LED Arrow turn signal

[bkuhr]

[ame=http://www.youtube.com/watch?v=beqi-UuiXq8]YouTube - flash arrow.AVI[/ame]

Above vid of my design for arrow led turn signal

At first I thought I would PM the only other member I know here who has experence in electronic design, Dingy, to see if he was interested in collobrating in my project, as he had stated in another post he was working on simular, then I thought this would likely be of interest to others, especially the final result, so I will provide my progress here.

 

My plan is to replace lighting on my Venture to LEDS. Many places to buy plug and play, but I thought I would like turn arrows, and started experimenting

 

I set my conditions for

1. turn arrows at 4 corners to sequence both front and rear on left or right as switched

2. turn lights also marker lights

3. turn lights also 4 way hazards

4. Minimum parts count/$

5. also design trim lighting/brake lighing/passing lights from LEDs

 

In my favor I found a source for cheap LEDS, buying 100ea 5mm for about $5 including shipping. For testing I have bought 100 each, Blue 5mm 6000mcd 20-50mA 3.0-3.4fV, Red 5mm 4000mcd 20-50mA 2.0-2.2fV, and Yellow 5000mcd 20-50mA 2.0-2.2fV. All have narrow view angle of 25-30deg and are very bright see straight on, not so much from an angle.

Item 5 above is pretty straight forward, so started design of flashing arrow first.

 

PDF is of my circuit so far, I am working on breadboard so changes are easy.

Circuit is based on HCF4094 8 stage shift registar driven by high voltage 555 clock called TLC555. The shift registar outputs are driving NPN transistors, but due to high part count I chose transister array CA3083 with 5 each independent transistors capable of 200mA each. I chose my IC's purposly as high voltage (18V) as opposed to TTL(5V) to avoid parts count in voltage control for our 12-14 volt MC systems.

 

Leds are arranged in series arrays. Had wanted to only have 1 major control circuit, but will need two , for left and right. Plan on 1 central control box with multiconductor wire to each corner light. As current to each segment is only 20-30mA this is well within capability of telco 24ga inside wire such as CAT5, and 4 pair wire gives 8 conductors for 8 chevrons, using main 12V already at each corner.

 

Flashing arrow works great so started designing item 3 above. A little experimenting accomplished lighting all LEDS in sequence till all lit then hold, then a timer to output enable of the 4094 will flash all lit LEDs as in 4 way flasher.

Then I started working on 2 above, meaning I thought I would need to dim the turn signals and have them fully lit to act as marker lights.

I designed a 555 based pulse width modulator which will adjustable dim the LEDs, but at a noticable dimness as opposed to fully lit arrows, I think the LEDs are too dim and can't seem to find a happy medium.

 

This is where I am looking for input. now I am thinking I need to have all LEDs lit full bright for marker lights, and when turning, only turn side will change from full bright to flashing arrow. also considering ordering White 18000mcd leds, but high fV will require changes in array.

Weclome all questions, comments, suggestions

[Bubber]

I looked at your schematic and all I can say is do you have that in english for those of us that are severly electronically challenged? LOL

Sounds like a great idea...best of luck with your project!

[dingy]

Here is a PDF of what I have at this point. It is still in a relay based format, as I have way more relay experience than digital. When trying to follow this, the five LED relays at the bottom need to be slow pick and slow release at about 200ms. It is fairly easy to do the slow release, but the slow pick is much more difficult with small relays.

 

This version incorporates the existing wiring in the turn signal switch & brake circuits function. I haven't done anything regarding front turns. I was concerned about doing fronts due to the viewing angle of LED's might not lend itself to good visibility with the design of the front turn signal housing on my 1st gen.

 

The R1 & R2 resistors are not needed that are shown. I would use 12v LEDs.

 

Tail lights LED's are not completely shown in this version as they will just be separate LED's on an independent circuit.

 

Timing shut down circuit is not shown. This would eliminate the stock turn signal & hazard relay. I was going to use the reed switch in the speedometer housing for a distance traveled input and a 555 chip for a timed input to cancel the flashers.

 

Basic description is a turn signal or hazard request picks the request relay which starts the cascade pick of the LED1-5 relays at 200ms interval pick time each. Once the LED5 relay picks, it picks the Pulse relay which starts a cascade drop of the LED1-5 relays at 200ms drop time each. When the LED5 relay drops it drops the pulse relay which starts sequence over.

 

The picking and dropping of the pulse relay every 1000ms would be used for the front flashers & dash indicators.

 

There is a right & left brake over ride circuit to cut out the brake light if there is a turn signal in progress. Brake lights illuminate same LED's as turn signals, only not cascading.

 

I have tried to design this circuit around the existing wiring in the turn signal switch. It is rather awkward to do since the turn signal circuit is sending a 12v positive signal to the circuit if the turn switch has not been manually canceled. The actual flashing is initiated by a 12v negative pulse when the switch is moved either left or light then released back to center. Hazards over ride turn signals with a constant 12v negative signal. Rewiring/reconfiguring the turn signal switch would be very difficult.

 

Gary

[bkuhr]

Gary, thanks for sharing. Wow, lot of relays. Not sure if you want critque or not, if not sorry-don't means to offend.

Just a couple of notes

 

I see you call for time delay on the LED relays, planning to use special purpose relays or design descrete delays at each relay? Would recommend simple RC circuit on each coil to delay the following relay from activating, then maybe another contact assembly in each relay to deactivate the previous relay.

 

I also see you are calling for +5v source for your LEDs. As you may be aware LEDs have certain operationing specs that must be met to light. Industry standard LEDs, both panel indicator and high output require 20mA to light. Some can operate up to 50mA with limited duty cycles, but below 20mA most will not light at all or only dimmly. They also have a Forward Voltage rating simular to the .6V drop seen on a standard diode. Due to different chemicals required to make different colors, most have different FV ratings in the 2.0-3.5v range.This means each LED will drop as example 2.0v in the 20mA circuit.

Standard resistor for 2.0v-2.2v LEDs in a 12v circuit is 470 ohms. If all LEDs in your chevron are in parallel, your circuit will require about 2.3 volts(use 470ohm resistors on 12v) and run about 60mA(3 leds) per chevron. If your LEDs are in series to each other, each LEDS will drop about 2.2 volts times 3 leds =6.6volt drop plus your signialing diodes drops and you will need to supply about 7.5volts with total circuit current of 20mA per chevron.

 

Have you considered duty cycle ratings for you relay contacts?

 

Finally a question, in your pic rear view MC, looks like you have yellow under red lenses. Is this LEDs?

[Owen]

bkuhr

 

Just a couple thoughts... I did not look closely at the LED specs but...

 

1. Using series leds in each chevron will cause the complete chevron to blank if one led fails.

2. Using 12v leds, not in series, would allow you to use one zener diode say of 9v range to dim leds as running lights. Simply switch it out of the power line and switch in the sequenced drive power for sequencing.

3. Using pulse width modulation for dimming the leds could cause interference with the audio/radio system or ignition noise could affect the pulse width.

4. The timer/driver design is solid and should perform nicely.

5. Auto turn signal cancellation could be accomplished with a second 555 timer.

 

Just some random thoughts to consider.

[bkuhr]

Just a couple thoughts... I did not look closely at the LED specs but...

 

1. Using series leds in each chevron will cause the complete chevron to blank if one led fails.

This is true, but running all all LEDS in parellel increases circuit current to 20mA x #leds. I currently have 40 per corner=160leds x 20mA=3.2 amps. Still less than bulbs, but severly exceeds other design specs such as multiconductor phone wire. Also usually, but not always when an led fails to light it will still pass current and still experience the diode voltage drop.

2. Using 12v leds, not in series, would allow you to use one zener diode say of 9v range to dim leds as running lights. Simply switch it out of the power line and switch in the sequenced drive power for sequencing.

Good idea, but the zener would not have any effect. The LED requires the specified voltage drop.To high a forward voltage can easily smoke an LED, and a voltage below spec and LED will not light at all. I did experimant with different resistors to effect dimming, but a difference of 10ohms+- goes from bright/dim/out and get into non standard resistor ratings, and with 10% tolerance resistors I am using very difficult to duplicate results and would likely be severly tempature sensitive. This is the reason PWM is preferred for dimming LEDs

3. Using pulse width modulation for dimming the leds could cause interference with the audio/radio system or ignition noise could affect the pulse width.

I agree with this, although suppresion cap could help, and I suspect this may be the cause of Rf noise on another current post about LED driving lights.

This is reason I am now considering leaving all LEDS on bright for markers, and turn off and activate arrows for only turn side leaving other side on steady bright

4. The timer/driver design is solid and should perform nicely.

Thank you. Only minor problem with design so far is I probably need to debounce the turn input signal(more parts), as I have experianced random flashing of chevrons on occasion rather than in sequence order, but the random pattern IS NOTICABLE. May not be that bad.

5. Auto turn signal cancellation could be accomplished with a second 555 timer.

I plan to interface odo reed switch with another timer to replicate orginal MC specs.

Breifly studied Garys 83 wiring dirgram, and think I can interface all parts of my circuit by removing turn, hazard, cancel relays and tapping into those respective sockets and rear harness quick disconnects, avoiding cutting any wiring on the MC

 

Just some random thoughts to consider.

 

Great suggestions, keep em comming

Brian

[dingy]

Gary, thanks for sharing. Wow, lot of relays. Not sure if you want critque or not, if not sorry-don't means to offend.

Just a couple of notes

 

I see you call for time delay on the LED relays, planning to use special purpose relays or design descrete delays at each relay? Would recommend simple RC circuit on each coil to delay the following relay from activating, then maybe another contact assembly in each relay to deactivate the previous relay.

 

I also see you are calling for +5v source for your LEDs. As you may be aware LEDs have certain operationing specs that must be met to light. Industry standard LEDs, both panel indicator and high output require 20mA to light. Some can operate up to 50mA with limited duty cycles, but below 20mA most will not light at all or only dimmly. They also have a Forward Voltage rating simular to the .6V drop seen on a standard diode. Due to different chemicals required to make different colors, most have different FV ratings in the 2.0-3.5v range.This means each LED will drop as example 2.0v in the 20mA circuit.

Standard resistor for 2.0v-2.2v LEDs in a 12v circuit is 470 ohms. If all LEDs in your chevron are in parallel, your circuit will require about 2.3 volts(use 470ohm resistors on 12v) and run about 60mA(3 leds) per chevron. If your LEDs are in series to each other, each LEDS will drop about 2.2 volts times 3 leds =6.6volt drop plus your signialing diodes drops and you will need to supply about 7.5volts with total circuit current of 20mA per chevron.

 

Have you considered duty cycle ratings for you relay contacts?

 

Finally a question, in your pic rear view MC, looks like you have yellow under red lenses. Is this LEDs?

 

The relay version was my way of laying out the circuitry. I would like to convert this concept to a digital format. I haven't had the time to get very far with that.

 

The 5v notation on the LED's will go away, I had planned on using 12v LED's. That is why I mentioned doing away with the R1 & R2 resistors. When I convert to digital, though I don't know if I will need 12v or 5v LED's.

 

Circuit only shows one LED in each instance. In actuality there will probably be 3 or 5 to create the chevron pattern.

I have a modified version of the circuit that does not require the slow pick relay function, but it eliminates the flasher from sequencing the bulbs out in the trailing direction.

 

The yellow color in my rear end picture is from the flash, both the tail light and turn signal lenses are red.

 

The relays can be made to slow release by adding a capacitor across the coil. The coil itself will act as a resistor. Cap value will be determined using specific coil resistance.

 

Don't know how far I will get with this. I am working at a company that has a lot of resources for developing this type of application though.

 

Gary

[bkuhr]

Now that I decided to not try to dim LEDs and just leave them bright for markers and flasher, I started working on the demand, autocancel circuit.

This required in depth study of MC schematic to figure out how to interface new circut with MC.

 

My plan was to remove existing MC flashers, and tap new circuit into those connectors without damaging any stock MC wiring.

Easier said than done!

 

Attached PDF shows my best effort. But requires me to seperate the wiring at the hazard switch and the turn switch from stock MC wiring. I can do this at in harness connectors, but not sure if replacement MOLEX connectors are available in both male and female to create jumper harnesses to be installed between existing connectors and avoid any stock wire cutting.

 

On attached PDF I reduced stock wiring schematic to just affected wiring involved with stock wiring to left of all MC components. To the far right are necessary inputs for new circuit.

 

Does anybody have a source for molex connectors, and know how to identify specific connector types/parts numbers required to match existing connectors on MC?

 

Does anybody have any suggestions to modify my best effort pdf, preferable to avoid above need to modify stock wiring?

 

Questions/comments/suggestions welcome

 

Brian

[dingy]

Brian,

 

Would the AutoCad files for the 83 bike schematics help you any?

 

If they will, send me your email address and I will send them to you.

 

I can convert to DXF if you have another cad program.

 

Gary

[etcswjoe]

Did either of you look at Bar/Dot drivers such as the LM3914? I had thought about using one and splitting an led bar for an additional turn signal.

[bkuhr]

Brian,

 

Would the AutoCad files for the 83 bike schematics help you any?

 

If they will, send me your email address and I will send them to you.

 

I can convert to DXF if you have another cad program.

 

Gary

Thanks Gary, but I don't think so. I have an Autocad(98) program on my desktop(which I seldom use anymore) but it wont load on my Vista laptap.

I am using electronic design software called KICAD. I am learning the software but it is supposed to let you build a schematic, then it will automatically design a PCB design(even multiple layers) with input to component locations, then output photo resist ready transparencys for use on photo resist pcb's.

 

I do like and appreaciate all the CAD work that you do, and sure would make it easier to provide visual descriptions when helping someone. Maybe sometime I will purchase a new CAD program, but low on $ list.

Thanks anyway,

Brian

[bkuhr]

Did either of you look at Bar/Dot drivers such as the LM3914? I had thought about using one and splitting an led bar for an additional turn signal.

 

LM3914 is a good thought I had not considered, This chip senses an adjustable voltage to drive up the led strip simular to the analog bar graph on the bottom of a digital multimeter. I figured it was easier to drive a circuit with a clock to progress the led strip. When I was at Fort Lewis I build a LM3914 circuit and wired it accross this signal strength meter on my CB. Our CB group used to have CB "rabbit hunts"(hide and seek). Bar graph helped out much in finding a CB transmitter! Maybe idea for Ventureriders?

 

I like idea of additional turn signal, ( may have to design power output to drive aditional turn signals in my design) but there are a lot of ready made LED Bar Strip lights available for less than they can be made with descret components. Saw some at AUTOZONE the other day, and they even have a built in battery for testing on the shelf. Pretty neat.

 

Keep the ideas coming

 

Brian

[dingy]

Brian,

 

Attached is a ladder logic diagram & a Boolean diagram for my relay circuit version.

 

Don't know if they will help you any.

 

The Boolean has the symbols hatched that are already on bike.

 

These are also in AutoCad.

 

Gary

[bkuhr]

Brian,

 

Attached is a ladder logic diagram & a Boolean diagram for my relay circuit version.

 

Don't know if they will help you any.

 

The Boolean has the symbols hatched that are already on bike.

 

These are also in AutoCad.

 

Gary

Gary, have you figured out who to interface these with your venture?

[dingy]

Gary, have you figured out who to interface these with your venture?

 

All the diagrams I have posted are shown hooking into the existing turn & hazard switches with no modifications other than one jumper on hazard switch input that can be done in the wiring harness.

 

I will need to cut & splice some wiring, but I don't see any way around that.

 

The main problem is that the turn signal switch puts a constant 12v signal out unless it is manually canceled. This coupled with the actual initiation signal being on the ground side complicates this.

 

Gary

[a1bummer]

Looks good. You have a good start going, you'll get the bugs worked out sooner or later. I've done similar to my VR as well but I didn't go with arrows like you plan on doing. http://www.venturerider.org/forum/showthread.php?t=51686&highlight=turn+signals

 

I had to use a simple relay to turn the momentary signal from the turn signals into a constant signal to drive them. Plus my auto cancel still works as it should that way.

 

Bill

[bkuhr]

Looks good. You have a good start going, you'll get the bugs worked out sooner or later. I've done similar to my VR as well but I didn't go with arrows like you plan on doing. http://www.venturerider.org/forum/showthread.php?t=51686&highlight=turn+signals

 

I had to use a simple relay to turn the momentary signal from the turn signals into a constant signal to drive them. Plus my auto cancel still works as it should that way.

 

Bill

 

Very cool, I like. Any specs on how you did it?

[a1bummer]

Ya, no problem. I'll have to get them to you later tomorrow though. I've been busy trying to wreck a couple computers today and now I have a big mess to clean up first.

[bryan52577]

You know you could buy one for under $30 bucks?

http://www.towmart.com/sequential-led-turn-signal

 

"http://www.youtube.com/watch?v=4X8dtg0X_-A&hl=en_US&feature=player_embedded&version=3"

[a1bummer]

You know you could buy one for under $30 bucks?

http://www.towmart.com/sequential-led-turn-signal

 

"http://www.youtube.com/watch?v=4X8dtg0X_-A&hl=en_US&feature=player_embedded&version=3"

 

Now that would just be WAY to simple!

 

Here is the schematic I followed, and some scans of the masks I used to create the actual printed circuit boards (PCB's) I used. I Used a free program called ExpressPCB I found here (http://www.expresspcb.com/) to lay out my traces for my PCB's. I later found another program called PCB123v3.3.2 here (http://www.sunstone.com/pcb123/benefits.aspx) that I think will work better. I then went to these guys (http://www.philmore-datak.com/index.html) for all my supplies to make them. I also included their catalog (No. 12-760-A(found on bottom right corner of pg. 270 of their online catalog)) that might help you. I used the Positive method. I got my LED's from http://www.superbrightleds.com/cgi-bin/store/index.cgi?action=DispPage&Page2Disp=%2Fled_prods.htm, my relays and other components, like my IC's that I couldn't get locally, from https://www.allelectronics.com/index.php. There's also http://www.digikey.com/. But if you get their free catalog, IT'S HUGE!

 

I have the rear signals done and installed. For the front I have to make a small PCB with about 4 or 5 LED's in a series for the running lights that will be mounted over the center of the PCB's I've already made.

 

I had to put everything into "Zip" files to get it to upload to the site.

 

Any questions feel free to give me a pm.

Bill

[bkuhr]

Now that would just be WAY to simple!

 

Here is the schematic I followed, and some scans of the masks I used to create the actual printed circuit boards (PCB's) I used. I Used a free program called ExpressPCB I found here (http://www.expresspcb.com/) to lay out my traces for my PCB's. I later found another program called PCB123v3.3.2 here (http://www.sunstone.com/pcb123/benefits.aspx) that I think will work better. I then went to these guys (http://www.philmore-datak.com/index.html) for all my supplies to make them. I also included their catalog (No. 12-760-A(found on bottom right corner of pg. 270 of their online catalog)) that might help you. I used the Positive method. I got my LED's from http://www.superbrightleds.com/cgi-bin/store/index.cgi?action=DispPage&Page2Disp=%2Fled_prods.htm, my relays and other components, like my IC's that I couldn't get locally, from https://www.allelectronics.com/index.php. There's also http://www.digikey.com/. But if you get their free catalog, IT'S HUGE!

 

I have the rear signals done and installed. For the front I have to make a small PCB with about 4 or 5 LED's in a series for the running lights that will be mounted over the center of the PCB's I've already made.

 

I had to put everything into "Zip" files to get it to upload to the site.

 

Any questions feel free to give me a pm.

Bill

 

I congratulate you on a great design, and your effort to created a seperate control at each corner, as opposed to my effort to create a central control and rewire/modify vehicle, is defenitly simpiler, and the cost/components difference is likely minimal.

I have made photosensetive boards before, and was planning to do so again.

Gives me something to consider.

Thanks

Brian

[Snaggletooth]

I love it when the wiz kids talk............makes me feel so ..........alone.