Mazdamedic for the labor.

[Loren]

Just took advantage of a Cyber Monday special and ordered our next set of tires. They're not in stock. But, hopefully, they will be before we need them around the end of January. We went with the narrower 205/50 option . RE-71R should make up for that. The price was right. Just under $400 after rebate.

[Loren]

The FASTiva had a rough day last Saturday!

Dropped a bolt from the shifter assembly on the way to the event. Ben fixed that with a few zip ties. Good enough to get through the day.

Then at the end of the day, we learned that our start-up fueling is a little too rich. Steve managed to flood the engine to the point that it took several minutes of WOT cranking with no added fuel to clear it and get it to start again. But, it did, and it was fine after that.

But, that wasn't enough for Steve. Then he left the headlights on when we stopped for dinner and killed the battery! Soraya rescued us, no harm done.

So, all in all, a fairly uneventful eventful day.

Oh, and we just hit 100 runs on this set of tires and we haven't even rotated them yet!

Next up, FROC prep!

[Native]

Steve FTW!

[Loren]

Now that I've got the idle tuned pretty well, I decided it was time to begin working on improving the starting characteristics.

I knew I was giving the engine too much air on start-up (opening the idle control valve too far), but it was working with the amount of fuel, and it DID reliably start. That's good enough for a "race car", and I was okay with it until some of the more important stuff was tuned. The result of that big gulp of air and fuel is that once it starts, it would leap up to 2200 rpm and hover there for a bit. Not exactly what you want to do with a cold engine, and just "weird" in general.

So, I started playing with it this morning. What I learned right away was something I really should have thought about a long time ago. This car has a "cold start valve". I haven't researched it, but it's likely one with a wax pellet in it just like the thermostat in the cooling system. It lets in a bunch of extra air for a cold start, and slowly closes as the engine warms up. I didn't think about this, and I was tuning as if the only air going into the engine was a tiny bit through the idle speed adjustment screw, and whatever I was giving it with the ISC. Wrong. As it turns out, I can leave the ISC completely closed on a cold start, and it will stay closed even as the engine warms up.

Just fixing that problem took care of most of the high-revving on cold/warm starts. From there, I tweaked the cranking fuel pulsewidth to get it to fire up a little easier. It's a lot better than it was. But, the parts car (completely stock) fires right up with about a half revolution of the engine, even on a cold start. This one still takes 1.5-3 seconds of cranking to start. It's acceptable... but, I'd really like it to start like stock. Will have to play with it more...

And this, kids, is the curse of the Megasquirt. You have control of EVERYTHING. Which means you're responsible for EVERTHING. And you "can" make everything as good or better than stock. It just takes a lot of studying, thinking, tinkering... and time.

[Loren]

Warning: More geeky stuff follows.

Been playing with cold and warm start settings last night and today. Got a pretty consistent 1.5-2 second start, but knowing that the stock ECU will kick off a very happy start INSTANTLY, like within a half second, I kept playing with it.

What I learned from looking at data logs (how could anyone possibly tune this stuff without the ability to log data?) was that the MS doesn't actually see the engine as "cranking" until it's done enough revs to count. (now that I think about it, I think there's a setting that's set to tell it to ignore the first 3 pulses... I might need to play with that) And it doesn't start delivering the "cranking pulsewidth" until "cranking" has been achieved. Cranking RPM, btw, is pretty reliably right about 280 rpm.

Some quick math shows that 280 rpm = about 4.5 revs per second. So, basically, what I'm asking the ECU to do is get the car started in about 2 revolutions. This being a 4 cycle engine, that's one full cycle of all cylinders. 4 spark events. Pretty impressive, if you think about it.

Digging into the logs, I started noticing things. First of all, the MAP (manifold pressure) clearly starts going down as soon as you start cranking. And you can see fluctuations in it as it passes each open valve. You can clearly see that it makes one full revolution and then the MS says "ah-hah, the engine is turning at 280 rpm!" and goes into "cranking" mode. Then starts delivering fuel.

That one full revolution of spin-up at less than 280 rpm takes right about half a second. Once it hits that and triggers cranking, the engine was actually STARTING reliably in well under a second! But, there was still that dead half-second before that, and it was bugging me.

There's a half second of spin-up, and then a couple tenths more before the fuel injectors start firing. 3-4 tenths after that, the engine kicks over and runs. What I need is to get fuel going earlier! I tried larger initial cranking pulsewidth. (there's a pretty cool taper feature that lets you control cranking pulsewidth down to the 1/10 second, so I set it a little heavy for the first tenth then tapered it down from there) It did seem to help a little, but also made it easy to flood.

Then it hit me. I need fuel earlier, and the only way to really get it is to increase the PRIMING pulse! I quickly found that in addition to that, there's a "priming pulse delay" to allow a lazy fuel pump time to get up to pressure, and I had it set ot .5 seconds. So, it wasn't even getting its priming pulse early enough. I ended up setting that delay to zero. And then increased the priming pulse from 4ms all the way up to 12ms. NOW it's getting enough of a priming pulse to actually ignite before the cranking pulsewidth starts (which required me to reduce the cranking pulsewidth back to "normal").

From there, I did some more logging and tweaked the timing advance in the cranking area to make it a little "healthier" when it finally does kick over. And now the engine kicks over pretty darned quickly! I still get that full revolution of spin-up, but after that, the engine kicks over within 2-3 tenths of a second! Total starting time less than 1.2 seconds.

I'm pretty pleased with myself.

I'm sure it's not "done" yet. Will need to check some more cold starts. But, it's certainly getting there. And, most importanly, I'm amusing myself.

[AScoda]

Not sure if it applies to your MS version or the FASTiva, but when I was trying to get that insta-start, I found that it will not start until it gets the RPM synced. Mine va-rooms to life as soon as the rpm sync comes on, so I figured it wasn't going to get any quicker.

[Loren]

Yeah, Drew. I need to look closer at that. I'm probably hitting that same wall. The car seems to start pretty well when it's warm. Still fighting the cold starts.

The FROC was a complete success! The FASTiva didn't miss a beat through 48 runs between 16 different drivers!

Even the tires seem to have survived quite well. After a flip, we should get another hundred runs out of them. Pretty cool!

Front:


Rear:

[Carracer]

Even the tires seem to have survived quite well. After a flip, we should get another hundred runs out of them. Pretty cool!

You and your humble brag, I would kill just to get 100 runs out of a set of tires...

[Loren]

Call it what you will, I'm loving it!

[AScoda]

the RS3s last a long time. they will likely heat cycle out before the tread goes away.

[ImpostorDan]

Even the tires seem to have survived quite well. After a flip, we should get another hundred runs out of them. Pretty cool!

You and your humble brag, I would kill just to get 100 runs out of a set of tires...

Wait, I of all people, got 100 runs out of my last set...