[info] VW Aircooled 1641CC engine

[mikedjames]

I am setting off on the aircooled EFI path choosing a MicroRusEFI ECU.


At the moment my 1975 VW T2 camper van runs a wasted spark crank fire ignition system with a 36-1 trigger wheel driving an EDIS-4 module.

The ECU is an STM32F103RB Blue Pill. It runs custom firmware that I wrote in C/C++

https://www.hambleonline.co.uk/websvn/listing.php?repname=EDIS_Ignition&path=%2Ftrunk%2F&#a0b40758157c8f16fa703ca3be466fa8a


It relies on other software modules I wrote over the years.

In fact it began life as a rev counter add- on for a PLX Devices MFD..

I have rebuilt two engines so far, a VW aircooled for the camper van and a Yanmar single cylinder diesel.

I did start assembling a Bosch CIS mechanical FI for the engine but realised that there is increasing trouble getting spares. I had it all running on the bench but every time I went to fit it in the camper, some huge rust repair project would take over..

At the moment I am hoping to keep costs down, using ITBs off a Honda VFR800 , and manifolds intended for IDF carburettors.

[AndreyB]

Nice custom ECU!

Welcome to the forum, most of activity is on the discord these days

[mikedjames]

Going loco down in .. .Hamble

I have been spending some time trying to get just the ignition system to run, using my existing timing wheel, and learning how to use TunerStudio, and some issues with earth loops causing watchdog timeouts when it did fire, not realising I had to remove the STM32 UART driver from Windows so that it could see the HID serial device and the mass storage device for the logs.

And it would not run, just fire once maybe. Could alter the timing and make it backfire or stop the starter motor by too much advance, by playing with the zero point and the static timing advance.

After a while I got to some of the settings in TS and managed to look at the engine timing phase ramp. Which was ramping 240 degrees and then stopping, resetting a while later and repeating.

What it turned out to be was that the VR signal from the timing wheel was good enough for a Ford EDIS-4 but was still pretty nasty.

The wheel was offset just enough on the end of a stack of hardware to be swinging about 0.5mm - the UAEFI high speed input would count 24 teeth then lose 12 every revolution as the signal got too small. .

I put my trusty Fluke Scopemeter on the VR signal and there it was swinging away - over a volt at the missing tooth but going down to 300mV at the low point.

I swapped to the alternate VR input and that did not work either, counting a different but wrong number of pulses for part of the rotation of the crank.

I am going to centre up the wheel, and have another go soon.

[mikedjames]

Well, that was embarrasing.

In an effort to space the timing wheel away from the engine, I had removed it, and replaced it - It has a pair of alignment bolts exactly 180 degrees apart, so I put it on 180 degrees out. So every time the engine wanted to fire, it fired the wrong pair of cylinders on the wrong part of the timing cycle.

As well as having a low output from the VR sensor.

That low output was fixed by the added spacer that allowed the centre of the tip of the sensor to run over the centre teeth of the timing wheel, and reducing the gap of the sensor by resting it on a piece of cardboard on top of the timing wheel while adjusting the fixing bolts.

I also found as a result of picking away at the end of a damaged Ford Fiesta VR sensor (where it was hit by the timing wheel for a while) that it uses a (smashed) 7mm button magnet on the tip of the sensor.
This was replaced with a stack of 5mmx1mm neodymium magnets, effectively reducing the diameter of the tip of the sensor and sharpening up its response. I epoxied the magnets in to place.

It also showed me that these Ford sensors have no directionality to their operation - the sensitive area of the tip of the sensor is a circle.

And finally the engine started.

But not before the carburettor filled the exhaust with so much unburnt fuel that I got a jet of flame about 2 feet long coming out of the exhaust when it fired.

Tomorrow I will get out the timing gun and start going through the setup and tuning - I took a video of the engine running for the first time using the UaEFI as the ignition controller.