Removal of Multicontroller a.k.a. Wonderwheel

[BlueBird]

How it should be.

I removed my wonderwheel and replaced the grip with the "ordinary" grip today,

can reach the indicators and operate the clutch at the same time now

 

[roamer]

Was the grip the same pattern, good to know it works ok,

 

[BlueBird]

Was the grip the same pattern, good to know it works ok,

oops,
Sorry Mucka, yes its the same and works a treat, ill post some pics of it tomorrow.

 

[fred_jb]

I removed my wonderwheel and replaced the grip with the "ordinary" grip today,

can reach the indicators and operate the clutch at the same time now

Excellent

Did you manage to get the dealer to provide the grip?

 

[GS_or_bust]

OK - I have now hooked up a couple of multimeters to the wheel rotation outputs and have confirmed that it is a 90 degree out of phase switched output, with a detent on the wheel every 90 degrees. Unfortunately the output is 12V. This would complicate using a 5V encoder as I would need a 5V supply and probably something like a couple of optocouplers to translate the outputs to the 12V pulses expected by the bike. I am also a little confused by the terminology on the specs of alternative encoders when looking for a replacement. The specs refer to the number of detents versus the number of pulses but don't define a pulse. The BMW wheel outputs three complete positive pulses per rotation but the detent means the two waveforms are effectively sampled 4 times per cycle and 12 times per rotation. I'm not sure how that would correspond to alternative encoders so any input would be welcome.
Fred

The reason there there are 12 detentes for 3 pulses is because quadrature encoding allows you to determine four positions from the one pulse (on each channel) using the edges of the pulses on the two channels. When turning the wheel, channel A will rise then channel B will rise then channel A will fall then channel B will fall. It is possible to determine the direction of rotation by whether channel A rises before channel B or B before A (using a software state machine). The replacement controllers on the Farnell page have too few detentes, if you buy a 24 ppr 24 detentes then each time you move one detente you will zoom 4 times rather than just once. 12 ppr 24 detentes would zoom 2 time per detente. I think you are right in thinking that using the existing wheel is the easiest, perhaps 3d printing mounting parts would be a possibility.

An alternative would be to program a micro-controller to turn an up down switch into the quadrature pulse output. That way you could use a four way switch such as this one so left right would simulate the wheel micro switches and up down would simulate one click of the wheel in the corresponding direction.

 

[ReGenesis]

Stop dicking around and just fit an extension to the indicator if your hands are that small.

 

[fred_jb]

Stop dicking around and just fit an extension to the indicator if your hands are that small.

Nah - that would be too easy, and probably too ugly, and to me would just be a bodgy workaround to an ergonomically poor design.

Anyway, I've always been interested in design, and as a project I'm more interested in producing something better than the original, and yes I am confident/arrogant enough to believe I can!

Fred

 

[fred_jb]

The reason there there are 12 detentes for 3 pulses is because quadrature encoding allows you to determine four positions from the one pulse (on each channel) using the edges of the pulses on the two channels. When turning the wheel, channel A will rise then channel B will rise then channel A will fall then channel B will fall. It is possible to determine the direction of rotation by whether channel A rises before channel B or B before A (using a software state machine). The replacement controllers on the Farnell page have too few detentes, if you buy a 24 ppr 24 detentes then each time you move one detente you will zoom 4 times rather than just once. 12 ppr 24 detentes would zoom 2 time per detente. I think you are right in thinking that using the existing wheel is the easiest, perhaps 3d printing mounting parts would be a possibility.

An alternative would be to program a micro-controller to turn an up down switch into the quadrature pulse output. That way you could use a four way switch such as this one so left right would simulate the wheel micro switches and up down would simulate one click of the wheel in the corresponding direction.

Thanks for the explanation - I suspected it was something like that.

I am currently working on fitting the existing wheel inboard of the switch cluster, and I'm making good progress. I have mechanically modified (hacked some bits off!) and rewired my spare wheel and just need to try a couple of ideas for fixing it in position. Will post a few pics once it is done. I feel that the indicator switch will then fall easily to hand, and that it is the multicontroller that will take a little more effort to reach, which to my mind is the way round it should be.

Interesting suggestion regarding using simple switches (which is what I hoped to do originally) - but then converting their output to a pair of quadrature waveforms, so that the bike's electronics can effectively convert them back into simple switch presses!

I still really like the idea of a small four direction switch/navipad and wonder why BMW didn't do this in the first place, though there is not enough space on the left switch cluster to incorporate this without moving other controls elsewhere, or making the cluster bigger.

Maybe it was just thought that a wheel design looks cooler, but having completely dismantled it and seen the level of engineering effort that has obviously gone into it, it seems ridiculously complex and expensive for something that is ultimately just emulating a simple up/down switch pair, while incorporating an actual switch pair for detecting the left/right sideways tilt.

Fred

 

[fred_jb]

Some pictures of before and after access to indicator switch, plus a couple of pictures of the modified wheel showing how I am intending to mount it the other side of the switches. It will be less easily accessible there, but I mainly use it while stationary anyway, just occasionally zooming in or out while moving, and I can choose my moment if I need to reposition my hand to use it, unlike the indicators where use is dictated by road conditions.

I've trial fitted it to the bike, and it all works electrically, but I need to mount to the bar such that it won't move and won't rock when I try to operate the left/right tilt function, as it is not a tight fit on the bars. I'm going to replace the three screws which would normally attach it to the grip with three nylon nuts and bolts. These will hold all the parts together, particularly the modified moulding on the inboard side of the wheel which operates one of the left/right switches, and the nuts should stick out enough to make it a very tight fit on the bar.

Fred

 

[fatnfast]

I like that set up Fred, .

I occasionally operate the wheel and assume I have put my indicators on, especially on right turns.
It's a shame that the Nav5 can't be reprogrammed to have an indicator function where pushing the wheel left/right turns them on.

 

[mrtee43]

Your next complaint if you manage to fit the wheel in that inboard position will, "every time I go to operate the wheel in and out, the bloody indicators come on !

 

[fred_jb]

Your next complaint if you manage to fit the wheel in that inboard position will, "every time I go to operate the wheel in and out, the bloody indicators come on !

Haha - yes that did occur to me!

 

[Wapping]

On the GS bikes, is there no way to 'lock' the wheel so that touching it has no effect?

On the 1600 you can simply turn the wheel on and off.

 

[GS_or_bust]

Since I have an old Garmin 660 without the BMW integration my wheel is of no use so I have been experimenting with trying to get it to control the music on my phone via Bluetooth. Controlling the volume and next/previous track.

 

[fred_jb]

Since I have an old Garmin 660 without the BMW integration my wheel is of no use so I have been experimenting with trying to get it to control the music on my phone via Bluetooth. Controlling the volume and next/previous track.

That sounds pretty complicated. I would be interested to hear how you get on. Are you tapping into the wiring to get raw input from the wheel, or are you pulling the data off the data bus?

 

[GS_or_bust]

That sounds pretty complicated. I would be interested to hear how you get on. Are you tapping into the wiring to get raw input from the wheel, or are you pulling the data off the data bus?

I'm tapping into the Linbus at the connection to the GPS cradle.