| MIATA BUILD |
|
|
 | |  | July 19, 2006 - These two front uprights are the same except for one thing - the one in the foreground has an ABS sensor.
Our car won't have ABS, but we're going to use the upright with the sensor because it may come in handy for our rally computer. The front wheels don't spin so they should be more accurate in reading distance and speed.
entry 13 - tags: computer | | |  | March 31, 2007 - A similar big brake kit is used on the front.
There's a significant weight advantage and the Wilwood calipers make it really easy to change pads.
Hidden behind the rotor is the 44-tooth sensor for the ABS. If we can find enough technical information, we'll use this as a sensor for the Coralba rally computer. Coralba is not being helpful at all - we know exactly what the ABS signal looks like, but they won't tell us the required signal for the computer. Instead, we're told to buy expensive peripherals and patch them into the car instead of using the sensors that Mazda spent a lot of money developing. I want to just plug this one in, figuring that there's not a lot of variation in hall-effect sensors. Eric's not so sure, probably because he's the one who gets to use the computer! If anyone out there knows what sort of voltage range is acceptable as an input for a C-Giant rally computer, let us know!
entry 111 - tags: brakes, wiring, computer | | |  | January 26, 2008 - I made a comment a while back about accounting for the weight of the rally computer.
Well, it arrived from Eric a couple of days ago and there's not really a whole lot of grounds for concern. It's tiny! It's a Coralba C-Giant, although it's also quite possibly the least accurately named piece of electronics I've seen for a while.
entry 390 - tags: wiring, weight loss, computer | | | February 12, 2008 - While rooting through the Coralba instructions last night, I got inspired to revisit my plan to use front wheel ABS sensors as triggers for road speed.
I've fired off a technical question to Coralba to see if they can answer me, and done a bit of testing on a Miata to see just how the wheel sensor behaves. The biggest potential problem I see is that the Coralba wants no more than 10 pulses per meter. That means about 18 teeth on the trigger wheel - and as you can see in this picture, I have a lot more than that to deal with. I could grind off every second tooth, or maybe two out of three. First, let's see if Coralba responds with any decent technical information.
The fallback position is to read off the speedometer. This will be affected by wheelspin of course, but it's a very simple installation.
entry 402 - tags: wiring, computer | | | February 13, 2008 - Since I posted the comment about altering the ABS trigger wheel, I've received a number of emails from folks with suggestions.
Unfortunately, trying to find a different way to mount the sensor isn't going to solve my primary problem: how to use the signal. Coralba has told me (in a one-sentence email) that I should use their PGE-V-pres8. There's no record of that unit anywhere on their site or anyone elses that I can find, so I'll see if I can get some information out of a Coralba dealer somewhere.
As for the wheel sensor, Mazda put a lot of effort into mounting that very well for me. It's easy to chop a few teeth off the trigger wheel, much harder to come up with my own mounting that will stand the abuse. So if I can figure out how to set the signal up properly, I will use the factory design.
Adding to the fun is the fact that the wiring harness for my used C-Giant actually comes from another computer. It's missing the wires for the fuel sender (not critical to me, really) and the remote reset (okay, that one matters). I can use the reverse wire as a reset but Janel is really happy about the fact that this unit will count backwards if I have to reverse at all. I'll either have to buy another harness (expensive!) or figure out which of the pins is the remote reset and solder in a new wire.
If it was easy, everyone would do it.
entry 403 - tags: wiring, computer | | | February 14, 2008 - After a conversation with Pete at P-Sport (a US Coralba dealer) as well as Janel, I suspect I'm simply going to spend the money to solve the rally computer hookup.
Coralba does not share technical information and won't even confirm if my existing harness will work at all. Given that these computers retail for huge amounts of money new, it's probably not worth taking a chance. Janel commented that I should spend my time working on something else instead of reverse-engineering her rally computer.
P-Sport does have both the correct C-Giant harness and the PGE-V-pres8 converter I need to hook the factory Mazda ABS sensors to the Coralba.
entry 404 - tags: wiring, computer | | |  | February 15, 2008 - The appropriate Coralba wiring doodad has been ordered so that I can use the factory ABS sensor as a trigger.
But first, I need to lower the resolution. The factory setup sends 44 pulses per wheel revolution. That's about one for every 41mm! The maximum resolution the computer can cope with is one per 100mm. So I ground away 3/4 of the teeth. Problem solved!
entry 407 - tags: wiring, computer | | |  | February 27, 2008 - Time to mount the rally computer on the dash.
My original clever idea was to embed it in the dashboard, but a quick check using a handy co-driver pointed out that this would put the odometer out of reach. Not a clever idea then.
So, after a fair bit of looking and thinking and turning the computer over in my hands, I came up with what should be a fairly strong solution. I'm concerned about the mount vibrating around, causing various failures and making it hard to read. So this box is meant to be nice and solidly mounted.
First step is to prototype it using cardboard. And not just any cardboard, race cardboard! It's the box from Janel's helmet. Reduce, reuse, recycle kids!
entry 415 - tags: ergonomics, computer | | |  | February 27, 2008 - The cardboard template, transformed into aluminum.
I've never welded aluminum before, but after all of the TIG practice with the header I found it really easy. The biggest difference was how the filler rod melted - and it was really fun to simply melt the seams together with no filler at all. I could get to enjoy doing this. After the picture was taken, a bit of work with a wire wheel cleaned up some rough edges on the cuts.
Next, to make it lighter (not that there was a problem in that regard!) I decided to cut some holes in it. Looks as if my hole saws were intended for wood, not metal. Whoops. The holes were rather ugly looking and the saw kinda wandered around. No worries, I was planning to paint the bracket black anyhow. Raw aluminum looks cooler and more racy, but is far more likely to reflect on the windshield.
entry 416 - tags: computer | | |  | February 27, 2008 - The Coralba, fully installed.
The bracket turned out to be immensely stable, making the computer a solid part of the car. That's good. It also weighs next to nothing, also good.
The computer is partially wired in, enough that I can verify that it's functioning. Good news, the modified ABS wheel sensor appears to be working perfectly. I did a rough calibration using the car's odometer and drove around town pretending I was on a stage. It's a very cool little unit with a lot of capability, and thus it will take a bit of work to understand it properly.
Other than the wheel sensor, I also wired in the reverse signal. This prevents the odometer from adding distance when the car is reversing. I can add a second remote to the computer if this feature is ignored, which can be used to finish a stage and display all the useful information. Hmm, I'll ask Janel which she wants. She was very excited about the reversing feature though. I sure hope she wants to keep it, it took me an hour to splice in one wire due to the position of that wiring harness. Ouch.
entry 417 - tags: ergonomics, wiring, computer | | | February 29, 2008 - I did a bit of autotuning on the way in to work.
The long way, of course. After looking at the changes, it appears the engine is hungry for all sorts of extra fuel up top. This is excellent, as it implies the engine is making more power. Yay! The car should hit the dyno on Monday and then I'll know a bit more. Jeremy at Flyin' Miata took a look at the maps and massaged them somewhat, so I have a better starting point for more autotuning on the weekend.
I also discovered that the Coralba shuts down when the headlights are on. Interesting. I'll have to trace that one back and find out what's going on. Running lights are fine, it's just the main lights that shut it down.
entry 423 - tags: tuning, wiring, computer | | | February 29, 2008 - Aha!
The relationship between the rally computer and the headlights is now clear. It started when I went looking for an unswitched power supply for the computer. The last thing I wanted was for me to accidentally reset the computer by turning off the ignition, you see, as then I would have to face the wrath of Janel and the potential of a retaliatory "flat left over crest into rock wall" call further along in a stage. I installed a separate switch for the computer instead. Anyhow, right beside the switch was the headlight retractor switch. It has unswitched power. So I tee'd into that.
Well, it turns out that when the headlights are powered up, the electricity to the retractor switch is cut off. Makes sense, you don't want someone dropping the lights when they're supposed to be up. But this nicely cut the power to the Coralba at the same time. I can't see this in the wiring diagram, but a check with the test light confirmed that's what was going on.
I had a handy wire coiled up behind the dash that was labelled "BATT +". See, I'd been thinking ahead. But it didn't work. I pulled out a wiring diagram and it seemed as if it should. But it wasn't used for anything else. So I checked the fusebox. Aha! A blown fuse. Popping in a new one lit up the wire and it was a simple matter of teasing it out of the harness and rerouting it to the back of my computer switch.
This was one of those examples where doing something else paid off. My brain kept flipping through options until the "retractors don't work when the lights are on" solution came to mind.
entry 424 - tags: wiring, computer | | |  | March 10, 2008 - What about in the car?
Once I'd roused Janel from the comfort of a warm tow vehicle, we discovered that our concerns about nausea were unfounded. With the rally computer to play with, she was distracted and forgot to get sick. Excellent!
The intercom worked well, allowing me to hear her clearly. She couldn't hear me quite as well - probably because the speakers were further from her ears - but that's a good setup. I did discover how hard it is to concentrate on driving quickly while also carrying on a conversation about how the trip computer works. Once the communication became more one-way, I was able to work better on dealing with the inputs.
For one session, she brought along a book and read it to me. Seems like an odd test, but we didn't have pace notes for the track, you see. Again, no problems with nausea as long as she kept her legs braced and didn't let them flop around. So the track day was a good test then. We learned a few things, that was the goal.
Eric asked for a picture of the whole interior in the mostly-final configuration. Here you go!
entry 432 - tags: testing, skills, computer, intercom | | |  | June 7, 2008 - Janel, hard at work watching for instructions.
A TSD is more of a treasure hunt than the Targa is, and there wasn't as much navigation as we would have preferred. They have to be run in rural areas, though, and rural in Colorado means you don't get a lot of turn-offs because you're usually running along a valley or climbing a pass. Still, we did get much more familiar with communicating with each other, dealing with problems ("I said slow down!" "But it's sooo boring!") and Janel got to spend lots of quality time with the trip computer.
All joking aside, despite the warnings of others never to do a TSD with a loved one, we had no compatibility problems. Both of us were in pretty good spirits throughout the day. This was probably in large part due to our rally computer. See, the required average speed varied fairly frequently thoughout the stages. But for us, we set the Coralba to display the average speed and simply reset the tripmeter at each change. So I could simply look at the display and see that I was 0.1 mph high, or we could stop and wait until our speed dropped to the ideal level. Normally there would be much work with stopwatches and calculators along with a lot of stress. Our biggest problem was my apparent inability to drive very very slowly at times. We didn't get lost, though. That's the important thing. And any frustration was tempered with humor instead of anger.
We also learned quite a bit about the car. It turned out to be an excellent shakedown test as we'd never spent 12 continuous hours in the car before. First off, both Janel and I found we need more padding in the seats. My memory foam must not be thick enough. Our butts were asleep before we even got to the start line, over an hour from our house. Also, the car is hot and noisy on the highway and cracking the windows seems to draw in exhaust fumes through a number of small holes in the transmission tunnel. The latter is easy to fix. Hot, well, that's probably not going to be a big concern in Newfoundland. Noisy - it's time to look for some transit headsets. We won't be at a sustained 80 mph on the transits, but even the background nose at 40 would get tiring over the course of the day. I'm going to throw some light sound deadening at the back of the car to see if I can help this at all. I think I'm also going to go to a 4.10 rear end instead of my current 4.30, as I don't need such short gearing.
Mechanically, the car was faultless. Some sadistic organizer put a 20-minute stop in at Leadville. That's about 10,000' up, and you should have heard the poor carburetted British cars trying to struggle back to life. Let's hear it for fuel injection!
Okay, almost faultless. The suspension got a nice dirt bath after we spent about half an hour trying to maintain 38 mph on a fairly rough road, and it's creaking badly. I think I just need some bushing lubrication underneath.
A long day, but a good day.
entry 467 - tags: skills, testing, ergonomics, computer, seats | | |  | August 18, 2011 - Not all the work on the Targa car is big cool projects like skid plates.
It's also getting some maintenance. For one, I'm changing out the front hubs to make sure they're at full strength. The current hubs had a total of about 10,000 miles on them but a large proportion of that was with hot race rubber bolted on. And, just like the original build, I had to grind off 75% of the teeth on the ABS trigger wheel in order to drop the number of pulses per mile to a reasonable number. The ABS sensor is used to trigger my Coralba rally computer, and it can only deal with 16,000 pulses per mile or so. With the modified ABS wheels, I get about 9,900.
And then I found a problem. The replacement hubs have an ABS ring that's about 2 mm smaller in diameter than the stock part. This meant the air gap to the non-adjustable pickup was too big, and it wasn't triggering the rally computer. That's not good. Luckily, the rings are just pressed on so I pulled off the old stock one and put it on the new hub. If only I'd noticed that before I spent all that time grinding!
entry 914 - tags: hubs, wiring, computer | | |  | August 27, 2011 - A final check of the undercarriage.
The car's just back from the alignment shop, where they set up the caster and rear toe. The control arms are set for maximum camber using the factory adjusters, leaving the extra adjusters on the V8Roadsters control arms for me to play with. I'm taking a last trip to the track tomorrow to just double-check the handling balance and scrub in the race tires a bit. I haven't spent anywhere near as long on the handling setup this time as I did last time, but that's because the car is such a good, solid platform now. I've been experimenting with it for about four straight years. So I know how to quickly make the adjustments I need to make it work the way I want.
Meanwhile, Brandon is busy doing the last-minute prep on Nancy. Mounting emergency triangles, packing his on-board tool kit, etc. He's also chased down a couple of odd noises. The one thing that is giving him trouble is the trip computer. It's a Terratrip, and we originally purchased an interface box so that it could use the car's electronic speedometer signal. But it turns out that the NC Miata doesn't actually have a separate speedo signal if the car's fitted with ABS. The car figures out road speed from the wheel sensors, and it's all passed around via the CAN-BUS network. So scratch that idea.
Then he discovered the Vehicle Speed Sensor that's used on the non-ABS versions. So he ordered one of those and popped it into the trans...and it's didn't work. It turns out there's a small pin in one of the shafts of the transmission that triggers the VSS, and it's not there on the ABS cars. The hole is, but short of disassembling the transmission it's not an easy retrofit. So now he's going old-school and putting a Hall effect sensor reading off the driveshaft bolts. Lots of fun!
The big trailer with the cars on board will be leaving Grand Junction on Wednesday morning. Almost there...
entry 929 - tags: trip computer, alignment, suspension | | |
|
THE DIARY