[WIP] Renault Megane III EV Tech conversion

[arber333]

I decided to buy a previous made conversion and remodel it a bit to my specifications.
This is the car that i bought…. Renault Megane III 1.6 16V Grandtour already converted to LV about 60V

I will bring this conversion up to standard with Mazda MX3 i converted recently to Outlander components and my own VCU controller.

This is the blog for re-conversion;
https://evmeganetech.wordpress.com

Edit: I had to change the website a bit....

[arber333]

Front battery box mount

I will use multiple battery boxes. Each box will consist of 24S connected CALB cells. There will be 24S ZEVA BMS module in each. Also i will build in 10mm copper tube with alu plate to transfer heat/cold. This will be my thermal management for each box.

To mount the front battery box i will use the 30mm x 30mm angle on each side to keep the box lifted to correct height to let me to close the hood.

Under the box i prepared two C crossmembers with 3mm inox mounts.
I bolted height adjusting inox plates on the sides of C members. On the front its 30mm in height and on the back its 120mm so battery box is canted towards the front a bit (20°).

Battery box will be fitted by M6 bolts from the front and the back. I fit the M6 nuts on inox plate from the rear so i can effectively assemble/dissasemble the battery box.

First i fitted the battery box on to the metal bands. From the rear i marked the hols for drilling.
For the second mount i had to go under the bonnet to mark the holes

I decided to tack weld nuts behind inox plates to make assembly and dissassembly easier. From the front i will simply fit the bolts through the holes and tighten them.

[arber333]

Driveshafts

To use Outlander motor and gearbox I needed to splice Renault Megane outer driveshaft parts to each of the Outlander inner tripod parts.
This was made on a lathe since this is the best way to balance the end product, however it was also the diameter difference between the slimer Outlander shafts and thicker Megane outer end.

We first thinned out Megane shafts to fit along the limits of Outlander diameter. Then we tack welded both pieces and prepared surface for C45 steel sleeve to fit over both shaft ends. This sleeve was cut in the middle to provide more welding area and provide homogenous effect to the splicing… as theory states if material is welded in 90°phase consecutively it will act as homogenous material.

Drivashaft ends were greased and fitted with new rubber boots.
I needed to get new snap rings to secure the tripod side of halfh shaft. I lost one when taking the thing apart and it seems normal to order a pair.

A company that works with driveshafts got me a pair of snap rings so i could assemble driveshafts to fit them to gearbox.
I see snap rings can be ordered from Aliexpress as well. You just have to order 20!
https://www.aliexpress.com/item/1005006 ... 18029RcnBf
Also i ordered new Oil and oil seal as it is very normal to damage the seals when inserting the shafts.

Gearbox oil i use is ATF SP III Mannol MN8209-1
Gearbox shaft oil seal P/N is 3515A056 (21293)
Gearbox fill plug oil gasket is P/N MF660036 (21286S)

I fitted the halfh shafts into one side and then the other. I removed the bolts from under each wheel knucle to remove the wheel to the side to fit driveshaft splines. On the other side i needed to use some sticky grease to stick the snap ring so i could fit the splines into gerarbox simply by moving the shaft longitudinally and shock of tripod hitting its stop would lock the snap rings across threshold. You need to be very carefull not to damage oil lip seals. This is why i ordered spares.

Driveshafts gave me a scare… when i assembled them i tested the length at tripod open and closed and it checked out. But when i inserted the driveshafts i mistakenly torqued the main wheel nuts. This fixed the outer part of the shafts. When i tested for shaft movement i got none!!! Did i made shafts too long???!!! After some time to calm down i thought to untorque the main nut and voila… shafts were able to move out and in with room to spare…DUH!
Of course they were good, i triple measured before cutting.

[johu]

Nice work!
Any road worthiness tests required? Since it's already EV and you just change small details

[arber333]

Nice work!
Any road worthiness tests required? Since it's already EV and you just change small details

Might be because I replaced the engine, there is no transmission anymore and voltage is changed a bit... no biggie .

If I can't do it here I will go to some other EU country...

[arber333]

To successfuly start the car and for every aux item in the car to work,, RPM signal is needed. In the case of 1.6 16V ICE engine this signal is picked up at the transmission flywheel. On the flywheel is toothed belt with 60 teeth pattern. This pattern is broken by two teeth missing at the RPM count starting point. This helps determine the 0 position of the flywheel and the engine Top dead Center.
This particular configuration is called 60-2-2. 60 imaginary pulses with 56 actual pulses and 2 periods of gap and 2 periods of long pulse.

How this works you can see here: https://www.tekniwiki.com/overview/how- ... t-sensors/

For our use we only need about 800RPM to incite ECU that engine is running. This will be done in code and transmitted on the same lines via PWM signal from VCU.

The code has been prepared so that:

Teensy 4.1 Arduino code to simulate a 60-2-2 crank wheel pattern using the Metro library.

57 evenly spaced pulses (1-tooth duration),

1 gap (1-tooth duration),

1 longer pulse (2-teeth duration),

1 gap (2-teeth duration),

Then repeat.

#include <Arduino.h>
#include <Metro.h> // Metro timer library
#include <FlexCAN_T4.h> // Teensy CAN library

const int tachoPin = 5; // Output pin for crank signal
const int teethCount = 60; // Total number of teeth (including missing ones)
const float rpm = 800.0; // Target engine RPM

const float revPerSec = rpm / 60.0;
const float revPeriodMs = 1000.0 / revPerSec; // 1 revolution duration in ms
const float toothPeriodMs = revPeriodMs / teethCount; // Time per tooth in ms

volatile int currentTooth = 0;
Metro tachoMetro(toothPeriodMs);

void setup() {
pinMode(tachoPin, OUTPUT);
digitalWrite(tachoPin, LOW);

Serial.begin(115200);

Can0.begin();
Can0.setBaudRate(500000);
Can0.setMaxMB(16);
Can0.enableMBInterrupts();
Can0.mailboxStatus();
}

void loop() {
Can0.events(); // Handle CAN events
tacho(); // Simulate crank pulse pattern
}

void tacho() {
if (tachoMetro.check()) {
// Standard 57 pulses
if (currentTooth < 57) {
digitalWrite(tachoPin, HIGH);
delayMicroseconds(100); // 100 µs pulse
digitalWrite(tachoPin, LOW);
}
// Tooth 57: 1-tooth gap — no pulse
else if (currentTooth == 57) {
// No pulse: simulate 1-tooth gap
}
// Tooth 58: single long pulse (2-tooth duration)
else if (currentTooth == 58) {
digitalWrite(tachoPin, HIGH);
delayMicroseconds(200); // Double pulse duration to simulate 2 teeth
digitalWrite(tachoPin, LOW);
}
// Tooth 59: 2-tooth gap — no pulse
else if (currentTooth == 59) {
// No pulse: simulate 2-tooth gap
}

currentTooth++;
if (currentTooth >= teethCount) {
currentTooth = 0; // R


I get a good 57 pulses per one "rotation" with a missing gap where the cycle repeats. Now i use a simple transistor module with 2.7kohm pullup so it will amplify signal to 12V.
Since the inductive sensor reads signal passing below 0V I needed to add one 1uf film cap in series which inverts the returning signal so what will be seen by the car will not be square wave, but rather pointed signal that goes down through the 0V...

4.png (79.02 KiB) Viewed 1749 times

This will give the ECU the appearance of a 60-2-2 wheel, synchronized to 800 RPM.

Connection to the former RPM connector... Since the sensor is induction type receiver needs to see a signal passing through zero to count it. My 50% PWM signal will not move the needle directly.

I solved the problem to cross the 0V line to trigger the count. I simply added one 1uF film cap in series with signal (!).

image.png (24.52 KiB) Viewed 647 times

First i tried to connect RPM signal to one wire of the connector and the second one to GND. It moved the needle a little, but not enough. Then i simply removed one side of connector and car came to life.