Tilt/Bend lighting.

I've been asked to document my solution. I am not going to do a “the red wire goes to pin z” write up. If you do not have a basic understanding of electrics and angles – For god's sake find somebody who does.

Also this is a SIMPLE solution. It switches lights on and off – it does not make them track with motors. Imo to make a servo actuated system big enough to move decent sized lights – you'll might as well re-engineer the headlights.

You could have multiple sets of lights witched at different lean angles too – but I have no wish to look like a 1960's Lambretta. I did consider two sets I will admit.

Some of the numbers used in later configuration are very much based on how I ride – and also some legal guidelines outlined el ECE Reg053. Much of the inspiration comes from a University paper “Corner adapting motorcycle headlight - May 29, 2002 - Department of Mechanical, Industrial and Manufacturing Engineering College of Engineering, Northeastern University Boston, MA 02115”

The way I've done things is based on some existing conditions.

I already had the TRIUMPH Fog light kit wiring installed and was NOT prepared to remove it. This due to the need to be able to reverse engineer FAST and EASY for biannual testing. And also eventual sale of vehicle.

This meant I had to add things that I won't be documenting here which allow me to use the lights both as TILT as well as normal Aux lights (or in future iterations) allow me to have both activate on Main beam.

This was an inconvenience – more than a help. I had to add DIODES and all manner of electrickery to existing wiring.

I have a top-case fitted – this made sensor locating a doddle.

PARTS.

1) You need a configurable and adjustable sensor/switch assembly. There are lots of TILT switches out there – but most are MECHANICAL and the centripetal forces of cornering a bike will counter how they work.

So you need something that either has a built in Gyro or uses something external like the earth's magnetic field to provide a stable reference.

I happened upon Hummingbird Electronics in Australia who make this type of sensor for heavy mining equipment. They have several types of sensor – some are by design ROBUST in sealed metal boxes. Not BIG but waterproof and twit-resistant. My thought was – If this works in an ore mine – a bike is no hassle.

Note they do some boxes JUST sensors – careful. You need a switch.

The device I have is their usb-configurable-tilt-switch (link : http://hmbe.com.au/products/usb-configurable-tilt-switch) .

This has two switchable relays built in (one for each light). It has dual axis detection so you can mount in-line or to one side of the axis of tilt. It is configurable to 6 minutes of deflection (1/10 degree) . Has built in electronic “damping (adjustable)” and has hysteresis control (explained later) . It can be zeroed (in case you don't get the mounting EXCATLY perpendicular – and 1/10º is nothing.)

It's downside is the relays are relay designed to switch small alarms on/off – so one needs to beef that side of things up. And it's NOT CHEAP.

Hummingbird have units with bigger relays – but NOT as configurable.

They also have an even more robust sensor/switch and if you do a lot of riding on bad/rough roads you should consider the Gyro stabilized version. I had some email conversations with the mfr and he was sure that for me it would be unnecessary – but if you ride gravel etc – maybe…….

BUT – IF ANYBODY FINDS ANY OTHERS – PUBLISH THE LINKS.


2) LIGHTS. No really! – you'll need some extra lights!.

I used LED ECE/DOT certified LED Fog Lights. J.W.Speaker 6150. You can of course use any light you see fit – BUT as the idea is to GIVE YOU LIGHT TO SEE BY AT NIGHT AND WITHOUT ANNOYING ONCOMING TRAFFIC. This is not a “hello I'm here” lighting solution.

I did toy with the idea of using two additional DIP BEAM HEADLIGHTS – one UK one Europe – but was warned I'd fall foul of a homologation issue. Using AUXILIARY homologated lights removed the issue.

Imo – it makes absolutely NO SENSE at all to use generic lights with no dip cut off. This rules out a shed load of aux lights on the market.

3) Power relays.

Based on you choice of 1) and 2) you'll maybe need intermediate relays that can pass enough current to your lights without burning out tiny relays in the sensor. Personally I used TWO ROWE ELECTRONICS PDM60s – But I also replaced a load of other relays and fuses etc at the same time. I personally believe that using GROUND TRIGGERED SOLID STATE RELAYS is the way to go. You need one for each light, obviously.


FITTING.


This is always going to come down to what options are available to you.

The sensor I used needs to be able to detect the earth's magnetic field so some basic common sense thoughts were.

1) Don't hide it behind metal. Or surround it with metal.
2) Put it as high as possible (as far as possible from where the tyres meet road) – thus any angular deflection will be detected with greater accuracy.
3) for me it made sense to align it with the central axis of movement – i.e. In the middle rather than to one side of the bike. This makes movement relative to angular detection consistent.
4) Make sure you can get to the configuration port once it's fitted – it has to be calibrated.

I have a small top-case with a sort of pack-rest hump that is pretty much impossible to pack anything into. It's High up – in the Centre line and Plastic. And I can get to the config-port.

Fitting Lights.


OK Now this you'll need to work out for yourself. Whosoever you fit them you need lot's of adjustment options. If we use a traditional light in shell analogy – you need to be able to angle the shells left and right, up and down and need to be able to rotate the light units in the shells.

The JW Speaker 6150 I used DO NOT OFFER THIS FINAL OPTION – But I knew my mounting solution would resolve this.

Now – simple logic would dictate that one fits the LEFT side light to the left side of the bike and Right to right. And this is how my system is now. BUT think on this- As you bank left the left side moves closer to the ground – this reduces spread as you close down angles – I will be re-enginnering so that the right side light illuminates left and visa-versa. Equally the higher you can mount the lights – the better it will be. Mine are pretty much Headlight height.


DIALLING IT ALL IN.

OK this is VERY MUCH a work in progress thing.

But some starting points.

ECE Reg 053 states tilt lights must not illuminate at less than 3º tilt.
At 30º tilt frankly you're getting to the limits of many road tyres getting towards their limits - and probably you're past caring how well you can see at night.

From previous attempts at this and reading that university paper I have a gut feel that 20º is pretty much about where I want my final max illumination to be. Past this I'm happy for the cut off to go below horizontal on the left side on left handers – It might well be considered a “safety feature”

So rotate lamps approx 20º from horizontal Clockwise for the R/H illumination – anti-Clock L/H looked at from front. You also need to aim the lights into the bend you're going to illuminate. In my case this is achieved by the mount being mounted at the same angle as the forks and integrated into the light itself. The more adjustment you have the more fiddly but finally more precise it'll be. Some vertical adjustment is necessary just to get it all right. Light the bend but not oncoming drivers eyes.

A big hindrance to my progressing much is the lack of DARKNESS in Spain during summer. One thing I will be doing is fitting some small repeater LEDs to the lighting side so I can see even during the day when the aux lights switch on. This will give me a better feel for angles of lean etc.


Configuring the sensor.

As stated I will have them switch off when I come up to less than 3º tilt. This using the hysteresis setting. Frankly at just 3º tilt I really don’t NEED additional bend lighting - but found before that if hysteresis is set to “0” - I. It switches on/off at one angle it could be a bit annoying.

So the idea is to say – OK when I hit 6º (whatever) switch on – but then set hysteresis to 3º such that until you get to 3º the other way from switch on don't switch off – clear?

Also after some cornering can to approx 1 second damping – i.e. Unless the angle conditions are met for more than 1 second – stay as is – clear?


The final angles I choose will depend entirely on how it all feels in use on the road. Nobody said this was easy.

The clever amongst you are saying – Yes but if you set the light to 20ª and have it come on at 6º then the light points UP at it's outer extreme by 14º. This is true.

On the Kerb side this is not so much of an issue as ECE regs determine the dip cut off at 15º. On the non kerb side – it could be distracting to oncoming traffic – hence it will be a process of OBSERVE and ADAPT.

I'll add some pictures as I can. My Mother in law is due and one has to at least appear every so often.
 
Interesting read. I will admit my knowledge regarding inclination sensors is somewhat limited, but I would assume that one with internal gyro would be less susceptible to being surrounded/near metal (correct me if I am wrong).
 
Interesting read. I will admit my knowledge regarding inclination sensors is somewhat limited, but I would assume that one with internal gyro would be less susceptible to being surrounded/near metal (correct me if I am wrong).
A pure Gyro sensor as used in Dead Reckoning GPS solutions - yes - I think it could actually be better. But I've not found any prebuilt with switches etc.

The higher spec' Hummingbird one is still a Magnetic field sensor but with GYRO stabiliser.
 
First Road run.

Was at 7.2º Activate - 4.1 Hysteresis and 16 sample Damping.

Was getting some odd false on and the odd false off. But generally pretty **** good.
Get the sensation that Left and Right settings may well need to be different depending on which side of the road one drives.

I've changed to 8º on 5º Hysteresis and 32 sample Damping. Will test when it's cooler again.
ALSO I will try and get out into the middle of nowhere to zero the sensor.
There may be interference inside the garage (wires, pipes etc).
 
Bloody hell - but calibrating these blighters is a LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOONG process.

The lights work pretty much as they should but the sensor is VERY F*CKING sensitive. And frankly it may not be the right type of sensor.

I had a fair number of false positives/negatives yesterday

Things like switching on while upright - I have it set to a 3.2second decision minimum so occasionally one would activate 3.2s then go of.
More annoying occasionally switched off 3.2s in bends. It was BRIGHT SUMMER SUN so actually just technically annoying.

Part of this was that somehow I'd calibrated LEVEL (0º) off by 1.4 degrees - fixed that today.
All that seems level on an R3 apparently is not - and my garage floor is off by a degree.

The other part has has me stumped. Seems certain types of (here) common road feature must generate magnetic fields that confuse the sensor.

One of these is almost certainly POWER cables. I know these can generate monster magnetic fields. And every time I was close - things seemed to happen.

Another that seems to confuse things is long rows of lamp-posts on motorways - seemed to depend how fast I passed them and how close.

In bends - Crash barriers occasionally blipped things - especially the extra height ones on Bridges over rivers, other roads etc.

So I've upped the angle to 9º from 8º. See how we get on.

BUT I need to search for a different type of sensor - Gyro at a guess.

Hmmmm

If anybody sees a GYRO based SENSOR/SWITCH advertised - please post links.
 
I've had to deactivate it.

The sensor I have simply is too sensitive to electromagnetic fields. I've tried everything I can but the sensor gives FALSE ON's and OFF's.
My neighbour saw them and commented in the spirit of R&D input - if he did so will LEO's - but their comments will not be R&D.

Out in the middle of nowhere running at normal-ish speeds - the lights work very well indeed. Light the road where you'll be not where the headlights point.

But under severe (R3) acceleration - high speed cornering changes the IMU (Inertial Measurement Unit) sensor simply cannot cope with the high change of moment of inertia.
Any structures with large amounts of steel in (bridges) - large trucks with metal containers - and it can start false on/off switching.
Power lines - very interruptive.
At speed on motorways - regular spacing of Light posts, crash rails etc sets them off too.

Just the WRONG sensor for the job. I need to find something else - but still unsure what. Gravity I doubt will work.
Gyro - maybe but most reset every time they're switched on. And I'm not buying another expensive one from OZ. It's not worth it to me.

I've put the JW Speaker Fog lights back horizontal - and will reposition inboard a bit (less vibration).

Oh Well!
 
Shame.
I don't know if there are any gyro sensors that will keep their calibration. As you say, doubt gravity based ones will work well, (perhaps a gyro using gravity to set?) as I suspect the centrifugal forces while cornering will play havoc with it.

Will try to keep an eye out for anything that looks promising.
 
I'm going to see if I can convince SMIDSY to part with a couple of their inertial switches for Auto-Hazard flashers. Metal Balls on ramps. You never know.
Gyro chips are cheap - maybe using an Arduino or "Pi" family computer. Seen some projects based on the technology.
 
Ooh, very interesting. Never occurred to me to look into the Pi for something like this, but there isn't really any reason it wouldn't work. The computers themselves are pretty cheap too, not sure about the sensor board but doubt it would break the bank.

Do you know anything about programming such a device (although I suspect there will be someone who has made a program for what you want, all sorts of things available.)
 
Back
Top