No announcement yet.

It's a Chilly Bin Bro! XR4 Daily Novelty FiST

  • Filter
  • Time
  • Show
Clear All
new posts

    Originally posted by Aaron View Post
    It’s probably the best and worst time to buy one.
    - Things like the gearbox being made from unobtainium glass mean either spending $5k on another used gearbox, rebuild it, add the LSD or fabricate a MTX75 gearbox swap into the car.

    Work colleague got bitten by this. Bought one for his son to drive. 200k on the odo got it from a dealer for $5500, I think it had been a trade in. Day 16 of ownership whilst his nopics is driving their son to work she hears some nasty noises, drives straight home and parks it up, when he gets home he gives it a run and hears some nasty noises and then sees oil coming out the bottom and what he originally thought was a piece of engine block on the ground. But the tell tale diff gears rattled out when they loaded it on to a flatbed to go back to the dealer. So blew the diff and punched a hole through the bellhousing. Dealer originally tried to claim the car didn't come come with a warranty which he called bullshit on since as a dealer he has to sell with some form of warranty (unless it's sold unregistered I think) which for a car of that age and km's is 1 month. When he was discussing with the dealer what they were going to do to make it right the dealer tried to get him to sign backdated signup documents for one of those 3rd party warranty companies, colleague rightly told him to get fucked he wasn't getting involved in the dealer trying to commit insurance fraud by backdating paperwork.

    After a week of the dealer trying to stuff him around and trying to claim he had it on tape him telling them it had no warranty the dealer relented and money is back in colleagues account.
    Last edited by Luke352; 16-10-21, 04:28 PM.


      It was one of the (less important) reasons why we bought the Camry - having a high-mileage XR4 as a daily isn’t an issue right until the Gearbag pops. Over the next few months I’ll start either collecting Gearbag parts, although I have most already, or Gearbag swap to MTX75 parts which is a hodge lodge of mounts to fabricate, custom driveshafts and shifter stuff to sort.
      Shed Project: 1994 Laser Lynx with BP-T


        Originally posted by Aaron View Post
        Then I came across Rapid ( - they do an incredible range...
        Thanks I just ordered 250 metres of the striped combinations I couldn't find elsewhere.
        Jaguar XJR, Freelander 2 HSE, Jaguar XKR, MINI Cooper S
        Originally posted by nutttr
        People must assume you are some sort of drug dealer with all these nice cars turning up to a fibro home


          Got the DXFs for a decent looking set of MTX75 into XR4 mounts. Uses a combination of the XR4 factory upper mount, and a Mazda CX-7 rear mount. The latter is a larger version of the Fiesta design so once the swap is done there’s very little obvious change from stock.

          Now need to source Focus LS clutch, flywheel, concentric slave and Gearbag. Focus LR shifter cables. Hybrid driveshalfs (Focus LR inners in XR4 outers) and custom clutch line.

          Will give taller gearing especially in 5th which probably will make the car 11nty times nicer as a daily.

          I’ve also started work on the code and hardware setup to replicate the CANBUS messaging that the ECU needs when running without the ABS module present.

          The plan is to read a conventional Hall/Inductive vehicle speed sensor - in the case of a typical Clubbie that means reading pulses at the tailshaft flange.

          Initial steps were to trial a few calculation ideas and do cross comparison tests in Excel to look at accuracy and impact of false/missed reads etc.

          What I’ve settled on is the “Pulses per KM” method, a starting value can be mathed out from tyre size and diff ratio (8356 in this case). The computation is very simple maths so the Arduino code is easy to write and very robust in execution. I’ll have to write a user interface and add some hardware to allow for it to be calibrated without hacking the code. It looks like updating the roadspeed every 200ms is practical with the speed sensor where it is. If we move to a tone wheel setup in the same location instead of tailshaft bolts we can increase the refresh rate.

          I’m trying to stick to ADR speedo requirements in terms of accuracy and error over operating range.

          The Arduino will need to broadcast the 0x4b0 message every 20ms to keep the ECU happy, along with a 0x080 and 0x090 message (that the ABS/Stability system generates). The ECU converts the 0x4b0 message into roadspeed and rebroadcasts that for the cluster and OBD2 diagnostics. It also increments the odometer.

          We are going to run this standalone setup driving a Haltech IC7 Dash via CANBUS.

          All in all it’s a relatively simple exercise, code wise I’m just setting a couple of timer variables, one for the CANBUS broadcasting, another for the speed refresh.

          When elapsed time condition is met, the count of received speed sensor pulses is used to calculate the speed over that period. This sets the Vehicle Speed variable which is used by the CANBUS message code.

          Where things can get really cool is I can also integrate GPS based speed, and use that to act as the speed of the non-driven wheels. The 0x4b0 message is roadspeed for each of the 4 wheels. As the ECU is responsible for Traction Control we can actually bring that to the cutdown system without having to get individual wheel speed sensors.

          It would also be possible to use an accelerometer along with a steering angle sensor and start tuning a very rudimentary stability control system by influencing the wheel speed inputs. The Traction Control is broadly two stage, a torque limit, and a hard torque cut. By testing to determining the threshold ratio between the 3 stages (None, limit, cut) we can then effectively make the ECU feather power. Interesting ideas are interesting.

          Those of you who know arduino will be wondering how the fuck this is all going to fit in the memory space, let alone computationally fit in some of the timings…. Yeah, already hit memory space issues playing with the dash screen ideas. So now I use Cortex based hardware, way way quicker, native CANBUS and relatively shit-tons of memory.
          Shed Project: 1994 Laser Lynx with BP-T


            So the ABS emulations device is up to the testing phase, success so far with the message structure and conversion of input frequency to roadspeed maths.

            My choice of using a significantly more powerful STM processor based arduino is a problem. Doing simple shit like storing the PulsesPerKM setting in EEPROM or Flash Memory (so the user can adjust and save on the fly without a laptop) is a pain the dick, it seems finding functioning libraries that work universally on the hardware type and have documentation is like finding a virgin in a brothel.

            The flip side is that the device has the processing grunt to parallel process/multitask so I’m not bound to using interrupts in the sensor inputs, I simply put the CANBUS transmission in one process stream and run the sensor inputs separately preserving message timing and not consuming interrupts.

            In other news the maker of the highly specifically sized Diff bearing as stopped supplying them, so it’s taken a fair whack of other peoples effort but this is the new Diff-bearing pair to suit IB5 gearboxes.
            Shed Project: 1994 Laser Lynx with BP-T