Tearing into the Instrument Cluster

Soon after I bought the ’87 K75S, I
noticed (well, it was pointed out) that the neutral light didn’t work. I never noticed because I didn’t particularly care. I spent the majority of my time growing up driving a trail bike with either a burned out bulb or just plan broken N light, so I never look for one. Besides, I have that fifty gear indicator telling me when I’m in 0. Well, after awhile, thanks to the friend that pointed out the N light, it started to really annoy me. That and the fact the clock was displaying a bunch of odd ‘Chinese type’ symbols. So, a couple of weeks later I decided to open up the Instrument Cluster. To my surprise, I found the problem. At some point, there was a lot of water in here because I have a couple of rusted screws and rust deposit all over the back cover. I found all kinds of info on the net and tried them all out, but couldn’t find the source of the problems, at least with my limited soldering ability. The best I could figure was the N light must have something to do with the main circuit board. I happened to see a decent deal on ebay on a used cluster that was completely operational, except it was in imperial. That didn’t mater because I wanted to keep my original mileage. Besides mine was lower too. Another issue, the main board in the used cluster works, but whoever had it apart last had broken the plastic frame. To make a long story short, I went through a number of trials and found that my board was not the problem. In the end, I swapped the Gear indicator board and the clock and re-soldered the power circuit on the main board supplying the clock and voila, everything works.

One last thing before I seal, or at least attempt to seal, it back up. I have this thing about inaccurate speedometers. My last bike showed a constant 20km slow. It’s very annoying when the speedo show 80kph, it feels like 80kph, but you are actually traveling 100kph. On that bike, I just installed a bicycle computer, worked great. Last week, I happened to find this site showing how the K series speedo could be adjusted. So I followed the instructions, found some old Cat5E network cable, a hunk of steel and made an electromagnet. Then I just had to make a couple of test tones, uploaded then to my MP3 that ran through an old amp and just like magic, the speedo started spinning just like clockwork. From what I remembered when I checked my speedo using my GPS, I had a +7kpm error. Using math and what I found for a tire diameter, I was showing a +4 to +5kph error. In the end, I went with the math and rolled back the speedo 4kph. I figure I will at least be within 3kph if the math is off.

Your article is very interesting and it would be very nice if you were able to give us some details (with pictures) about the adjustment of the K bike speedo.
According to my GPS (when riding in straight line on the Highway), my speedo is pretty accurate.
I will point to you that on every speedo (European vehicles) the speed indicated is always about 4km/h higher than the reality.
The reason for that: if you get a speeding ticket, you cannot sue the manufacturer of your vehicle because of the inaccuracy of the speedo.
You also have to consider if your tire is brand new or worn.
Does your bike has the same size tires as the one sold with the bike?

Bert

Yes, mine has the standard size tire sold on a K75S. In 87, the standard K75 model had an 18" with a drum brake and the 'S' received the same rear rime as the K100S, a 17" with a disk brake. I guess I was thinking the speedo was setup for an 18” rim, but as you mentioned the manufacture may have intentionally set it high. I never sat down to figure how much difference an 18” compared to a 17” would make.

The theory behind this is building an electromagnet and feeding a sine wave or tone that when picked up by the speed sensor on the rear hub imitates a spinning wheel. I took 6 strands from a 12 foot long piece of network cable and wrapped it around an old nail punch, a nail should work fine. I will take a picture of it tomorrow.

With some simple calculations you decide what speed you would like to calibrate the speedo to and record a tone at the frequency that imitates that particular speed. Here is where I found the information.
http://ibmwr.org/ktech/speedo-cal3.shtml

I knew approximately how far my speedo was out and I found the sample tones from the previous link didn’t give me the results I expected. So I download a tone generator from here. http://www.nch.com.au/tonegen/index.html
I produced three new tones that should produce 80kph, 100kph and 160kph. As I mentioned, the GPS was showing me ~ +7kph and the test tone was ~ +4 or +5 so I still expect 2 to 3 kph error. I fault this to the tire diameter. When spring arrives and I can get out in a parking lot, I will roll the tire over a longer distance and with the bike loaded to get a more accurate diameter. Here is the link discussing the adjustment pot and with a link to a photo of it. http://ibmwr.org/ktech/speedo-adjust2.shtml

Here is a rundown on the formula I used. Every time the wheel turns, the pickup sensor receives six hits.

Tone(Hz)= cycles per second
Speed(mps)= speed in kph * 1000
1 hour = 3600 seconds

Tone(hz) = 6* (speed (mph)/3600) / (tire dia)

6*(100(km/h)*1000(m/km)/(3600(sec/h)*2.03m)
= 82.10 cycles/sec(Hz)

100kph=82.10Hz assuming my tire dia is 2.03m