• Required reading for all forum users!!!

    Welcome!
    Register to access the full functionality of the GSResources forum. Until you register and activate your account you will not have full forum access, nor will you be able to post or reply to messages.

    A note to new registrants...
    All new forum registrations must be activated via email before you have full access to the forum.

    A Special Note about Email accounts!
    DO NOT SIGN UP USING hotmail, outlook, gmx, sbcglobal, att, bellsouth or email.com. They delete our forum signup emails.

    A note to old forum members...
    I receive numerous requests from people who can no longer log in because their accounts were deleted. As mentioned in the forum FAQ, user accounts are deleted if you haven't logged in for the past 6 months. If you can't log in, then create a new forum account. If you don't get an error message, then check your email account for an activation message. If you get a message stating that the email address is already in use, then your account still exists so follow the instructions in the forum FAQ for resetting your password.

    Have you forgotten your password or have a new email address? Then read the forum FAQ for details on how to reset it.

    Any email requests for "can't log in anymore" problems or "lost my password" problems will be deleted. Read the forum FAQ and follow the instructions there - that's what we have one for...

  • Returning Visitors

    If you are a returning visitor who never received your confirmation email, then odds are your email provider is blockinig emails from our server. The only thing that can be done to get around this is you will have to try creating another forum account using an email address from another domain.

    If you are a returning visitor to the forum and can't log in using your old forum name and password but used to be able to then chances are your account is deleted. Purges of the databases are done regularly. You will have to create a new forum account and you should be all set.

3 Colour LED Charge Indicator - opinions?

  • Thread starter Thread starter fatnfast
  • Start date Start date
F

fatnfast

Guest
I have just fitted a DIY LED charge indicator to the clocks of my GS750. Its a single LED that has been set up to display whats happening in the battery charging dept.

Basically it functions as follows:
<11.99 volts - the LED is Orange and blinks 4 times (repeat)
12 - 12.49 volts - the Orange LED is constantly on
12.5 - 13.09 volts - the LED changes to Green and blinks 4 times (repeat)
13.10 - 15.49 volts - the Green LED is constantly on (Charging correctly).
15.5 volts and above - the LED changes to display a constant Red light.

I have wired this up to come on with the ignition, and it now gives me a good idea as to the health of the charging system. It should now let me know that all is well, and if a fault develops I will be able to pick it up quickly.

I have used a 10mm tri-colour LED (smaller ones are available) and it is a simple two wire connection (earth and a switched positive feed). It used a very small PCB and all the voltages can be pre-programmed.

Is this a good idea, or simply a device for the paranoid?!

The attached pics shows the green LED display.



View attachment 10716
 
How much .. where do I send the paypal :) ?

I think it is an awesome idea !
 
Great idea, and it wouldn't be a bad idea for you to make up a couple dozen. I am sure a bunch of the members here would love to have that on their bike(s).

I am planning a similar pic circuit, but the purpose of mine is to change the color of the backlight in my analog volmeter gauge.

Kudos, and seriously, consider making a bunch up and I am quite sure they would sell, as long as GSR members get a discount of course!
 
Sounds like a nice option for not adding alot of space and has nice functionality.

I assume you can water proof the device, also you said it is programmable so I'm guessing you have a PIC in there.

How big is the unit behind the LED? 1" square? would be nice if it fits inside of a gauge housing.
 
Hell, Id kick a few bills for one for my 650 and my Chop. Nice idea.
 
I am running something similar in my cage at the moment, with a view to putting on on my bike .....
Except I only use two LEDs, orange for under voltage (<12.5 V) and red for over voltage (>14.8V).
I am not a huge fan of lights on a dash to tell me everything is fine.
What did you use a PIC?, mine is just a zener, some resistors and a few transistors, so it can't be adjusted (easily) once it is on the car/bike.
It is a good idea for sure, any early indication of a fault is a bonus seeing as how the GS charging system can be a bit delicate.....
 
I'm pleased it is seen as a useful device and its not just me wanting to add some lights to my bike!! Sadly, it was not made by myself. I was searching for the parts to make up a simple device to give a basic red/amber/green charging status and stumpled accross a guy on ebay who had done all the hard work (and I struggled to find the parts to make it up cheaper).

It would quite easily fit into an unused 'dash' light (I was to lazy,and just wanted to see if it could survive on a motorcycle!). I have protected it from the elements by using a little bathroom silicon.

I have wired it via a basic relay to read direct battery voltage (I have also used the relay to provide the sensing input for my 6-wire Honda reg/rectifier).

He is a UK seller, but I'm sure he will ship to you guys (if not, I would happily act as a middle man for no cost).

The price is a stagering $12.50 at the current exchange rates!

He goes under sparkbright_products and if you cut and paste 160457896281 you can see a typical product.

Turn around in the UK was less than 24 hours from order to delivery including a custom set up for your bike.:D
 
I am running something similar in my cage at the moment, with a view to putting on on my bike .....
Except I only use two LEDs, orange for under voltage (<12.5 V) and red for over voltage (>14.8V).
I am not a huge fan of lights on a dash to tell me everything is fine.
What did you use a PIC?, mine is just a zener, some resistors and a few transistors, so it can't be adjusted (easily) once it is on the car/bike.
It is a good idea for sure, any early indication of a fault is a bonus seeing as how the GS charging system can be a bit delicate.....

I breadboarded a 3 color solid/blinking LED with a PIC a couple of years ago using a Red/green/yellow LED. Wrote about it here too. I personally didn't care for the look/functionality as much as a 20-LED "bargraph" built with 2 LM3914s (looked kinda like posplayers 2nd link but with 20 LEDS)

FWIW, what I mean by functionality is: With the bike runnning, the voltage is not a constant voltage, it oscillates quite a bit around the average, and the 3914 type shows this as a "smear" in the display. I not only thought that looked neat, but it also gave LOTS of info about the charging going on ... the smear was darker where the voltage spent the most time, and showed both the high and low voltages that the oscillation was going to. response time was VERY fast, so you could easily see how the charging changed with blips of RPM (as opposed to digital meters, which have to average over time) ... :)

But anyway, the PIC version is dead simple (and cheap ... probably under $5)
It only requires a PIC, red/green LED, four resistors and 1 bypass cap.
If you do it as SMD you might be able to fit the whole thing UNDER a 10mm LED.
Even if you use leaded components, its only an 8 pin DIP for the PIC.
You would want to use the HV615 PIC: it has a voltage regulator (to make it easy to run from the bikes power), an A/D converter (to sense charging voltage), and a PWM circuit (to change color easily) built in.
The default project people use to learn about PICs is making a LED blink ... this project only adds a little bit of complication in that it does so in response to voltages sensed by the A/D ...

If anyone wants to build a PIC version, I'll give more advice, and sample code (not completed though, I used a pot to emulate changing the voltage cause it was easier to run on the breadboard in developement ... but thats a pretty trivial change)

If anyone is interested in the LM3914 version, I still have a writeup I did with pictures, schematics, layout, etc somewhere.
The 3914 voltage monitor was done on premade circuit boards available at radio-shack (part #276-170), and required no particular electronics experience, just the ability to solder. Parts are probably $10-$15.
I liked this much better than a digital display (I rubberbanded a multimeter to the handlebars for a while)
It was easy to read at a glance, gave good information, was easy to read in all kinds of light (night to bright sunlight)
Size was about 3 inches wide, 1 inch high, 2 inches deep.
If and when I get around to it, I'll probably redo it to be 2 x 0.5 x 2
I rubberbanded mine to the top of the instrument cluster.
 
You would want to use the HV615 PIC: it has a voltage regulator (to make it easy to run from the bikes power), an A/D converter (to sense charging voltage), and a PWM circuit (to change color easily) built in.
The default project people use to learn about PICs is making a LED blink ... this project only adds a little bit of complication in that it does so in response to voltages sensed by the A/D ...

If anyone wants to build a PIC version, I'll give more advice, and sample code (not completed though, I used a pot to emulate changing the voltage cause it was easier to run on the breadboard in development ... but thats a pretty trivial change)

.

Those pics are unbelievably cheap :eek:.

I just looked over the data sheet. I'm assuming the development environments are basically free, but what are you using to program the device? Is it "in-circuit programming" or are you using some type of ROM programmer? I saw there are several methods, but what is cheapest for a 1 off design?

This might be a good candidate for an "anti-slosh" filter.
 
Those pics are unbelievably cheap :eek:.

I just looked over the data sheet. I'm assuming the development environments are basically free, but what are you using to program the device? Is it "in-circuit programming" or are you using some type of ROM programmer? I saw there are several methods, but what is cheapest for a 1 off design?

This might be a good candidate for an "anti-slosh" filter.




Microchips IDE
Write Assembler, debug, fairly involved simulator with stimulus, convert asm to hex.
Think it does C too, but I never used C.
(you can learn a LOT about the PICs and what they are capable of just by writing programs and using the sim to see if they do what you want)

WinPIC
Loads Hex to PIC using hardware below

Hardware
Needs Parallel port and 14 volt power supply.
I used the schematic unchanged, but changed the layout to fit the ubiquitous radio-shack #276-170 board and save myself etching.

These are by no means the only or best solutions, they are just what I tried and used. They are all free. Everything worked pretty much 1st try with no gotcha's.
There are Parallel port, Serial port, and USB programmers. Also, Microchip sells the PICSTART Plus and PICkit, which are complete ready to go solutions (I think).

On the hardware end, I put a solderless breadboard next to the programmer, and ran the wires to that.
Push a DIP PIC into the breadboard and its ready to program.
Put LEDs, MosFETs, Pots, etc on the breadboard to build test circuits.
WinPIC has a hardware debug mode which lets you toggle individual pins, so turn on Vdd (but not Vpp) to boot the PIC and let it run without even removing it from the breadboard or changing any wiring.
You can toggle the clock and data lines to provide input to the PIC too if desired.
For In Circuit Programming and SOIC PICs, run wires out of your project board and stick them in the breadboard. Clip them when done programming and debugging.



I've got a quicker, easier project for you than the anti-slosh filter:
Connect the Coil relay mod to a PIC output vice the ignition Switch. PIC boots to a default of "off"
Read the ignition's signal generator output, determine the RPM. (some PICs have op-amps and/or comparators built in, use those to read sig gen)
If RPM is greater than "X" (determined empirically) turn on the coil relay. Once RPM hits 1000, the relay stays on till power (the ignition) is turned off.
... Automatic Anti-kickback ...

Options:
pot that allows X to be adjusted between two limits,
thermistor (or diode) to sense temperature and automatically adjust X lower on cold days if required
If RPM goes over 10,000 drop the relay again (rev limiter)
(Hidden) toggle switch that changes rev limiter to 5,000 rpm ... useful for testing rev limiter, and also in case your inexperienced brother-in-law wants to borrow your bike.

If you build it, let me know what you come up with to read the sig gen ... If I get motivated I'll do a rev limiter, but I have no need of anti-kickback
 
Man... I wish you brains could figure out a better or replacement way to do the electronic tachs in some of the later model GSs.. After so many years, they'll get awful screwy... Ive installed 3 different boards in mine, and they were all wildly off at one side of the powerband or the other..
 
Microchips IDE
Write Assembler, debug, fairly involved simulator with stimulus, convert asm to hex.
Think it does C too, but I never used C.
(you can learn a LOT about the PICs and what they are capable of just by writing programs and using the sim to see if they do what you want)

WinPIC
Loads Hex to PIC using hardware below

Hardware
Needs Parallel port and 14 volt power supply.
I used the schematic unchanged, but changed the layout to fit the ubiquitous radio-shack #276-170 board and save myself etching.

These are by no means the only or best solutions, they are just what I tried and used. They are all free. Everything worked pretty much 1st try with no gotcha's.
There are Parallel port, Serial port, and USB programmers. Also, Microchip sells the PICSTART Plus and PICkit, which are complete ready to go solutions (I think).

On the hardware end, I put a solderless breadboard next to the programmer, and ran the wires to that.
Push a DIP PIC into the breadboard and its ready to program.
Put LEDs, MosFETs, Pots, etc on the breadboard to build test circuits.
WinPIC has a hardware debug mode which lets you toggle individual pins, so turn on Vdd (but not Vpp) to boot the PIC and let it run without even removing it from the breadboard or changing any wiring.
You can toggle the clock and data lines to provide input to the PIC too if desired.
For In Circuit Programming and SOIC PICs, run wires out of your project board and stick them in the breadboard. Clip them when done programming and debugging.



I've got a quicker, easier project for you than the anti-slosh filter:
Connect the Coil relay mod to a PIC output vice the ignition Switch. PIC boots to a default of "off"
Read the ignition's signal generator output, determine the RPM. (some PICs have op-amps and/or comparators built in, use those to read sig gen)
If RPM is greater than "X" (determined empirically) turn on the coil relay. Once RPM hits 1000, the relay stays on till power (the ignition) is turned off.
... Automatic Anti-kickback ...

Options:
pot that allows X to be adjusted between two limits,
thermistor (or diode) to sense temperature and automatically adjust X lower on cold days if required
If RPM goes over 10,000 drop the relay again (rev limiter)
(Hidden) toggle switch that changes rev limiter to 5,000 rpm ... useful for testing rev limiter, and also in case your inexperienced brother-in-law wants to borrow your bike.

If you build it, let me know what you come up with to read the sig gen ... If I get motivated I'll do a rev limiter, but I have no need of anti-kickback

thanks for the run down.:D

I did a search on ebay and a bunch of under $50 OPTIONS pop up for programers/dev kits. I have a Needhams EMP-11, but they seem to be out of business now. Maybe I should put that on e-bay while people still know what it is.

Anyway I forgot about the anti-kick back. I did do some experiments and found that the logic did need some more smarts than what I got from simple relay based relays. A pic would be good.

I still have my INNOVATE LM2 and LMA3 to integrate with O2 sensor, and will see what is required to get the RPM stable.

While these PICS are pretty cool mainly because they are so cheap, in college I was doing hand assembled hex code on 6800's back in 1978 and at work brought my first wire wrap Z-80 CPU up from scratch for a pic type application in 1981( actually that turned into a major program for Raytheon at the time). So I really dont have a burning desire to run off and play with one of these without a compelling need. :-k

My starter clutch with the 1166 has been doing well as long as the battery is fully charged. Trying to start with a poor battery will invite the kickback though.:mad:

If mine goes out I might change my mind about need in a hurry. :o I'll keep it mind thought.
 
Man... I wish you brains could figure out a better or replacement way to do the electronic tachs in some of the later model GSs.. After so many years, they'll get awful screwy... Ive installed 3 different boards in mine, and they were all wildly off at one side of the powerband or the other..

Sorry, Josh no time for that anymore. Even my mechanical one bounces alot. :eek:
 
Man... I wish you brains could figure out a better or replacement way to do the electronic tachs in some of the later model GSs.. After so many years, they'll get awful screwy... Ive installed 3 different boards in mine, and they were all wildly off at one side of the powerband or the other..

Are there pots to adjust, and maybe bring it into calibration ?

Also, my understanding is that electrolytic capacitors dry out over the years and change a lot in value. A first quick and dirty shot to try to fix it woud be to just replace any electrolytic caps with new ones of the same value. The value should be printed on them.
 
I breadboarded a 3 color solid/blinking LED with a PIC a couple of years ago using a Red/green/yellow LED. Wrote about it here too. I personally didn't care for the look/functionality as much as a 20-LED "bargraph" built with 2 LM3914s (looked kinda like posplayers 2nd link but with 20 LEDS)

FWIW, what I mean by functionality is: With the bike runnning, the voltage is not a constant voltage, it oscillates quite a bit around the average, and the 3914 type shows this as a "smear" in the display. I not only thought that looked neat, but it also gave LOTS of info about the charging going on ... the smear was darker where the voltage spent the most time, and showed both the high and low voltages that the oscillation was going to. response time was VERY fast, so you could easily see how the charging changed with blips of RPM (as opposed to digital meters, which have to average over time) ... :)

But anyway, the PIC version is dead simple (and cheap ... probably under $5)
It only requires a PIC, red/green LED, four resistors and 1 bypass cap.
If you do it as SMD you might be able to fit the whole thing UNDER a 10mm LED.
Even if you use leaded components, its only an 8 pin DIP for the PIC.
You would want to use the HV615 PIC: it has a voltage regulator (to make it easy to run from the bikes power), an A/D converter (to sense charging voltage), and a PWM circuit (to change color easily) built in.
The default project people use to learn about PICs is making a LED blink ... this project only adds a little bit of complication in that it does so in response to voltages sensed by the A/D ...

If anyone wants to build a PIC version, I'll give more advice, and sample code (not completed though, I used a pot to emulate changing the voltage cause it was easier to run on the breadboard in developement ... but thats a pretty trivial change)

If anyone is interested in the LM3914 version, I still have a writeup I did with pictures, schematics, layout, etc somewhere.
The 3914 voltage monitor was done on premade circuit boards available at radio-shack (part #276-170), and required no particular electronics experience, just the ability to solder. Parts are probably $10-$15.
I liked this much better than a digital display (I rubberbanded a multimeter to the handlebars for a while)
It was easy to read at a glance, gave good information, was easy to read in all kinds of light (night to bright sunlight)
Size was about 3 inches wide, 1 inch high, 2 inches deep.
If and when I get around to it, I'll probably redo it to be 2 x 0.5 x 2
I rubberbanded mine to the top of the instrument cluster.

Martin, this is an ancient post of yours, but I stumbled across it just now looking for info on LED brightness.

Do you still have the details for your bargraph at all? I'm keen to have a look at that myself.

I want to see under and over voltage on the dash the instant it happens, but I don't like the idea of having to interpret blinking rates of LED's, I just want a simple guage like your bargraph.
 
Martin, this is an ancient post of yours, but I stumbled across it just now looking for info on LED brightness.

Do you still have the details for your bargraph at all? I'm keen to have a look at that myself.

I want to see under and over voltage on the dash the instant it happens, but I don't like the idea of having to interpret blinking rates of LED's, I just want a simple guage like your bargraph.

Attached is a bitmap with the layout, and following is the text description.
I have pictures somewhere of mine, if I can find them I'll post.
... then you can OOOH and AHHHH the cardboard and duct-tape housing :eek:
(Really ... I'm not kidding ...)

it uses a Radioshack 276-170 breadboard so you dont need to etch a board


******************************************************

This is the description for the Voltage Monitor.

The monitor is based on a pair of LM3914 Chips and a pair of
10-LED "bargraphs"
(like those sequential LED "power" displays that are on stereo amps and
tapedecks)

I used 2 of each for a 20 segment display, but you could use only 1 of
each for a 10 segment display or lots of each for an N-times-10 segment
display

The upper and lower endpoints of the displays can be adjusted to
indicate anywhere between about 8 and over 20 volts or so.

You can get the data-sheet for the LM3914 at
http://www.jameco.com/Jameco/Products/ProdDS/24230.pdf

The 3914 is pretty easy to use, and the monitor is more or less right
off the datasheet ...

The picture is the layout, and it is so simple, I didn't even draw a
real schematic.

it uses a Radioshack 276-170 breadboard so you dont need to etch a board

Jameco part#s are

300003CK 2 LM-3914
334529CK 2 LED Bar (Red)
334511CK 2 LED Bar (Green)(alternate to above ... pick your favorite color)
330798CK 1 10uf electrolytic
25523CK 3 .1uf ceramic chip (bypass)
41822CK 2 10K 15-TURN POT
41873CK 1 20K 15-TURN POT
Assorted resistors

an explanation of how it all works:

----------

The LED displays are at the top of the board.
Only the cathodes of the LEDs are soldered into the breadboard, where
they connect to the LM3914s.
The LED assemblys are then bent almost 90 degrees so that they can be
seen when the board is seen edge on from the bottom (see pictures)
This makes your display 1 inch high and 3 inches deep (and 3 inches wide)
All the anodes are then connected to a hookup wire which is connected to
V+ (see picture) (actually, I divided mine into 3 sections ...
electrically the same thing, but mechanically stronger)

Instead of bending the LEDs, you _COULD_ just solder the anodes directly
to the board at V+ (you would need to drill 3 holes), but then your
display would be 3 inches high and an inch deep (and 3 inches wide)
(I.E. it would not mount on the bike as well)

---

In the middle of the board are the 2 LM3914s and the associated
circuitry.

Each LM3914 has resistors connected to Ref-out and Ref-adj to set the
LED current and the voltage refrence.
(LM3914-1's Ref-out is not actually used, but is programed anyway so that
it roughly tracks LM3914-2's Ref-out)
Please note that the resistor values for LM3914-1 and LM3914-2 are
different.
(since those for LM3914-2 also account for the current drawn by the
internal divider chain)

There are also resistors connecting pins 9 and 11 of LM3914-1 to V+
which are required for proper daisy-chaining

---

On the bottom left corner are four 22k resistors.
(I think I may have changed these to 10k in the end, to make the LEDS a
bit brighter)
These connect ground to the cathode of four of the LEDs in order to
dimmly illuminate these LEDs at all times. The purpose of this is to
give a sense of edges of the display when it is dark outside. Kind of
like "background illumination". You could eliminate any or all of them
if you wish (or add more ...) The only requirement is not to connect to
LED 9 of the first 3914 or LED 1 on the second, since that would screw
up the daisy-chaining. When the 3914s illuminate one of these four LEDS,
they just add to the resistor's current, and swamp it. These LEDs are
shown in a lighter color in the layout for refrence. This idea is not
shown in the LM3914 datasheets

---

The caps are all just for bypass. A 10uf electrolytic and .1uf chip
where power comes on the board, and one more .1uf at each IC.

Thats probably way overkill, the first 2 are probably enough...
(and its possible that you could get away with no bypass anywhere at all)
But the caps are cheap, its recomended practice, and I'd rather not
track down oscillation problems because I tried to save 10 cents on a
cap.

---

On the far right are 3 multiturn pots. These set the high and low
settings for the Displays.

The rightmost pot is just a voltage divider between V+ and ground.
Its center terminal goes to the inputs of the LM3914s.
It is primarily used to set the voltage for LED 10 of the 2nd LM3914.
The other 2 pots do not have _MUCH_ influence on LED 10 of the 2nd LM3914.

The middle and left pots parallel the internal voltage dividers on
LM3914-2 and LM3914-1 respectively. The high side of the divider on
LM3914-2 is connected to the REF-out pin, and the low side of the
divider on LM3914-1 is connected to ground via a 10k resistor.
Thus the internal dividers and this 10k resistor form a 3 part voltage
divider which divides the ref-out voltage.
Since you can adjust the effective resistance of the internal dividers
with the parallel potentiometers, you can adjust the relative
proportions of the 3 part divider.

So you can adjust LED 10 of the 1st LM3914 (and LED 1 of the 2nd LM3914)
with the middle pot, and LED 1 of the 1st LM3914 with the left pot

Unfortunately, the other pots also interact with each of these settings,
so you have to iterate to get the settings correct...

You will need an adjustable powersupply to adjust the pots
It could be as simple as an LM317 (with a resistor and a pot)
http://www.jameco.com/Jameco/Products/ProdDS/23579.pdf
and two 9-volt batteries.

The setting procedure is as follows:

1) set the supply to the desired voltage for LED 10 on the 2nd LM3914
(for example 15 volts)
2) adjust the right pot till LED 10 on the 2nd LM3914 illuminates

3) set the supply to the desired voltage for LED 10 on the 1st LM3914
(for example 14 volts)
4) adjust the middle pot till LED 10 on the 1st LM3914 illuminates

5) set the supply to the desired voltage for LED 1 on the 1st LM3914
(for example 10 volts)
6) adjust the left pot till LED 1 on the 1st LM3914 illuminates

goto step 1 and repeat
(continue repeating until no more adjustments are needed in any step)
 
Awesome, thanks Martin!

I've dumped the text and bitmap into a document so I can follow at my leasure.

I'm going to be getting some LED's for my guages soon so I'll get these bits and pieces at the same time.

This will give me something to do while I wait until I can budget for my exhaust and things.

And if you find those pic's, that will help me visualise it properly too, duct tape and all :)

Not sure if you saw it, but my exhaust mock up is cardboard... :rolleyes:
 
Back
Top