Topic

The revised code has been uploaded to the repository and the link above will find it. It now works completely and I spent half an hour turning the ignition on, lowering the windows on the one touch then switching off again and operating the fob without any glitches. Hooray! I have a revised circuit diagram and have chosen a much cheaper pair of transistors for the power switch. The revised diagram is annotated with the connector and switch assignments from the MGF diagrams on Dieter's site to make the connections easy.

I shall continue to develop the software next month when I will be spending some more time with my cellar in France. A further revised diagram using a relay for the switching is in existence but I want to build it before publishing it. For the moment though this can be regarded as definitive and ready to construct.

The unit has been installed in my MGF. The hazard warning light switch has an unused pin position and the indicator wire is simply bared, poked through the hole in the socket and pushed into the green/white wire's receptacle. The electronics fit into the Rover one touch module box and it all works very nicely. I have ordered parts to start constructing another one for development purposes and will be ordering components for a PCB shortly once I have the design finalised. The circuit diagram is now replaced above.

Awesome.

When you say is now replaced above, where is that please?
I have checked the GitHub and can't see it. I just want to be sure I have all the parts now before I look to start construction.

Also looking at PCB design and a possibility of ordering them online.

Thanks

That is already in hand. I have received enough interest to commercialise the device so am at present creating PCB layouts. Much of the interest has come fromowners of another marque which expands the potential of the idea. I shall be spending the next month production engineering the module and making the first production batch.

Boards drawn. Components ordered. It is the details that make the difference between prototype and product. I managed to buy matching wires for all but the grey/ pink which will have to be pink. I shall be making a unit for a TF as I have one set of switch cables but this batch will be for the MK1 cars.

The software has now been improved to know where the glass stopped so that it can run the correct amount of time to destination without stalling the motor. The position at ignition switch off is memorised and there is a calibration routine to match it to the car so if one side is slower than the other, the timing will be correct. This allows it to be fitted to any car and dialled in to fit using the window switches.

Holding both switches in the up position while switching on the ignition will cause the module to go into learn mode. Press the left down button and release when the glass is down. Press it again and release as soon as the glass stops going up. The left side is calibrated. Repeat for the right down button to do the right hand side. The settings are now memorised.
This needs testing fully in the car, which will take a couple of days but it is working well on the bench so I think it is probably finished from the software design angle and now I want to refine the installation.

The circuit is still evolving. I have replaced the window relay and arranged to power the control circuit from both sides at once so that if one fuse blows, the other window will still operate. The relay is a fail safe so that if one window control relay welds in the on position, it will not flatten the battery.

The installation is complicated a little by this though. It is necessary to remove the relay socket from its bracket, take the front off it, remove the green wire with a ty rap to push down the latch and replace it with a twin wire before clipping the front back on, replacing the socket and plugging the relay back in. It isn't difficult but it is fiddly as access is pretty limited. The reason for this is my determination to avoid cutting any wires and to allow the car to be returned to originality if required. I have removed the fusebox and its bracket for access which I needed to do in order to find the best way to do the relay mod. Tomorrow I shall try to do the job whilst keeping it in situ. I think it is possible- I just need to make it easy!

Parts now arriving daily. The cable and crimps are here for the first ten modules, as are the relay boards and Arduini. Most of the PCB components, save for the board itself are here and the boxes are due in three weeks' time. I shall get started on the cable cutting and crimping this weekend now the circuit is finalised. The next project will be a module to give variable heater motor speed and just maybe temperature control for the heater if I can find a suitable electric valve. This will be integrated with an electronic boot release, potentially a windbreak lifter/lowerer and a delay for the coolant monitor sensor as the processor won't be very busy. Before then, however, I have some spheres to change...

Getting closer. The current sensing trip method works better than the elaborate software solution, switching the drive off in 80mS when the motor stalls. This gets around the necessity for calibration and will allow me to simplify the software next week while I'm in France. In the meantime, I have finalised the PCB and printed out the space model which fitted a treat These will now need etching and drilling, a job for when I get back.

A rather busy kitchen table in West France Tuesday. The wires are cut for each module and the modules themselves are largely complete, needing only the daughter boards and testing before final assembly into boxes and then there will be ten ready to go to work. I shall put one into my car and keep the second as a development hack so that modifications can be made and tested, should they become necessary or desirable. The other two will be used to prototype the TF and Mark two car systems and work out what will be required for the different wiring harnesses to make it plug and play.

So making progress. Great news.

Haha! That's what my table looks like when I'm making my low coolant alarms