TOPIC:

I've got a '79 KZ750 B4. It's my first and it's been an interesting bike, but it's had this really annoying problem which came down to the previous owner(s) hacking away at the wiring harness haphazardly. Not to mention my handlebar switches were less than reliable, and I'm replacing the handlebars, and going to do WG's coil mod. With all this going on, I decided to strip and completely rewire the bike.

Perfect time to make modifications, right? So I'm about a third of the way though this, I've got the "unbuild" done for the most part and my parts are sourced, and I'm working on the new wiring diagram, and trying some ideas out. Here's where I need some help.

I've taken off the marker reflectors that face out laterally from just forward of the gas tank, and replaced them with amber LED clearance lights of the same size. Simple mod, direct fit, and they're bright as blazes and look really sweet. One option is to wire them into the ignition-switched power and have them on as running lights. I may do that, but I'd much rather wire them to flash as a side-firing turn signal. They would stay on as running lights while I'm riding, and when I turn on the left turn signal for instance, I want the one on my left side to turn "off" when my turn signals flash "on"...in other words, the marker light is flashing opposite the turn signal, and stays on when the t/s is off.

My idea for doing this is to have each side of lights on a 5-pin relay, with the switch terminal on the relay driven by an electronic flasher. I'd have my main turn signal lamps on N.O. terminal 87 and my marker light on N.C. terminal 87A. This way, the marker light is on whenever the ignition is on, but switches off when the flasher turns on the relay (and the signal lamps). This sounds reasonable and should work.

It doesn't work as expected and I'm not sure why. I wired this up quick-and-dirty on my bench to test it out...what happens is the relay doesn't have enough time between on-off pulses to turn off all the way. After the first cycle, my turn signal lamps stay on or flicker as the relay slowly switches off and then quickly switches back on.

Am I doing something wrong with the relay/do I need to source a better relay, or different type of relay? Is there an alternate way I can wire this that dispenses with the relay entirely? Or is this delay within the relay something I need to account for, and maybe try to find another way to drive the relay? (Thought about an electronic "pulse maker" with a 555 timer and maybe an op-amp to drive the signal up to 12V for the relay...but I'd really rather go with the automotive flasher since I bet it's probably a lot more robust.)

Any suggestions? Anything? Have I bored you guys to tears? Would pictures and/or video help?

Could you get dual filament bulbs and wire it up just like a regular tail light? (1 cicuit for tail, 1 for turn)

Or in your case here, maybe drill and add an extra led into your housing? I would be interested to see your idea as you have it now.

Post edited by: jjdwoodman, at: 2008/02/17 02:15

Could you draw a picture of the circuit?

What you want to do should only require 2 simple relays. You should not have to alter the wiring for the normal turn signals. That way you can use the stock signal flasher.

If you are running all of the signals through the relays, and using the turn signal flasher only to power the relays, you will have a problem with flash rate unless you use a 3-prong electronic flasher.

I would suggest keeping the normal signals as stock, and just tap into the signal wiring to add the relays. Each relay coil will be powered by it's respective turn signal feed. Then the common contacts of the relays will get steady 12v, and the 87a terminals will feed the respective LEDs.

Here are my notes on flashers about half way down this page:
www.geocities.com/loudgpz/GPZturnSigInteg.html

loudhvx wrote:

Could you draw a picture of the circuit?

What you want to do should only require 2 simple relays. You should not have to alter the wiring for the normal turn signals. That way you can use the stock signal flasher.

If you are running all of the signals through the relays, and using the turn signal flasher only to power the relays, you will have a problem with flash rate unless you use a 3-prong electronic flasher.

I would suggest keeping the normal signals as stock, and just tap into the signal wiring to add the relays. Each relay coil will be powered by it's respective turn signal feed. Then the common contacts of the relays will get steady 12v, and the 87a terminals will feed the respective LEDs.

Here are my notes on flashers about half way down this page:
www.geocities.com/loudgpz/GPZturnSigInteg.html

Yeah, though all I have is MSPaint right now, but here you go:


img84.imageshack.us/img84/676/proposedsetupie7.png

You've got some interesting information. I read your note on trying to use a 2-prong flasher to drive a relay. That sounds like what's happening. I'll have to trade it for a 3-prong and see what happens.

However, it is true that I could use the lights wired in a stock fashion to drive the relay. I had not considered that before--see that's the kinds of ideas I was hoping for! Anything else you've got I'd like to hear.

Also--your headlight wigwag and T/S circuit uses a pair of 555s in sequence. I've actually got a couple of the 556 chips from Radio Shack and they're itching a hole in my workbench. What's your experience using these on a motorcycle? Will they run ok on 12 or 15 V? Will they output 12 V? Would EMI/RFI pose a problem? (I'm running points ignition with copper wires.) I had thought about using a 555 since they output a clean square wave and you can adjust duty cycle to take care of switching delays. What sort of precautions did you use on your hardware?

Hope this helps!

If you don't mind experimenting a little, there is a way you can do it without any relays and maybe no extra parts.

Keep the bikes wiring stock. Feed 12v to the LED's as if they are to be on all the time, then wire the ground side of the LED's to the positive wire for the respective turn signal bulb. This will work as long as the LED's use much less current than the turn signal bulbs. When the turn signal is off, the turn signal bulbs will act as a ground for the LED so it will be on. When the turn signal lights up, the ground for the LED becomes 12v, so the LED turns off.

(Applying 12v to both sides of an LED won't do any harm to the LED, it will simply turn off.)
If the LED does not completely turn off, then you just need to put a diode or two in series with the LED.

Post edited by: loudhvx, at: 2008/02/17 14:47
About the 556 chips, the chip itself is ok with the 12v from the bike, and work well, but the actual one-shot circuit is susceptible to voltage fluctuations. It's inherent in the nature of timing circuits using RC time-constants. Fluctuations in supply voltage alter the charge/discharge rate and threshold voltages, so the circuit becomes less consistent when the supply voltage acts as that of a motorcycle. My quick fix was to install a giant filter capacitor "Cfilt" in the circuit. A better solution would have been to add a 7805 or 7808 regulator to the timing logic and a few transistors to buffer the logic from the relay coils. But I got lazy and just threw in a filter, which luckily worked well enough.
The only shortcoming is that under severe voltage fluctuations the flash rate may alter very slightly, or the automatic-cancelling may occur a few seconds earlier or later.

Post edited by: loudhvx, at: 2008/02/17 14:59

loudhvx wrote:

If you don't mind experimenting a little, there is a way you can do it without any relays and maybe no extra parts.

Keep the bikes wiring stock. Feed 12v to the LED's as if they are to be on all the time, then wire the ground side of the LED's to the positive wire for the respective turn signal bulb. This will work as long as the LED's use much less current than the turn signal bulbs. When the turn signal is off, the turn signal bulbs will act as a ground for the LED so it will be on. When the turn signal lights up, the ground for the LED becomes 12v, so the LED turns off.

(Applying 12v to both sides of an LED won't do any harm to the LED, it will simply turn off.)
If the LED does not completely turn off, then you just need to put a diode or two in series with the LED.

Best idea so far. Someone on another board mentioned this too; the concept was floating in the back of my head but I wasn't sure how to implement it.

I'm gonna try this and see if it works. Thanks!

I assume the LED's you are using have built-in dropping resistors? If not you'll have to use some. If the LED is called a "12v LED" then it has the resistor built-in.

Yes, they do, it's a sealed unit. It's interesting because I don't know if there's a rectifier in line, but the LEDs will light no matter which direction the current's applied. There's about 1/2 MOhm resistance in the forward direction and 5 MOhm in the backward direction--but both directions do require about 2.2 V across the terminals before they will light. There's no immediately obvious difference in brightness either.

I've been playing with them some--I'm looking to mount two more in the back half. But I don't have a factory reflector to cannibalize, so I'll have to dream up another mounting solution.

That's interesting that it'll light in either direction. Yeah a simple bridge rectifier would do it. Do they give you a reading for the milliamps it uses at 12v, or did you happen to measure it?