1 2 3 4 5
e46potenza
e46potenza Reader
4/17/15 6:10 p.m.

Jesus dude impressive work. I'm looking at doing some front end upgrades myself but nothing this intense

series8217
series8217 Reader
5/8/15 4:16 p.m.

Pretty good weekend overall but the car seems to be losing a significant amount of power as it gets hot. My fastest lap was the first hot lap of the entire weekend, despite not having driven the car for 6 months and having a bit of lap traffic. By the end of the weekend my friend's S2K was easily walking away. My fastest lap for each session was always one of the first two or three, with each subsequent lap dropping by a few seconds each. I thought the issue might be due to tire pressures.. I'm currently running the Hankook RS3s at ~37 psi hot. Perhaps they like a lower pressure and my lap times are going up because the tire pressure is rising too high. I'm going to try lower tire pressures at the next event to see if that might be the problem. However, the sidewall flex I get from these tires is not too encouraging to lower pressures... refer to pic below. I'll probably book a dyno test soon so I can see if my motor is doing OK (I think it's not).

series8217
series8217 Reader
3/22/24 12:39 a.m.

Hello GRMers, it has been a long time! Brief update leading up to 2022 before I get into the really really good stuff: In 2016 my clutch pressure plate failed during a track day at Buttonwillow and I limped the car to my friend's shop in Tehachapi. Then I had to move from my rental house (the owners decided to move back in), and didn't have a garage to work on the car. Having to drive 2 hours to see my car, combined with scope creep (rewiring the firewall bulkhead, and developing a new cooling system), it took awhile to get the car back together again. Then I bought a house, started a job that had me traveling 30% of the year, and well, yeah. tl;dr: life happened.

About a year ago I really started getting the drive to work on the car again. I dusted off my old CAD work for front uprights, started scanning a bunch of parts using photogrammetry, eventually bought a 3D scanner, built a 3D printer, and got into a really good workflow with a focus on durability, driveability, and big MFing tires.

I'll be spreading out the updates over the next few posts, so stay tuned. For today, we start with the first step toward getting the Fiero ready again for track duty: front wheel bearings.

The 1988 Fiero has a unique bolt-on bearing/hub cartridge that isn't used in any other vehicle. It has the lowest flange-to-flange offset of any bearing package made for a production car, and the knuckle bore is tinier than most rear bearings on FWD cars. The OEM 1988 Fiero bearings were long ago discontinued by GM, and various aftermarket replacements have been attempted, with some successful enough to survive a single weekend at the track on OK tires, but nothing better than that. Even the best suffered from occasional flange separation. Not for track use.

I've worked through a number of designs over the past 15 years to adapt larger more durable hubs to the stock knuckles, but never got anywhere I was happy with. Finally I decided it was time to build a new knuckle.

First I needed reference geometry. I scanned the stock Fiero front knuckle by taking a lot of photos and running them through RealityCapture, which generates a 3D mesh from those images. This technique can be as accurate as a 3D scanner, but it does take quite a bit more work as you have to manually add control points to relate images to each other.

1988 Fiero Knuckle photogrammetry

Next I designed a new knuckle that retained the same hard points as the Fiero but took a new hub/bearing package and brakes. I decided to standardize on modern GM because there are so many interchangeable rotor+caliper options with the same offsets and mounting patterns. With a single knuckle I could go swap between little W-body sliding caliper brakes up to 14+" CTS-V 6-piston Brembos, and even C7 ZR1 Carbon-Ceramic brakes.

Initially I based on my design on the 3-bolt pattern hub that allows for interchange between C5/6/C7 (5x4.75), Camaro/CTS (5x120), and Chrysler 300M (5x114.3). Below are some 3d-printed prototypes with this design.

Unfortunately the 3 bolt pattern is very hard to package within the Fiero ball joint and tie rod boss locations. The swaybar endlink also wants to interfere with the bolt bosses. No matter how I clocked it, something got in the way. Then I discovered the C8 Corvette front hub, which uses a 4 bolt pattern but is otherwise the same as the 6th-gen Camaro hub, including the 5x120 bolt pattern. This meant no more 5x4.75 or 5x114.3 option, but there are PLENTY of wheels for BMWs that will work with a simple hubcentric ring. The same brake interchange would work, although rotor fitment becomes a bit restricted due to GM putting a step inside the hat that's just a few mm smaller than the 5x120 hub flanges.

Here is a size comparison between the 1988 Fiero front bearing unit and the C8:

The image below shows the C8 hub mounted to a 3D printed ABS knuckle for test fitment. I also at this point in the development moved to a modular knuckle with a bolt-on steering arm. This allows me to use the same knuckle on the left and right sides of the car, and fabricate different steering arms to adjust for bump steer, ackerman, and steering ratio.

 

After a few iterations of 3D-printed prototypes with tweaks including adjusting to the caliper angle and overall dimensions for clearing other parts at full lock both ways and full compression, I created a manufacturing drawing package, crossed my fingers, and pulled the trigger on CNC machined parts. To be continued...

series8217
series8217 Reader
3/22/24 1:03 a.m.

It would take at least 6 weeks to get the CNC machined knuckles, but that wasn't all I needed. I had only purchased a single hub, rotor, and caliper for test fitment. Now I needed another of each. I also needed to fabricate bolt-on steering arms, some little heat shields I designed to bolt onto the new knuckle and protect the balljoints, and much much more: hardware, brake pads, steering rod ends, tie rod adapters, rod end spacers, ball joints, ball joint sleeves for the lower control arms, and upper control arm ball joint plates. I also needed to fabricate steel ball joint taper inserts that I determined were necessary for the LBJ to spread the load in the aluminum knuckle.

I had settled on using K6145 lower ball joints which are bigger than the stock Fiero ones and are commonly available with different stud heights and friction levels. For the uppers I'm using C4 ball joints which bolt into an off the shelf adjustable upper control arm plate from SPC, and have the right taper pitch and length to work with an aluminum upright. This will make replacement parts easy to find, and I know they'll be available for a long time.

After a few weeks of sourcing and ordering, everything started to show up. It was looking like a complete kit minus the knuckles themselves!

I picked up a lathe during the pandemic, which came in very handy for cutting these knuckle ball joint taper inserts.

These spread the load to a larger surface area of the aluminum knuckle to prevent it deforming and loosening up under high loads.

series8217
series8217 Reader
3/22/24 1:34 a.m.

Finally, CNC-machined aluminum uprights!!! I very very very carefully inspected them, checking all the critical dimensions against my drawings, and they were perfect save for a minor cosmetic dent.

It was so satisfying to feel my design in solid metal. The 3D-printed ABS prototypes were cooler than CAD, but they didn't feel real enough. The real thing feels SUBSTANTIAL. And everything fit! In under an hour I had fully assembled an upright with all its accoutrements.

Here you can really see the difference between the stock knuckle and bearing and the new one. Yes, the weight gain is non-trivial (16 lbs per side with C8 one-piece 320x30mm rotors and C7 calipers) but 80% of that is from the brake upgrade. Two-piece rotors save 5 lbs per side, and lightweight race caliper packages for the C8 and some C7 will fit.

Finally, mounted on the car.

Upper ball joint mounted in the correct orientation in this shot unlike above blush

This is the C8 front rotor with a C7 front caliper. The C8 front caliper would also fit, as I 3D scanned a C8 front knuckle to confirm the geometry.

The next upgrade would be a two-piece rotor to drop some weight, and perhaps also the Z51 package rotor which is larger and uses a similar caliper with longer mounting lugs to bolt to the same knuckle.

You might have noticed the brake line dangling in the above photos. I love little projects that I can solve with laser-cut and bent parts, because I can quickly design and make them with a service like Send Cut Send or OSH Cut.

I whipped up a design and 7 days later had parts in hand. I tried out Send Cut Send's hardware insertion service this time. That's an M6 insert pre-installed in the bracket.

After test-fitting I spray-painted it black and bolted it in. Perfect!

Next I needed to finish the other side, then do something about the rear bearings. Despite my upgrade to J-body bearings in the back they start to get play in them after a few track days. More soon...

DrMikeCSI
DrMikeCSI Reader
3/22/24 5:27 a.m.

This is some next level work!

Somebeach (Forum Supporter)
Somebeach (Forum Supporter) Dork
3/22/24 10:01 a.m.

Awesome project. Great job on all the custom work. I just looked at the recent update, I will have to sit down and read through the whole build. 

iansane
iansane GRM+ Memberand Dork
3/22/24 10:59 a.m.

That's beautiful. I really like the modular nature.

tuna55
tuna55 MegaDork
3/22/24 3:58 p.m.

Wow. That was worth waiting for.

series8217
series8217 Reader
3/23/24 3:14 p.m.

With the front hubs and brakes upgraded to C8 units, it was time to address the rear. I had previously bored the stock rear knuckles out to accept J-body front hubs, which are a bit larger than the stock Fiero units, and still available new. Unfortunately they weren't enough to keep up with 200-treadwear 275s and 1600 lbs of weight on the rear axle. It would also help with wheel selection if I had the same bolt pattern front and rear. I settled on Camaro5/6 5x120 hubs which use a 3x116m mounting flange with a 91.5mm bore. The mounting pattern is shared with C5/6/C7/300M/Camaro5/ATS/W-body/etc. The hubs I selected accept a large 33T spline, making them compatible with various Porsche 930 style CV axle stubs as well as some W-body outer CV joints that accept my stock Fiero shafts.

The reason I didn't use the C7 rear hub is because it has a 5x4.75 bolt pattern, 12mm studs, and a different size brake and wheel pilot than my front C8 hubs. Would rear C8 hubs have worked? Maybe, but the 4-bolt pattern would have been harder to package around the strut mount and suspension link mounting points.

The Camaro5/6 rear Brembo calipers are identical to the C7 rear Brembos save for the paint colors. Plenty of options for calipers from the GM parts bin.

I started my design by 3D scanning the 1988 Fiero rear upright, C7 rear knuckle, and Camaro5 rear knuckle. From the scans I could figure out where I needed to place the hub mounting surface to maintain the stock Fiero wheel offset. I also wanted to make the drum-in-hat parking brake functional, so I scanned the C7 backing plate separately to get a good view of both sides. The Camaro parking brake shoes are the same, but it has a different cable arrangement that looked more difficult to package than the C7 parts, so I decided to design for compatibility with the C7 backing plate and cables.

 

 

 

You might also notice the caliper for the Fiero and Camaro is trailing instead of leading. GM gets away with using the same calipers between the Corvette and Camaro by just flipping them left to right... and not using staggered piston sizes.

The rear upright is a much more substantial piece than the fronts. Rear axle weight on my car is 1600 lbs (vs 1250 for the front), and the size is much larger due to the tall mounting boss for the strut. Because of this, I chose to go with a fabricated steel design for the rear. Machining this from billet is possible, but it would have to start from a massive block of material, and require a lot of machining time. Steel is also a lot easier to rework if there are any fitment issues.

This image shows the new upright with the stock 1988 Fiero rear upright overlayed on the model. The major differences you can see here are the lowered strut mount, the caliper bracket moved from trailing to leading, and all of the multilink attachment points moved downward by 45mm to improve camber gain and roll center characteristics. This effectively results in it being a drop knuckle.
 

It's not the most efficient design in terms of weight but I wanted to get to usable part quickly as I was really eager to drive the car, and the parts are relatively cheap to iterate on compared to the billet aluminum front knuckles, which I had to get right the first time. I did run some simulations with traction, braking, and cornering loads and it's way overbuilt.

In a future iteration I'll reduce the weight of the strut mount and use a double shear clevis for the trailing link instead of that big heavy spacer at the bottom. I can also thin out a lot of the overall structure.

I was happy enough with the design so I created the cut and bend drawings, sent them off to Send Cut Send, and went on a work trip for a week.

 

series8217
series8217 Reader
3/23/24 3:31 p.m.

Got home from my work trip to a pile of boxes sitting by the garage.. including one containing these parts.

Everything looked pretty good out of the box. I had to tweak a couple bend angles but it was easy enough to do with my big vise and a hammer.

I designed the assembly such that it's self aligning. Everything that defines the suspension geometry bolts together, then slots into the front plate that holds the hub and caliper.

NICE!

Since the parts locate against each other, the welding fixture is very simple -- it's just a plate that the upright bolts down to thorugh the hub, caliper, and trailing link holes. The strut mount bolts and the lateral link rod end spacers keep their respective pieces in place.

Painted and done!

... and successfully test fit onto the car.

I only needed to adapt a few parts -- I used stock 1988 Fiero front brake hoses instead of rears, and I swapped the outer CV joints out for early 2000's W-body CVs, which have the larger 33 tooth spline but have inner starts that accept the Fiero axle shafts.

The rod end lateral links are 100% off the shelf parts. I designed the pockets in the upright to have the correct width for the 5/8 to 12mm misalignment spacers I'm using, and the inboard side just needed a small shim/washer to fit in the Fiero subframe pockets.

So now I have big bad brakes and bearings on all four corners, and no wheels that fit....

series8217
series8217 Reader
3/23/24 3:43 p.m.

Logically, if I'm going through the trouble of putting heavy duty bearings on this car with big M14 studs, I should install the fattest tires that can fit... right?

After going back and forth on various levels of aggressiveness I finally decided to just F it. Go big or go home. I ordered 315 Nankangs for the rear and -- in proportion to the car's weight distribution -- 265s for the front.

Apex Wheels has a great selection of meaty rims in Camaro and BMW fitments -- perfect for my new 5x120 hubs. I picked out satin black SM-10s in 18x9.5 ET35 for the front and 18x11 ET35 for the rear. The rears are Camaro fitment so they didn't even need hubcentric rings, while the fronts have a small ring to adapt BMW centerbore to the GM hubs.

Happy delivery day!

The next day a mobile tire installer came out and by sunset I had the car back on the ground with all the meats.

Yes, the rears are WAY too close to the fender even with my 500 lb/in springs, but I had to take some pics before I raised it up and lengthened the bump stop.

Believe it or not the fronts were fine like this, just some minor rubbing on the plastic fender liner under hard braking.

MadScientistMatt
MadScientistMatt UltimaDork
3/23/24 5:33 p.m.

DIYing your spindles? Wow. Just wow.

jfryjfry
jfryjfry UltraDork
3/23/24 7:18 p.m.

I didn't realize you were in the Los Angeles area. What mobile tire shop did you use? My buddy owns zip tire and I'm trying to invest in it as he and the company are phenomenal and I was wondering if you happened to use him.

series8217
series8217 Reader
3/23/24 8:45 p.m.
jfryjfry said:

I didn't realize you were in the Los Angeles area. What mobile tire shop did you use? My buddy owns zip tire and I'm trying to invest in it as he and the company are phenomenal and I was wondering if you happened to use him.

Yep I used ZipTire. Other than being 3 hours late for the appointment they did a great job. The tech was very careful with my new wheels. I would hire them again.

series8217
series8217 Reader
3/25/24 11:18 p.m.

The overly-observant of you may have noticed something peculiar about my knuckles -- an extra feature that doesn't belong on a Fiero.

I'm talking about the one circled here:

 

Yes indeed, that is a wheel speed sensor boss.

In my absence from tracking the car, I almost abandoned the Fiero platform due to the lack of ABS and its propensity to lock the front right due to horrible weight balance. Many times have I flat-spotted a brand new front right tire going into Off-Ramp when running Buttonwillow clockwise.

Then I discovered MK60 retrofits... particularly the MK60e5 variant which has more detailed CAN output, proportional valves, and built-in pressure sensors. It is no coincidence that I used hubs with magnetic encoder rings that are directly compatible with E90 BMW wheel speed sensors.

Here is the E90 front wheel speed sensor installed on my custom front upright. In hindsight, I should have mirrored the sensor boss so that the cable lengths would be correct from left to right, but I made it work.

The rear uses --- you guessed -- a rear BMW E90 wheel speed sensor.

I routed the wheel speed sensors along the brake hoses using 3d-printed Nylon clips. I had to reprint them all after a week as I had skipped the annealing step, which turns out to be critical for preventing the material from creeping even under light loads. I reprinted in Nylon and then in Nylon+CF. So far it looks like the carbon fiber fill isn't necessary. The annealing is the important part. I annealed both the plain Nylon and Nylon+Carbon Fiber parts at 80*C for 6 hours. No more creep!

With the wheel speed sensors in place, I still had another accessory to install. The DSC sensorcluster is an inertial measurement unit with gyros and accelerometers for measuring braking and cornering forces, as well as the rotation rate of the car. This feeds into the MK60e5 brain so it can decide how much pressure should be applied to each corner of the car. The MK60e5 isn't just an ABS unit but also a DSC (dynamic stability control) system. I don't have the DSC portion working yet, but the DSC sensor still feeds into some of the ABS calculations, especially for dynamic front-to-rear biasing.
 

I hoped to mount the sensor under the front edge of the driver seat where it is on a stock E90, but my floor plan would have me kicking it with my shoe every time I got in and out of the car. Instead I located it on the centerline of the car directly in front of the shifter.

Unlike the MK60, the MK60e5 can actually have a custom position for the sensor coded in. I'll get into my custom coding a little later. Suffice to say: this won't mess up the ABS, because I told it what I did.

Now I had wheel speed sensors, an accelerometer, and a stock Fiero booster and master cylinder with the stock prop valve... and only one brake line going to the rear of the car. Ok, lots of work left. Let's get to it.

New brake line to the rear left caliper... Done.

 

3D scan the stock booster area, E90 booster & master and MK60E5 ABS/DSC... Done.

 

Design and fabricate an adapter for the brake booster.... Done.


Design and fabricate a bracket for the MK60e5... Done.

 

Connect the lines... Done.

Wiring and fuses.... Done.

Using a laptop with BMW software to bleed the brakes on a Fiero is a very odd situation to be in. But also... Done.

... and finally out in the sun ready for its maiden voyage with ABS.

Now you can see the extra height I had to add to keep the new tires off the bodywork. It's a little exaggerated here with the camera angle, and doesn't look as bad in person. Really need to get around to fender flares though...

Anyway, it looks sunny, but it had actually just cleared up after a long rain. Great conditions for testing ABS! I went for a short 15 min drive, testing full panic mode braking through puddles and sandy spots on the road --- no lock up and NO DRAMA. Wow. I was logging nearly 1.0g on wet dirty roads with cold tires and un-bedded pads from Autozone. Bananas!

series8217
series8217 Reader
3/25/24 11:21 p.m.

A few more notes about this setup:

  • The E90 M3 front and rear brake volumes are very close to the C7 brakes, so the master cylinder volume works out very well
  • In the E90, the master cylinder and booster are angled upward, so the reservoir is tilted. I used a 5 series brake fluid reservoir to keep it level with the ground in the Fiero. I had to tap the E90 master to accept the retaining screw for the E60 reservoir
  • The MK60E5 doesn't require any wake up signals from the CAN bus. It just works.

My custom programming for the MK60e5 is as follows. I entered these using NCS Dummy then coded with NCS Expert.

  • einspurmodell -> set to custom_esm
  • brake pad wear monitoring -> disabled
  • pre engine drag-torque control -> disabled
  • ready assist braking -> disabled
  • brake disc drying -> disabled
  • brake fading compensation -> disabled
  • yaw moment of inertia -> 0x7A (E87 4-cylinder?)
  • front lateral suspension/tire stiffness -> 6812 (N/mm?)
  • rear lateral suspension/tire stiffness -> 10397 (N/mm?)
  • steering angle characteristic curve -> E87 for all LWK parameters
  • distance from front axle to center of gravity: set to 1.34m
  • distance from the rear axle to center of gravity: set to 1.02m
  • vehicle mass: set to 1331kg
  • sensor cluster distance from front axle: set to 770mm
  • sensor cluster distance from vehicle center line: set to 0mm
iansane
iansane GRM+ Memberand Dork
3/26/24 10:02 a.m.

This is so cool. Very well done.

series8217
series8217 Reader
4/25/24 10:19 p.m.

Update for March/April 2024:

Another MAJOR update! Some autocrosses and track days, a new seat, completing the front half of my custom center console, a digital dash, and a fire suppression system.

I took the car out to a few autocrosses to shake down the new setup before hitting higher speeds at the track.

The car is doing great but I'm still dusting off my driving skills. Almost there. Been slowly creeping up from finishing last to almost on the podium.

No major issues in the shakedown runs -- just some suspension link jam nuts coming loose, camber adjustment slipping, and a blown fuse, things like that.

I let a very fast driver borrow my car out for a few runs and he loved it. Based on the times he was running my car could have won the event if he was driving instead of me laugh

20240224_153439.jpg

 

After I felt confident nothing major was going to break, I wrapped up the last few things necessary to get back out to a road course.

New Sparco Evo driver seat and new upholstery on the Corbeau A4 on the right side. I designed the seat mounts by mocking up the seat where I wanted it, then 3D scanning the seat and floor to get the relative positions. Send Cut Send lasered and bent the mounts, and I painted them black (no time for powder).

20240315_161028.jpg

 

I also added a new tow hook that bolts to the rear wing mounts, along with nicer wing mount hardware. No more shiny silver on this bad car...

I also added a LifeLine 360 Novec fire suppression system.

This also required completing Rev A of my totally custom center stack. If you've seen a Fiero interior before you will notice nothing is familiar here, because I designed everything but the dash. The center vent stack HVAC control surround, and Fire suppression button panel are 3D-printed in ABS. The center tunnel cover is aluminum wrapped in vinyl.

You might notice something else new behind the steering wheel... yep that's a digital dash. I'm using a Joying 10.25" head unit running RealDash, which connects to the CAN bus. I'll go into a LOT more detail on this in a later post, because there's a lot of custom work software work behind it.

I finished up the fire system install and got the digital dash working the same day I loaded my car onto a trailer to take to the track -- a weekend with NASA at Buttonwillow Raceway.
 

My first few sessions involved a lot of confidence building in the new car setup, but I spent some time in my simulator to make sure I wouldn't also have to re-learn the track. I was gradually getting faster and faster and about to break my personal best lap in the #13CW config.

Then as I was braking from the sweeper, I scanned right to find the exit and noticed a giant smoke screen in my rear view mirror. At first I thought it might be time to press the big red button, but the smoke went away as my speed went down.

I pulled off the track to find a steady stream of oil shooting out of the pan. At speed the airflow was blowing it onto the muffler and causing all the smoke.

At first I thought I must have blown the front main seal from crankcase pressure. Then I lifted up the right side of the car and poked my head under to find this....

The back side of an internal baffle spot weld had broken out causing a 3mm hole to appear in the pan.

With the car lifted to an angle, the oil level was below the hole so I blasted it with brake cleaner and JB welded a coin over the defect.

Look, it's the 1 yuan coin that Spectra saved by not spot welding the oil pan correctly!

A few hours later the JB Quick was set and... no more leak!

Unfortunately that's about when the sky started leaking instead.

I didn't get another dry run all weekend. Still, I deemed it a success. Nothing I built or designed broke, the car was getting faster and faster, the ABS worked amazing even in the wet, and I'm really motivated to continue taking this car to the next level.

jfryjfry
jfryjfry UltraDork
4/25/24 11:07 p.m.

Amazing update!   Some pics aren't coming through but it's still great. 
 

I'll be running sow with speed ventures on 5/12 if you're interested.  Black z06 #77

series8217
series8217 Reader
4/26/24 2:00 a.m.
jfryjfry said:

Amazing update!   Some pics aren't coming through but it's still great. 
 

I'll be running sow with speed ventures on 5/12 if you're interested.  Black z06 #77

Not sure what went wrong with the pics but I reuploaded and I think they're working now. I'm not a huge fan of SoW but Speed Ventures is a great group to run with! I'll be back at Buttonwillow in May or June. 

wyndscreen
wyndscreen New Reader
4/26/24 10:11 a.m.

Just a fantastic read!    Love the wrap, love the tech, love the fab work, and love the entire car!

Still have AC?

Plans to cage it?

Purple Frog
Purple Frog GRM+ Memberand Dork
4/26/24 10:14 a.m.

Very impressive build!  Wondering how much of the original unibody will need reinforcement due to the bigger loads you will be able to achieve with improved suspension and larger/stickier tires.

series8217
series8217 Reader
6/11/24 12:58 a.m.

My big project for April/May was to finally add some true datalogging capability to the car, and put a cherry on top of the interior.

I picked up a Joying Android head unit, and permanently installed a USB CAN interface inside that works with the RealDash app --- a professional quality digital dash engine that runs on Android, Linux, and iOS and is fully customizable and programmable to do pretty much anything you want. It's basically a 2D game engine designed specifically for car dashboards.

I designed and 3D printed mounts to place the head unit "brain box" directly on the steering column, and the display in front of it.

I originally tried to hide the head unit under the left speaker grill, but it was tricky to find enough room even after I had 3D scanned that whole area with the dashboard removed. That's ok, this will get covered with a shroud anyway.

Here you can see the head unit without the display.

20240428_210638.jpg

Since the dash communicates via CAN bus, which Fieros don't have, I needed to get all of my vehicle sensors onto CAN. For the MK60E5 ABS module, this was trivial -- it already outputs a plethora of data from its CAN interface, including:
- individual wheel speeds and brake pressure, so I can monitor brake bias in datalogs or in real time
- inline and lateral acceleration and yaw rate
- ABS control state (e.g in intervention or not)

For supplementary sensors such as water pressure, oil temperature, and fuel level, I'm using an AEM 30-2226 6 Channel CAN Module.

Engine data is a bit trickier. I'm still running the stock GM "9396" ECU from 1991, which is ancient and slow by today's standards... even by 1996 standards. This ECU does have a serial data output line. It's a single wire ALDL interface that operates at a baud rate of 8192. It's possible to connect this interface to a standard UART with just a few passive components -- a resistor and a diode to split the send/receive lines into the microcontroller. Using a Seeed Studio CANBedFD development board for it's CAN interface and automotive power supply, I implemented firmware for an ALDL-CAN bridge and connected my stock ECU to its UART. Now I have the full diagnostic data stream from the 1991 ECU on CAN so I can datalog it with pretty much anything -- including my dash.

You can find the ALDL-CAN bridge source code on my github: https://github.com/series8217/aldl-can-bridge

Very exciting to see live data streaming into the dash!

20240421_222525-dash-can.jpg

After I confirmed the electrical system was working and the display was in a good spot with real graphics running on it, I designed and 3D printed a shroud and sun shade. The print is a little rough as I only had time to print it with draft settings but I'll do a reprint this summer when my printer has some idle time.

20240523_203118-interior-dash.jpg

I used the Porsche 992 GT3 R technical manual as a reference for the gauge layout. I have multiple pages including the race display shown here, as well as a warmup page (monitoring engine vitals, oil temp, water temp, water pressure, etc), a track map, chassis diagnostics (individual wheel speeds and brake pressure), etc. The diagnostic pages are very helpful for making sure systems are working properly after repairing or replacing components.

20240520_185557-dash.jpg

With the CAN bus sorted and RealDash working, I could log performance and diagnostic data, and see a lot of real-time information that previously required a GM or BMW scan tool. However, due to the way the Android OS isolates files between applications, it's a bit of a pain to get data logs off of the dash. Also, once you get the data off it has to be converted to a format compatible with race analysis software in order to get insights for improving vehicle and driver performance. All of these steps waste valuable time between track session. Time better spent looking at the data or working on the car.

As fortune would have it, another racer friend was in town to visit the same weekend as a Speed District track day at Buttonwillow. So he flew down with more datalogging gear and race safety equipment than normal luggage, and we headed to the track to test them all. At one point we had 4 different datalogging systems and 3 cameras on in the car. By the end of the weekend, one prevailed... Stay tuned!

 

wyndscreen said:

Just a fantastic read!    Love the wrap, love the tech, love the fab work, and love the entire car!

Still have AC?

Plans to cage it?

 Thanks!! Yep I still have AC in the car, and it's registered and street legal. Even passes California smog checks!

Every time I've been to the track recently I've gotten closer to thinking it's time to gut and cage it. This year I finally decided I'm ready to have a stripped and caged Fiero in my life. I just haven't decided yet if it's going to be this car or if I'll get another to build up so I can keep operating my current car in the meantime.

 

Purple Frog said:

Very impressive build!  Wondering how much of the original unibody will need reinforcement due to the bigger loads you will be able to achieve with improved suspension and larger/stickier tires.

Thanks!

I've already had to box the subframe mounts as they were tearing out. The whole space frame could use seam welding as GM really skimped on the number of spot welds. There are just enough to hold the various components in place.

series8217
series8217 Reader
7/24/24 2:41 a.m.

Beating the Heat

I managed to squeeze in two more events before the full heat of summer arrived.

At the end of May, my friend Shane Reetz (@trackdailycrx) was in town to hang out for a few days and co-drive my Fiero at Buttonwillow Raceway during a Speed District track day. Shane's track car is a tastefully modified and upgraded 2nd gen Honda CR-X that he tracks & autocrosses nearly every weekend in the PNW. He's also a performance driving coach and has a good handle on using data to improve driver and vehicle performance.

 

buttonwillow-from-the-sky.pngButtonwillow Raceway Park from the sky, with the fresh black asphalt of Track 2 still curing in the foreground

 

Shane showed up with small bag of clothes and a giant bag of race gear and data logging equipment.

The event was Saturday morning, and we had Friday to prepare with a reasonably short checklist of items to complete:

  • Mount AIM SOLO 2 DL and configure CAN protocol for ALDL->CAN bridge and BMW MK60E5 ABS
  • Mount tablet for CircuitStorm app
  • Mount GoPro
  • Mount Android phone with RaceChrono Pro
  • Mount Qstarz Bluetooth GPS on the windshield (for RaceChrono)
  • Shakedown run up Angeles Crest Highway with all dataloggers on
  • Fill the fuel tank
  • Load the truck and put the car on the trailer

By around 9pm, we had tested all the loggers and were seeing reasonable data. The AIM Solo was recording everything from individual corner brake pressures, wheel speeds and accelerometer data, to engine coolant temp, oil temp, throttle position, and fuel level. The only really useful things missing are ride height sensors and steering angle, but I will be adding those soon.

After we got back from the Crest we filled the fuel tank, jacked up the front of the car to do one last check of the front brakes, and took a break for dinner. An hour later we returned to the garage and were assaulted with gasoline vapors. There was a giant puddle of fuel on the floor and a steady drip from the back of the fuel tank. 10pm...

tank-install.png

 

We got the back end of the car up in the air and quickly assessed the situation. The vent pipe on the back of the fuel tank was loose where it had been swaged/crimped by the factory. The only option to resolve it was to replace the tank. As luck would have it, I had purchased a new tank from Rodney Dickman Automotive Accessories last year to use for a future swap. By 1 AM the new tank was in and filled and the car was on the trailer.

loaded.png

 

After 4 hours of sleep, we were on our way to Buttonwillow.

Track Day


If you are lucky enough to have Speed District events within driving distance of your home, you probably know how exceptional these events are. Unlike the racing + HPDE orgs that cram track day activities in with lots of club racing, Speed District hosts HPDE-only events with limited car counts and give drivers so many sessions per day you will pack up and go home due to exhaustion before you're out of track time. It's a great environment for testing, tuning, and driver development. Attendees also tend to be very fast and very courteous, so you're likely to have a few unimpeded laps per session as getting stuck in traffic isn't a problem.

 

grid.png

 

It was going to be a long day and we had a lot to get done, so we loaded up on coffee and got to work.

The goal of the first session was a more thorough shakedown run for the chassis (the Crest drive was just to make sure the loggers and dash were working) and to get the tires up to temp, set hot pressures, and check brake temps. No problems there -- we bled 3 psi out to achieve a hot pressure of 29 psi, checked the rotor temps (they reached bedding temp), and confirmed the brake fluid was staying cool. No engine error codes, no ABS codes, all looked good on the car.

 

1988-fiero-track-car-series8217-steven-snyder-buttonwillow.png

 

As for the data systems, the AIM Solo 2 DL had recorded a full trace, RaceChrono was looking good, my dash was logging but CircuitStorm never even started recorded and didn't trigger the GoPro either.

 

1988-pontiac-fiero-track-car-data-acquisition.png

 

For the second session, Shane hopped into the passenger seat for his first ride around Buttonwillow in real life. He has set some great lap times in the simulator but some of the surface irregularities (especially the bumps/dips around Riverside and Sunset) are different. Still, a simulated Buttonwillow is accurate enough to serve as a refresher if you've been to the track before. Time in the sim also gives you a huge head start if you've never been to the track before, as there are multiple blind corners.

 

1988-fiero-track-car-series8217-buttonwillow-speed-district-1.png

 

We found the issue with CircuitStorm -- sort of. If the GoPro goes to sleep, CircuitStorm won't be able to start it on its own and complains about the camera being off. Since I couldn't see or reach the tablet from the driver seat in my first session, I didn't know what was going on. With Shane sitting shotgun we he turned it on for this run.

For the next few sessions, Shane and I traded off between passenger and driver duties. Air temp stayed under 90*F and we had no problems with oil or coolant temperatures. Shane adapted fairly quickly to the car and we were both setting reasonable lap times, but no personal bests and we were getting slower as the day went on. This is typical for sunny days as the track temperatures skyrocket by the afternoon and soft tires get greasy. However, neither of us could consistently brake for some of the corners on the track, and the steering alignment seemed to be getting progressively worse as the day went on.

So we dug into the data to try to figure out what was going on. The drawbacks of some of the systems became immediately clear, and I'll summarize here as there is really a clear winner.

CircuitStorm was too brittle and clunky to manage, and we couldn't get it to connect to the online Podium service which would have been a nice benefit. RaceChrono Pro is a nice tool if you're just looking to improve lap times and don't have a coach or need to develop the car. It's very easy to quickly compare laps and see what corners you need to work on. It also has a very useful lap timer display while you're in the car.


AIM Solo 2 DL and Race Studio 3

1988-fiero-track-car-aim-solo2-dl-race-studio-3.png

 

The AIM Solo 2 DL was the real star of the show. It started automatically when the car started moving, and recorded basically every CAN channel on the car at 10hz, plus it has built in GPS and accelerometers. When I pulled into the pit garage and turned off the car, the Solo 2 DL would stay on with its internal battery to allow downloading data to a laptop. Shane didn't even have to get up from his chair. The Solo 2 DL has a WiFi access point that you connect to from a laptop, and a few clicks later you have data loaded and visualized in a very professional tool.

The most immediately obvious issue was both of us were over-braking before most high speed corners, leaving a lot of cornering force and exit speed on the table. There were easily seconds to work off with improved technique. We were also inconsistent in braking into some turns -- particularly Off Ramp and Sunset -- and even overshooting the ideal turn-in point at times. It was something we noticed while driving but couldn't pin down what was going on.

It took some digging but I eventually found that in a few particular corners the brake pressure generated by the pedal (about 100 bar out of the master cylinder at peak force) didn't result in the same amount of longitudinal acceleration, and it wasn't due to ABS intervention. This meant the effective brake friction coefficient was going down. Sure enough, the thermal indicating paint on the brake rotors had turned a color that was literally off the charts of the operating temperature range for the racing brake pad compound I'm using.

The next event was two weeks away and with a guaranteed hotter air temperature, so brake ducting and improved braking technique would both be necessary. The first to get cool air to the center of the rotors so the cooling vanes can do their job, and the second to reduce the over-braking and the additional heat generation it causes.

There were more fixes to make too but I wouldn't discover those until the post-event teardown...

1 2 3 4 5

You'll need to log in to post.

Our Preferred Partners
fREpsRGyI4zc2SF8c6jmpFScDOwItIjL7ee4J6IwWfY5fJsWcAMhsXoqLuptwA1Y