More LED Turn Signal Woes

[slug]

Pardon me if I ramble, but I'm at my wit's end... (and I know, this should probably be in Technical Topics...)

 

So I bought a set of four LED turn signals for Dirty Bird, knowing full well I'd have to monkey with the flasher unit to get them to work. "No problem" I said, since I thought I knew a little about electrical systems... First I tested out the left side of the bike before I actually installed them. First, with the stock blinker with no resistors attached, I connected the new signal to the left rear and left the original signal connected at the left front, and the blinking was fine. Next, I replaced the stock left front with the new signal, and the lights came on, but no blinking, the signals just remained on. Probably means that the new signals aren't pulling enough current to operate the blinker, so I started testing out resistors...

 

I've attached resistors with values ranging from 5 to 40 ohms to each signal, no dice (on the positive and negative terminals, and both). I've attached resistors with values from 10 to 40 ohms on either side of the stock flasher unit, still no dice. I think I've tried every possible combination.

 

I gave up on the stock flasher unit and bought the FL-2 Red electronic flasher from Super Bright LEDs. Again, I tested one side before installing, this time front and rear were blinking. Next I attched a signal to the right side. Now, the left and right side signals are blinking at the same time!! However, when I attach the old signals with the new flasher installed, only one side blinks at a time. So, I tried the resistors with the LEDs again, still no improvement.

 

One thing I did notice was that the indicator light on the instrument cluster was very dim when it was on...

 

I've gone over the schematic for the electrical system ('02 - '03 Sport/LeMans) and in the turn signal circuit, the flasher unit is connected between the fuse and the switch, so at least in theory, the flasher shouldn't be the problem.

 

The only place on the schematic that things could get crossed up is the indicator light(s?) on the instrument cluster. I have two schematics, actually. One shows a single bulb being shared by the left and right indicators, the other schematic shows separate indicators for left and right. I'm going to assume that my bike matches the schematic with the shared indicator...

 

If I'm making proper sense of the diagram, it looks like the ground for the indicator light depends on which direction is on, so if the left side is blinking, the circuit for the right side is used as the ground, and vice-versa. The other possibility is that I'm on crack (no biggie since most people think I'm on it anyway.)

 

So here's my theory:

 

1: The higher resistance of the incandescent dash indicator and the reduced resistance of the LED indicators forces more current to go to the desired direction indicator.

2: Whatever current does flow through the dash indicator isn't enough to fully power the bulb, but...

3: The remaining current that went through the dash indicator will then travel through the opposite indicator (the side that shouldn't be blinking) on its way to the ground, but will light up the indicator because it takes much less power than an incandescent bulb to operate.

 

 

If this is somewhat correct, my problems now makes more sense to me because I was testing the resistors on the back of the bike where they wouldn't help ... So in theory again, if I yank the dash indicator bulb, the blinking should only be on one side as it should, since it stops any current from flowing through the indicators on the opposite side.

 

Of course I'll be busting out my multitester tonight to test this out, but does this sound somewhat right to you guys?

[Guest Nogbad]

The resistor must mimic the current draw of a 21W bulb. The resistors must therefore be wired in PARALLEL with each LED indicator. You need a 10 ohm 15 Watt resistor. This is a little less than the 21W but should be sufficient.

[slug]

The resistor must mimic the current draw of a 21W bulb. The resistors must therefore be wired in PARALLEL with each LED indicator. You need a 10 ohm 15 Watt resistor. This is a little less than the 21W but should be sufficient.

87607[/snapback]

 

Hmm, is the purpose of the the resistors to a) Provide enough current draw to operate the flasher thingy, or Provide enough resistance to keep the signal on the opposite side from lighting up?

[Guest Nogbad]

Hmm, is the purpose of the the resistors to a) Provide enough current draw to operate the flasher thingy, or Provide enough resistance to keep the signal on the opposite side from lighting up?

87614[/snapback]

 

Provide enough current draw to operate the flasher properly, and bypass any leakage current that would cause unwanted LEDs to light. Given the LEDs require only about 1V and a few milliamps to light them it is easy to get reverse earth currents lighting them or leakage current if the flasher is a solid state one. You need the parallel resistors to match the LED winkers to the flasher circuit designed for bulbs.

[slug]

Provide enough current draw to operate the flasher properly, and bypass any leakage current that would cause unwanted LEDs to light. Given the LEDs require only about 1V and a few milliamps to light them it is easy to get reverse earth currents lighting them or leakage current if the flasher is a solid state one. You need the parallel resistors to match the LED winkers to the flasher circuit designed for bulbs.

87617[/snapback]

 

Makes perfect sense. However, I'm using an electronic flasher designed specifically for LEDs, and its my understanding that this should eliminate the need for resistors... Again my suspicion is the dash light and (from what I can guess) its alternating ground path...

[BrianG]

Pardon me if I ramble, but I'm at my wit's end... (and I know, this should probably be in Technical Topics...)

 

 

so here's my theory:

 

  1: The higher resistance of the incandescent dash indicator and the reduced resistance of the LED indicators forces more current to go to the desired direction indicator.

  2: Whatever current does flow through the dash indicator isn't enough to fully power the bulb, but...

  3: The remaining current that went through the dash indicator will then travel through the opposite indicator (the side that shouldn't be blinking) on its way to the ground, but will light up the indicator because it takes much less power than an incandescent bulb to operate.  

 

87600[/snapback]

 

1. The current flow through the LED is LOWER than through the incandescent so the resistance is higher. If the current flow were higher the flasher would work.

 

3. There is enough power available from the battery to light any number of little incandescent bulbs in parallel. This means that the one bulb isn't drawing current away from another, at least not enough to notice. Consider the headlight (huge current draw) runs in the same circuit with the tail light (much less current draw) and both light to their full capacity. Also a LED will NOT light with back-flow current. Being a light emitting DIODE, they allow current in only one direction, and produce light only in that direction.

 

If you didn't know this about LED's.... are you certain that you observed the proper polarity when you installed the LED's in the signal light positions?

 

Excepting the possibility of mis-installed LED's, I don't know what's wrong with your set-up. If I were a betting man, I'd suspect a poor ground somewhere........ probably in the instrument pod area. This would cause current to seek some other path to ground..... like some unsuspecting light bulb.

 

I replaced my signal lamps with LED units and simply placed a 10ohm resistor across each of the 2 tail light's pigtails. Didn't even use them on the front.

 

Hope this helps some....

[Guest Nogbad]

Slug, looking at your wiring diagram, it seems the dash light is connected across both indicator circuits, and this would have the effect you observe.

 

In the standard set up, the dash light earths through the unused indicator circuit depending which way you are turning. With ordinary bulbs, the pilot bulb current is insufficient to light the main bulbs but IS sufficient to light a LED.

 

To prove this, remove the pilot from the circuit. The indicators should then operate normally.

 

I will post a solution shortly.

[Guest Nogbad]

OK, here is the resistor-free solution since you have a LED dedicated flasher.

 

1. Disconnect the indicator bulb from both flasher circuits and EARTH one side of it.

2. Connect each flasher circuit to the same side of the bulb through a 1A power diode such as an 1N4001. You need two diodes as each winker circuit must be blocked when the other is on.

 

 

 

PS any old 1A diode with at least 50V RRM is good, all the 1N400x series for example. You can get these for about 10c each from Radio Shack.

[slug]

BrianG: Thanks for the input!

 

OK, here is the resistor-free solution since you have a LED dedicated flasher.

 

1. Disconnect the indicator bulb from both flasher circuits and EARTH one side of it.

2. Connect each flasher circuit to the same side of the bulb through a 1A power diode such as an 1N4001. You need two diodes as each winker circuit must be blocked when the other is on.

 

PS any old 1A diode with at least 50V RRM is good, all the 1N400x series for example. You can get these for about 10c each from Radio Shack.

87657[/snapback]

 

Nog, may I take a guess that you're an EE?

 

I did that test last night and removed the indicator bulb, and voila! Only one side blinks. On one hand, I think the design of the indicator circuit is a bit goofy, but then again, I'm sure the designers never considered that people would be using clusters of LEDs to replace the original bulbs, and after all, it is a Guzzi.

 

Your solution is very close to what I'm planning to do

 

I was thinking of replacing the bulb with an LED, something along the lines of this (just need to make sure I find a proper ground in the instrument cluster):

[Guest Nogbad]

Nothing wrong with your idea, but it's more trouble than just keeping the existing pilot bulb. If you are on the seach for perfection though.....

[slug]

Nothing wrong with your idea, but it's more trouble than just keeping the existing pilot bulb. If you are on the seach for perfection though.....

87687[/snapback]

 

I am a total geek, and I proudly admit that!

 

Remember the three Os:

 

Overengineer

Overbuild

Overkill

 

[Guest Nogbad]

I am a total geek, and I proudly admit that! 

 

Remember the three Os:

 

Overengineer

Overbuild

Overkill

 

87698[/snapback]

 

Er... You forgot "Overspend"!

 

So there are 4 "O"s (And a fanatical devotion to the Pope)

[mike wilson]

Er... You forgot "Overspend"! 

 

So there are 4 "O"s (And a fanatical devotion to the Pope)

87700[/snapback]

 

I didn't expect that.....

[Guest Nogbad]

I didn't expect that.....

87706[/snapback]

 

NOBODY expects.........

[slug]

NOBODY expects.........

87712[/snapback]

 

Uh-oh, I know where this is going