Warning Lights (additional)

Hi,
Today I checked my brake lights to find the rear brake light switch was stuck on and the front brake light switch was only intermittently working!

Since I've had the bike on the road I've replaced the rear brake light switch once because of it getting stuck on when the adjustment is just slightly out...
So, I decided I just have to know what is going on back there when I'm riding.

[The Project]
The 'TAIL' warning LCD indicator works a charm if one or both of the lamps becomes open circuit which is great but... what if it gets stuck on ... or doesn't even come on?

So essentially I wired in a LED which is drawing 25mA and doesn't confuse the TAIL warning indicator(tested) which is expecting 880mA (2 x 27W bulbs).

Mounted inconspicously inside my fairing, near the dash, I will now know when my brake light is on, a non-directional(glowing) red LED is what I used.

Please see schematic below... any thoughts about any issues would be appreciated.

I also want to create a small, hopefully monolithic, circuit to light a LED when the current required to power two indicators (2 x 27W 880mA) is not flowing, to add a flashing warning indicator that one or both Indicator lamps has failed(based on an electronic flasher which doesn't slow down when one or more globes has failed)... soon. (unless someone has done it already of course, and if so, I'd love to see the schematic!)

 

Update:
I learned 2 things today.
1. 25mA is way too dull for daylight and changed the resistor to 330Ohms for 36mA(need to test if the LCD TAIL circuit still works though) instead.
2. It seems I have been riding around with my brake light stuck on for quite some time since every other time I hit the rear brake the switch stuck on and needed some flicking of the pedal to release.

The latter seems to be due to the non-standard spring connection of the switch to the brake, which I've now changed to a lighter duty spring...

Still had a whole bunch of fun riding in the cooler weather today though... 18C ish! Even if I was pararnoid about some car running into the back of me!

 

Hi,
Hoping for some electronic enthusiast, like I thought I was..., to give me a hand here.

Since I installed the indicator LED that shows me when the brake light is on or off, at the dash, I just love the assurance of knowing that the car coming up behind won't run into me because of no brake light... and in the same vein I've become obsessed with creating an indicator which will flash a LED when one or both indicator globes aren't working.

Basically the concept is,
1. run a high side shunt resistor which doesn't impede the flow of current to the globes too much. ie 4 x 0.1Ohm 5Watt resistors in parallel.
2. measure the voltage from the supply and after the shunt resistor and use that as the reference or standard, then when the voltage after the shunt resistor rises to high (which means not enough load and 1 or both bulbs aren't working) it will cause the LED to light.

Too easy you say... well that's what I thought

Since I now realise that anything I knew 20 years ago, electronic knowledge area wise seems to have been deleted, purged and militarily encrypted beyond my comprehension within my brain... Please help me!

Here is the schematic that I figured would work... but didn't seem to... if you think it should I'll get that damn breadboard out of the bin and try again!

Thanks in advance and if you work this out... You Rock!

PS: l1 is the load of the 2 bulbs(2 x 27w) and U2(OP-amp) is a LM741 and the NPN transistor is a BC547.

 

Way outta my knowledge range Always wish I knew more of it tho.

 

Search for a FAQ post I did a while ago titled something like "How to convert to LED's and still have everything work." Part of what it talks about is converting the 750 Seca/Maxim warning center to correctly check for operation of the tail and brake lights.

Sounds like your bike doesn't have such functionality and you're trying to add it.

The warning center is pretty clever on how it checks:

1. Tail lights run through a series resistor and the microcontroller checks for sufficient differential voltage drop at all times across this resistor.
2. Brake lights have a separate series resistor of a lower value (due to higher current draw), and a separate circuit to check voltage drop there, but only when the brake switch is closed.
3. The microcontroller also pulses the brake circuit for a fraction of a second after startup so it can check the brake lamps even though the brake switch is not closed.

In my conversion, I switched to a 3A rectifier for the tail circuit and a 5A Schottky rectifier for the brake lamps so that I get sufficient voltage drop with LED lamps but could also switch back to standard bulbs and have things work. You'd need series resistors, or rectifiers, plus at least one quad op-amp to act as a comparator for the voltage drops and to incorporate the brake switch on logic. Output from that would then probably need to trigger a 555 or something to flash your warning LED.

 

Hi SQLGuy,
Sounds like you had a bit of fun experimenting also after reading the write up. Although I am more interested in creating a plug in unit which I can just tie in, in series at the flasher relay for the current sensing.

Good work though! and if I do decide to get all LED'd up I'll be referencing your FAQ. Thanks.

 

Looks like I was working backwards with the Op-Amp and the voltage reference settings using the variable resistor.

Essentially,
The 2 x 27Watt indicator globes(plus the dash indicator globe) will now be powered through a set of 4 x 0.1Ohm 5Watt resistors in parallel. This will differentiate the voltage on one side and the other (Shunt resistor) but only create a very small amount of extra draw because of the low resistance.

We measure both sides and when the voltage on the globe side of the shunt resistor rises closer to the supply side we turn on a LED signal.

A bit of fine tuning needs to be done on the variable resistor to work out the best reference voltage. Which would involve setting it so the LED is off when both globes are in and taking a globe out and setting it so the LED just comes on. Put the globe back in and make sure the LED is off.

All the components are analogue and also not CMOS(ultra low power = sensitive) so... pretty robust as far as electronics are concerned and only about $10 worth of bits.

Unfortunately, it wont be a straight up cut into the indicator circuit as it needs a clean supply voltage also so the connections to the bike will be a cut into indicator circuit between the flasher and the switch (for the shunt resistor) and a power supply from the ignition circuit.

This is the principle and it works on my 'new' bread board but... I'm SURE there will be some sort of glitch somewhere so if anyone has words of wisdom I'd be appreciative.

In the meantime I'll be soldering it all together into a sexy looking lump of bits to see if it works!

 

Alrighty!
Now it works... off the breadboard plugged into the bike that is, but I feel happy with the performance after altering the voltage at the transistor supplying the indication LED(added R11).

Connecting the whole shebang in the indication circuit inline with the indicators works a charm also, so no need to connect the supply to a clean source... at this point.

I did a bit of a video to show the operation and it's simplicity. A couple of the clip-wires jiggled things about at one point but you can see the led flashes when 1 globe is disconnected and stops flashing, but stays lit, when both are disconnected. This is on the bike with the bi-metal type flasher unit and not the electronic flasher but I would envisage the electronic flasher would just keep flashing... something else on the check-list.

A few more tests and I'll see what happens. If all goes well I reckon I'll be implementing on my car also, for the brake lights and indicator lights, fingers crossed and maybe after a bit of time trialling it's durability too.

 

with one of these Arduino you could monitor everything on the bike and do things you never thought of

 

Polock... one step at a time! I'm trying hard to not over complicate the electronics but I've run into a bit of an oversight, on my behalf.

The indicator switch switches the output of the flasher unit to the indicators and that being said when the ignition is 'on' there is power at the flasher unit.
So... if I splice into the circuit between the flasher and the switch the unit I've just built is powered up and has barely any draw and as such sees the condition as both globes being blown and leaves the LED illuminated.

When I was testing I plugged the unit into one side of the harness after the switch and I neglected this 'small' extra piece of complication.

I really wanted to keep the hook up simple, like a general purpose plug in module...

Maybe a secondary input from the output side of the switch just to switch the indication circuit on would be the go... 1.5ma via signal diodes perhaps to a NPN base controlling the op-amp and LED power supply. That would require 2 signal diodes and another wire from the harness... Not pretty but still not too bad I think. Excuse me if it seems I'm babbling.

 

Well...
I spent so much time messing about with this it's driving me a little around the bend so I'm leaving it at this point.

It will be plugged between the feed to the indicator switch and the flasher relay so the warning indicator LED will be on when the indicators are off and off when the indicators are on, apart from if a fault conditions occurs, which sucks, but it's only drawing sub 25mA so not a big problem power wise. I'll see if it is distracting...

I'm pretty sure there may be some AC problems arising from the charging circuit but... I'll cross that bridge later.

I know this would be a worthwhile, collision saving, project but my kung fu is not as strong as it needs to be in this area.

 

After reading SQLguy's LED post I couldn't help but pick a few up when I was down at the local electronic store today (Jaycar).
Cost me $1.75 each and I spray painted the ends of the LEDs in the row of five indication lights with flat black as they are directional LEDs and as SQLguy points out are way too bright through the indication light panels (indicator, neutral,warning,highbeam,neutral,indicator).

The display lighting I decided to go for a non-distracting red LED backlight to give the dark-room effect and to top it off a blue LED for the police digital speed meter.

Not sure if the colours will be distracting or just groovy but I'll try it out at any rate.

 

Indicator warning LED is fitted and operational.
Short video to show it in action.

 

Is it just me, or is your tach like WAY off (reading too low)

 

I only just started it and it was a bit nippy.
Normally it sits just above 1k after I've gone down the road a few Kms but until it's warm I have it set so it barely ticks over, otherwise it idles way too high when its warm.

That being said I obviously have never compared it to anything else yet... another 'todo' on the list.

 

The indicator light being on when everything was off and flashing when there was a fault is just too annoying so I added a cutoff portion to the indicator warning circuit.

if draw is less than (2 globes)
then
if draw is (1 globe or more)
then light LED
fi
fi

Essentially the LED will not light if all of the globes are dead or the switch is off. So it is still a patch in circuit to before the switch and after the indicator relay.

Haven't implemented it yet but will get the soldering iron out again shortly.

On another note... I couldn't stand the LED's in the dash because of the directionality of magnified lens and ended up grinding the indicator LEDs flat and leaving the ends opaque. Now they work well.

The police meter lights need to be incandescent or 360degree versions of the LED wedge lamps but at $27 each... incandescents went back in.

The backlights for the speedo and tacho are now blue LEDs as the red LEDs made me feel ill strangely. LOL

 

Well, I have an '82 Maxim 750, but I'm not a mechanic or much of a bike rider ( Surely good enough to not get killed, but not hardcore ;-) ) but I've been into electronics for almost 30 years...

Most normal LEDs upper current limit is 20mA -- 15mA is closer to where you would want to run it. If that's not bright enough for you, you'll want to either:

1) "gang" them for more light, or
2) look for "High-output" versions (some LEDs can produce up to 20K mcd at 20mA - most older LEDs were around 1K or less, so that's quite an improvement in efficiency.

Running normal LEDs over 20mA current will seriously shorten their lifespan. There are some LEDs (usually called 1W or 3W) that handle much more current, but are also much brighter (prolly much too bright to use in your instrument cluster) and usually require heatsinks for temp. management.

This calculator can help with figuring out resistor values for LEDs at different source voltages & target current:

Single Resistor Calculator

Also, different colors of LEDs have a different forward voltage -- Red LEDs have a forward voltage of 2.1v. Yellow (IIRC) are 2.5v. Blues are 3.5-4v. This will make a difference for what resistor values you can use -- swapping a red LED for a blue LED in a circuit totally changes how much current you'll be drawing across the diode. Here's a chart for forward voltage vs. color:

LED Color / Forward Voltage chart

One other thing that I (personally) would be concerned with -- the reason that red light is used in an instrument cluster is because it won't b0rk your night vision nearly as much as blue -- too much blue or white light and it might impair your night vision if you do a lot of night riding. You might want to try orange LEDs and see if those don't make you feel ill, they may not be as "harsh" on your night vision as blue.

Anyway, I hope this helps, and I'll watch over this thread if you have any other electronics questions, I'd be more than happy to help!

Thanks,
Roger "Merch" Merchberger

P.S. Yes, since buying my Maxim earlier this year, I've been thinking about what I could do to "trick out" my instrument cluster as well... ;-)

 

Hi Roger,
Thanks for the great info.

You've definitely come to the right place to polish your mechanical skills. Within a very short period you'll find that these guys knows these XJs!

Just so you know the RED light in the instrument cluster was originally WHITE incandescent lamps(the 2 backlights for the speedo and tacho) and so far the RED and the BLUE T10 Wedge LED lamps have both been lower light alternatives as they aren't that effective for area illumination but are more directional spot type illumination.

I still haven't had the opportunity for a long ride at night yet, to test, after changing to the blue, but they do look cool on a stationary test.

I admit that about 15 years ago I was fairly adept at basic electronic circuit creation and troubleshooting as I had most of the basic principles stuck in my head, but I've become quickly amazed at the information I just don't have anymore.

I'm enjoying re-learning the basics... hopefully I don't blow too much stuff up in the meantime!