Bike computer / light interference

[PrettyboyTim]

I recently got a cheap Maplins wireless bike computer, and I thought it was pretty neat for the money. However, just recently I've been having some trouble with it - it sometimes measures my speed as being much faster than it actually is.

Today I worked out the what the problem is.

My bike computer is sat right next to a Cateye 3 LED front light on my bars, and when I have the Cateye turned on (I tend to have it turned on to blink mode), I get an extra 32Mph added to my speed, even when I'm stationary.

I'm not sure what frequency the wireless sensor on my computer uses, but it's quite evident that when in use my Cateye produces pulses at a similar frequency. I guess I'll have to move my computer over to the other side of my bars now!

 

[barq]

I've heard of similar things with HID lights too. Does the interference still occur when on steady mode?

 

[Elmer Fudd]

Had the same problem with my computer, I was doing 27mph when stationary, and I also had intermittent problems when the light was on constant. I'd go wired every time if you are using l.e.d. lights.

 

[HLaB]

I have the same problem with my SMART Polaris LED's sometimes interfering with the Raleigh Wireless computer. I think it only does it when the LED batteries are nearly dead and my other lights (LED's also) don't. My computer tends to underestimate when it goes awry, i.e when I'm flying downhill it only reads 0 or 1mph.

 

[the_mikey]

I decided to investigate this, using a software defined receiver with a wide bandwidth, I was able to test a number of cycle lights. I found that most modern cycle lamps do radiate noise..

Most lamps using Lithium Ion batteries seemed to not create a problem when using the brightest setting, and only started creating noise when using dimmed modes, this is down to the system of controlling brightness where the lamp is strobing on and off, if there are several brightness levels the ratio between the lamp being on and being off changes on each setting, the action of the LED's turning on and off rapidly creates a spike of wide band noise. Most of these circuits seemed to start creating noise between 100kHz and 300kHz, some were much lower with noise spikes starting as low as 40Hz, but the harmonics at frequencies above this at 40Hz intervals.

Testing some lamps that use AA/AAA cells also produced mixed results, particularly noisy was a Moon nova 80 which employs a buck converter to raise the single AA battery voltage to a level needed to drive a bright LED.

In general this is only a problem for devices using the low end of the radio spectrum, LF and MF , and most GPS devices that communicate using ANT+ and Bluetooth uses 2.4GHz and not LF so will be mostly unaffected, some cheaper computers that use their own dedicated wireless sensors may experience difficulties, signal drop outs or sensors that simply fail to function when the LED lamp is turned on.


If you are experiencing this problem then the solutions are: move the lamp away from the cycling computer, use a different cycling computer or use a different lamp.

 

[Ming the Merciless]

It’s called electro magnetism not noise! Any electrical current will generate such a field.

 

[andrew_s]

Most cheap cycle computers use inductive coupling rather than transmitting a radio signal - i.e. when the magnet is next to the sensor, a brief current passes through a coil in the "transmitter", which generates a magnetic field, which generates a current in a coil in the computer head.

The advantage is that it's short range enough that you don't need to worry about interference from a similar computer fitted to the bike being ridden next to you, and that it's very cheap.
The disadvantage is that you'll also receive signals from any other pulsed current that's close enough. Pulse width modulation in LED lights is by far the most common, but you can also get signals from other sources - when I used a wireless computer, I used to get a ~70 kph reading every time I rode over a particular traffic detector loop in the road (not one directly associated with traffic lights).

Options are:
a) move the LED light further from the computer head
b) use a wired cycle computer instead
c) use an expensive computer that does transmit by radio. These generally usecoded transmissions that do get distinguished from similar sensors on other bikes, and often a cadence option that needs enough range to require this.