Any ideas/success in shielding interference for wireless bike computers?

[Zero_Enigma]

I'm not sure if the term is EMF here that is disturbing the wireless computer. When I put in a freshly charged 1.42-1.46v x 6AA battery pack peaking at ~8.6v I find my 'quad-damage' 4xSSC LED 3.6-9v @ 800mA using the DX driver to really give me a hell of a bright light for my downhill runs or when I need to turn those burners on to ride through a dark path.

One thing I've noticed is that my wireless computer freezes and does not record any data sometimes and just stays at zero for the speed till I've used the light for like 3-5mins then the bike computer will start recording data on one of my battery packs. My other 6AA battery pack I find is ~8.2v and I can still run my bike computer then. I like wireless over wired as less wires the better so that's somethign I wish not to give up but there has to be a way to shield the light or the drivers or something that is causing the wireless bike computer to not record data.

When I turn off the 'quad-damage' light the wireless bike computer starts recording no problems even with my 1W NiteHawk in direct path to the wireless trans/reciever. I am a bit of a stat junkie for my riding and log all my data while I go riding in my log book and because I ride both road and trail paths I can't really measure out my shortcut or trail areas like I can on the road with Googlemaps if I needed to use the google maps thus why I like having the bike computer which also aids in my data collecting and timing for maintance and such.

Any help is appreciated.

 

[Zero_Enigma]

I've also thought about a GPS but .. if I'm under heavy canopy or clouding I may lose signal and my stats.

 

[znomit]

Check out the buckpuck datasheet.
Put a cap on the input if you're running long wires to the battery.

 

[pe2er]

If using only a capacitor across the battery connection does not help, put a (clamp-on) ferrite bead around the battery lead, near the driver.

It works like a choke on the EMI generated by the driver and prevents the battery leads from acting like an antenna.

You can buy one, or salvage one from an old computer keyboard/ display/ mouse cable.

 

[Zero_Enigma]

If using only a capacitor across the battery connection does not help, put a (clamp-on) ferrite bead around the battery lead, near the driver.

It works like a choke on the EMI generated by the driver and prevents the battery leads from acting like an antenna.

You can buy one, or salvage one from an old computer keyboard/ display/ mouse cable.

How much are those things and where can I find one?

 

[znomit]

How much are those things and where can I find one?

They are a few $
eg
http://www.radioshack.com/search/index.jsp?kwCatId=&kw=ferrite&origkw=ferrite&sr=1

Did you try the cap?
Shortening the battery cable and/or rerouting away from the speedo may help too.

 

[[email protected]]

The term you want to use is EMI (electro magnetic interference). The DX drivers have nothing to address it since they are typically used in a metal faraday cage (aka your flashlight).

Try 2 capacitors across the driver.

One largeish one, say 10uF to 100 uF. Tantalum caps work best but can be spendy, especially at higher capacitance and higher voltages. You can substitute a less expensive electrolytic cap for a tantalum, but it won't work as well. I try to be conservative and use a cap rated at twice the highest voltage I expect to subject it to (you can round down to the next available voltage if need be).

Use one small one, say in the 0.01uF to 0.1uF range. A small ceramic disc cap will work fine. In electronics, these are often refered to as decoupling caps. One rated for 50 volts will be quite small.

If you have a really long power cable, you might also try a couple of caps at the battery end.

The ferrite that covers both of the power wires will help a little bit, but not as much as a ferrite on each one. You can buy some quite small ones that you can solder inline on each power lead (one for the positive lead, one for the negative), and then just cover each one with heat shrink tubing.

Here are some like ones used that have worked very well in drivers I've designed: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=240-2513-1-ND

The reduction in noise is amazing. I do quick and dirty EMI tests by holding my light next to an FM radio. With a driver with no ferrites, if I get within 20 feet of the radio's attenna all I get is static. With ferrites, I have to practically wrap the antenna around the light before I pick up any noise.

Not sure if these suggestions will fix your wireless odometer issues, but it should cut down the noise a lot. The DX drivers have nothing for EMI.

Mark

 

[macforsale]

The newer GPS chip sets have better reception under less than ideal situations. I never found heavy overcast to be a problem. Downtown city canyons can always be a problem depending on satellite constellation configuration at the time. If you can try a trial run just to be sure. I am not sure how to determine chip set other than name brand new off the self should be the latest.
Mike

I've also thought about a GPS but .. if I'm under heavy canopy or clouding I may lose signal and my stats.

 

[Barry_Scott]

i have a garmin 705 and have never lost gps signal even in heavy forest i also never have an issue with the cadence and speed sensor. ( yes they do have a speed senor which allows you to turbo train)

 

[xiejol]

I used to have one light, which I kept on the left handlebar, the computer was on the right. I added another light and the computer went haywire.

I second pe2er's post, I'm using two chokes, one on each wire, and it has stopped the interference.

 

[UberLumens]

caps and hash chokes should work well,

if its still blasting shield the driver/housing in brass screen

 

[FX-32]

I tried with different caps (various from 10uF to 3300uF, and lots of ceramics too), and it's the same.
The electronics in the light isn't the problem (I also tried disassembling the light, and was testing it without all metal-case), the batteries aren't the problem either (I tried using the charger instead of the batteries)... I think the problem is the cable that acts as antenna.
It catches interferences mostly in low and strobe mode in my case (in Hi only when turn on but doesn't show any KM/h of difference), but also while charging the batteries sometimes if cycle-computer is too near catches some interferences.

I'm going to try re routing the cable, and adding some ferrite clamp to solve it.

 

[2_i]

I suffer from a similar problem, where switching on of the dynamo IQ Cyo lamp badly affects the reception of weak FM stations on my bike radio. I think I tried capacitors in the past without any success. I'll try the bead when I am reunited with my bike after my current travel.

With a folder in the current travel, I deal with a combination of a Philips Saferide dynamo lamp and the same-brand radio as on the main bike. It will be interesting to see whether there is any difference in interference between Saferide and IQ Cyo. The first lamp has an alu housing and the second - plastic.

 

[FX-32]

I tried with some ferrite rings, but it isn't enough. I think that adding lot of aluminium paper should work.
This is what I get with my scope from the LED:
http://i56.tinypic.com/34oeqtd.png

 

[Steve K]

Nice scope photos! It looks like the switching power supply is running a bit below 100kHz, with some pretty significant spikes when the switch transistor turns on and off. The spikes are really where all of the high frequency energy is. If you can reduce their amplitude or increase their rise and fall times, that'll be an improvement.

The advice that's been offered so far is good. The switching power supply can make noise at both the input (power from the battery) and the output, so you'll want to try improving both. Ferrites and caps are the right parts to use, but ... they need to be ones effective at the frequency of interest. The ferrite manufacturer will have a datasheet that shows the impedance over a range of frequencies.

Similarly, caps are effective only up to a certain frequency. Ceramic caps will be useful at higher frequencies than tantalums or aluminum electrolytics. You might want to put a 10uF or 100uF right at the input to the switcher, but also put a 0.1uF ceramic there, and a 1000pF immediately at the switcher terminals wouldn't hurt either. The goal is for the switcher to get its surge of current from the caps, and not from the battery located many inches away.

At the output.... maybe a 0.1uF right at the output, a ferrite after that, and maybe another cap (1000pF?) right afterwards.

Part of the trick is to minimize the wire length or trace length going to these filter elements. Even an inch of wire can have significant inductance at high frequencies, and could make a good capacitor ineffective. My practice is to mount the parts on copper-clad circuit board, using a dremel tool to cut pads into the copper. An example of this is a recent circuit board for my dynamo headlight. It's not pretty, but the idea works. Note that much of the board is used as a ground plane too, which reduces the inductance of the ground traces.

The noise is almost certainly radiating off of the cables; either the cable to the battery or the cable to the light. You might get some benefit from using twisted pair wiring. This reduces the ability of differential noise to radiate off of the cable. Shielded twisted pair wiring might work even better, as long as the shield is connected to the ground. Reducing the length of the wires should help, as should routing the wires away from the computer.

It'd be interesting to see how these ideas pan out. If possible, take pictures of the ferrites and caps and how they are installed.

best regards,
Steve K.

 

[2_i]

OK, some extra data point from the Philips Saferide dynamo lamp. While that lamp has an alu casing, its interference impact on my bike radio seems, subjectively so far, even more serious than for the IQ Cyo in a plastic housing. Thus, the Cyo produces a constant hum of interference that affects the reception of weak stations. On the other hand, the Saferide produces a sequence of chirpy spills that impact even stronger stations. The spills appear to depend on the speed at which one is riding, with downhill being the worst. A direct comparison for the reception of identical stations at the same location will come after I get back to my home base.

 

[2_i]

After getting home, I have a chance to experience the effect of Philips Saferide on a couple of my favorite tad weakish radio stations and the noise is awful, with the stations barely getting through at all. (It is a shame, as Philips is one manufacturer who actually pays attention to tuner reception in consumer products.) Winding the two wires going to the dynamo, onto a random ferrite ring, just made the situation worse. The ring seems to work as an antenna. Putting alu foil around the wires worsens the situation at first, but then it gets better after connecting the foil to the frame. Overall, at the moment it seems best when winding the wires around the fork's leg and covering them up with a grounded foil. Next is doing nothing aside from winding the wires around the leg. I will try next snap-on ferrites onto individual wires and get, otherwise, tubular copper braid for shielding.

UPDATE: I've got the snap-on ferrites from Radio Shack, stock No. 273-067 (273-069 is for slightly thicker cable) and put them separately on the two wires going to the dynamo. The radio interference is grossly reduced! This will now become my standard for LED lamp installations. My only worry is of the ferrites opening up. I guess I'll put a large diameter heatshrink encompassing both ferrites (I do not want to split the double wire all the way through.) Thanks guys for all those suggestions. I guess I'll get also some extra snap-on ferrites to keep around the house given the impact they can have.

UPDATE 2: I now installed ferrite beads (this time some off Ebay) over the wires of an IQ Cyo, right outside the lamp. For IQ Cyo these beads eliminate the EM noise from the lamp completely, as judged by the FM radio. By listening to the radio alone, one cannot tell whether the lamp is on or not. Naively, one might expect less success with the Cyo than Philips, because of the plastic housing of the Cyo. In parallel, I further bought those tube copper braids to shield the supply wires, but I doubt I will now bother with them anymore.