# Bicycle Turn Signal Electronics needed.

#### Ernst

Aug 18, 2012
13
Hello Everyone!

I am a bit of a fish out of water when it comes to circuit design. I did a bit of tinkering in my 20s' but I am now in my 50s' and wouldn't you know it in need of a bicycle turn signal unit.

Here is the story behind this quest.

I gave up a 9 miles per gallon 4X4 Dodge for a 100 MPG motor assisted Bicycle
Shown here in an earlier form

Now I went for the deluxe package and added all the Motorized Bicycle Bling Bling (MAB) including expensive wireless turn signals which fail on bumpy roads or when RF interference blocks the strobe signal between the front and rear unit. The Failure rate is 100% for the jostling and road shock at 30 MPH.

So I wish to use an external flasher circuit with a handy thumb switch and I lack the skills to design and build one. I need help.

This project will be shared with other MAB riders who have paid the $80 for a Bicygnals unit and find it is not a Safety device as we hoped. I have a thread going on a Motorized Bicycle Web Site http://motorbicycling.com/showthread.php?t=32046 At First the solution seemed to place the electronics for both units so close to each other and up front where they could "talk to each other" and I could reset them as needed but that idea faded and I am now looking at an unfinished project and friends wondering what happened. I intend to "Hard Wire" these units with the units being dumb displays and the electronics seperate. I found a few things as I searched and this site was one but a fellow on YouTube seems to have made a circuit I like the Transistor idea. I do understand Transistors and operational amplifiers in general terms. I know the difference between Capacitors, Resistors and Diodes. I know what a 555 timers is and have some understanding on basic ICs. I have etched a few boards in my day as well. I am handy with a dremel and can solder when I can find my good glasses I sure could use some help and in turn I will help others get some use out of their expensive investment and add a needed safety device to our Motorized Bicycles. My Photos of the dissasembled Bicygnals unit are here http://motorbicycling.com/album.php?albumid=439 I wish to simply route the LEDs out of their electronics and power them with an external circuit. Now it has a Head Light and Tail Light so I could use them but would power them seperate perhaps even from the running engine with a small 6V generator added to the magneto. That is a seperate thing I assume Possibly a seperate battery and simple switch situation. Thank You for your time and thank you Steve. Ernst Last edited: #### Ernst Aug 18, 2012 13 I thought to start. I assume then the heart of such a device is a switch which turns on and off at regular intervals and that function would be the domain of the 555 Timer chip. Wikipedia has a nice page http://en.wikipedia.org/wiki/555_timer_IC Now what form of use would I want? Bistable ? Perhaps it is the Astable? Opps perhaps it is the Monostable mode? Which is the one I am looking for ? Last edited: #### dahhal Jul 17, 2012 58 congratulations on the switch of transport. I'm a s/w engineer by trade and can't help with the circuit design. I'd be asking the manufacturer for an explanation on the faulty product. is it faulty design, or production? I don't believe a 2.4 GHz signal should be stopped by vibration at speed. is something loose, is it a possible short? #### duke37 Jan 9, 2011 5,364 If you are hard wiring the lights, this will be very simple. You could use a 555 or a 4093 to make an astable. Then you will need an amplifier to pass the light current. A power fet would do this. Two circuits, each activated by a button or a centre off switch would do. The main problem will be to protect the electronics from poor stability of the supply #### Ernst Aug 18, 2012 13 congratulations on the switch of transport. I'm a s/w engineer by trade and can't help with the circuit design. I'd be asking the manufacturer for an explanation on the faulty product. is it faulty design, or production? I don't believe a 2.4 GHz signal should be stopped by vibration at speed. is something loose, is it a possible short? I am replying in order: It seems to be a common issue with owners of this unit and in general I recall that the manual states operation is not guaranteed. As for why ? Perhaps the engine and magneto or the front light which can disrupt the speedometer which also intercepts some pulse from a sensor not hard wired and is located in-line with the wheel spinning; that is the main issue. However, this needs to be a safety item and dependable. Travelling at 30 MPH in traffic doesn't allow for looking back or fidgeting with anything in my opinion. So simply put I am not alone with complaint. I assume many have simply tossed the sleek looking device for non-performance. As for S/W design and engineering I too write and design but not as a job. I am working with the Collatz family of dynamic equations in an attempt to compress the Million Digit File to win the challenge. Kook I am not. Armature standing I have. Nice to meet you! Have a laugh at http://marknelson.us/2006/06/20/million-digit-challenge/ BUT, I promise you I am encoding 40 bits to 36 bits here and looking to encapsulate the whole file with 39 bits for the 40 bits sampled. Fingers crossed for sure. There has to be a way and yes I found "A Method." Again Nice to meet you. Enjoy the folly! #### (*steve*) ##### ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator Jan 21, 2010 25,505 The main problem will be to protect the electronics from poor stability of the supply And water! #### Ernst Aug 18, 2012 13 If you are hard wiring the lights, this will be very simple. You could use a 555 or a 4093 to make an astable. Then you will need an amplifier to pass the light current. A power fet would do this. Two circuits, each activated by a button or a centre off switch would do. The main problem will be to protect the electronics from poor stability of the supply Cool. May I have your guidance? I will wish to document and share this solution for others. I found this device an excellent brake light but the idea of an accelerometer as the AI of when one is stopping is beyond logic. However, once I watched the video of how it is in Taiwan for bike riders at night then it hit me that hitting a bump or slowing down; perhaps coughing hard, could be a goodly reason to alert an approaching driver something is different for you relative to some local logic. I took the item and made it into a wonderful brake light! The LED and lenses make it so bright that it is obvious in day-light and an excellent safety device! This is a place where folks will try and pass in less than 50 feet to a stop sign when I am travelling at 30 MPH because they "React" to a bicycle. More than once I have had a mad driver sitting on the yellow dividing line making faces at me because they misjudged what they were doing! Here is that brake light modification thread. You can see I am trying to be helpful with my threads. http://motorbicycling.com/showthread.php?t=31336&highlight=ernst Oh and there is a picture of the rear Bicygnals unit about page 2.. Nice looking rear unit! Would like to talk about power using the 6V generator for the LEDS for the centre rear red and the front White LEDs too later if you might That would be a bonus not expected. So you say you understand how. Good. I believe I can follow instructions well. Are you up to that? I am at that level. Developing a procedure for myself and others to follow would be nice. Are you comfortable with the tutelage? I will appreciate the help. Ernst #### CocaCola Apr 7, 2012 3,635 Why go wireless in the first place? Switch and logic hardwired together in or around the switch and a few hardwired to the lights and you have a rock solid design... This whole thing can be done with a micro and a few transistors in a very tight space, pot it in epoxy and it would be weather and vibration safe... Also if 'reliability' and 'performance' is really the goal, you should consider doing a professional run of professionally populated and built PC boards... As soon as you introduce inexperienced builders (solderers) and prototype build supplies (perfboards, point to point wire) your failure rate skyrockets and you are back at square one... #### KrisBlueNZ ##### Sadly passed away in 2015 Nov 28, 2011 8,393 Hi Ernst. I've designed something for you to get the detailed discussion started. This is a controller module that may do what you want. Please have a look at it, and my description, and tell me what changes, if any, you want to see. Also anyone else on this thread, please add your suggestions and comments. Here is a circuit description. The circuit receives power from an automotive 12VDC supply at the top right corner. This supply is filtered by LD and CR and further filtered by RD and CD. +12V emerges at CN1 and is fed to the indicator lever switch, and the hazard switch if present. The returns from these switches go positive when the switch is closed to the appropriate position. When the hazard switch is closed, DH1 and DH2 inside the controller simulate both the left and right indicator switches being closed simultaneously, so both indicators blink. U1 (the ubiquitous 555 timer) is configured as an astable (oscillator) with roughly 50% duty cycle. Feedback and timing is through RT and CT. Normally QE is OFF, and RS and RT form a voltage divider that pulls 555 pins 2 and 6 above the high threshold (which is 2/3 of the supply voltage), preventing the 555 from oscillating and forcing its output (pin 3) low. When either indicator switch is closed, current flow through RL and/or RR into QE base turns on QE. DS is now reverse-biased and the 555 starts to oscillate. Initially its output is low. This guarantees that the indicators illuminate as soon as the lever is activated. The first blink is somewhat longer than subsequent blinks because RS holds CT charged somewhat above the high voltage threshold. With an indicator switch closed, current flows through DL and RA, and/or DR and RB, into QA and/or QB. When the 555 output is low, the transistor conducts and applies bias to the gate of QL and/or QR, the main switching MOSFETs, which pass the +12V supply voltage through to CN7 and CN8, which are connected to external LED panels which are returned to 0V. If the hazard switch is not needed, CN2, DH1 and DH2 can be omitted. QL and QR each carry the current for the indicator panel(s) for one side, and must be rated for the appropriate current. Several possibilities are suggested on the schematic. Most of these are TO-220 devices rated at several amps or more, and 60V or more. LD must also be rated to carry the maximum needed current - this is equal to the total load of all the LED indicator panels, unless the hazard feature is not used, in which case it is the total of the LED indicator panels for either side. The voltage on CN7 and CN8 that supplies the LED panels is very close to the 12V automotive supply rail at the input to the controller. Multiple LED panels can be connected in parallel. Each will see nearly the full supply voltage. A typical panel consists of several series strings of LEDs with a single current limiting resistor each, as shown in the schematic. The forward voltages of the LEDs in each string is summed, and the total is subtracted from the nominal automotive supply voltage, to calculate the voltage across each current limiting resistor. The value of each current limiting resistor is then calculated from R = V / I where V is the voltage across the resistor and I is the desired LED current (in amps). Multiple strings can be paralleled; the currents will add together. Aim to keep several volts across the current limiting resistors, otherwise variations in supply voltage may noticeably affect the LED brightness. The schematic should also have 1N400x diodes from QL and QR drains to the 0V rail (anodes to 0V). The next version of the schematic will have these shown. #### Attachments • Ernst.gif 15.2 KB · Views: 1,189 Last edited: #### Ernst Aug 18, 2012 13 And water! Why go wireless in the first place? Switch and logic hardwired together in or around the switch and a few hardwired to the lights and you have a rock solid design... This whole thing can be done with a micro and a few transistors in a very tight space, pot it in epoxy and it would be weather and vibration safe... Also if 'reliability' and 'performance' is really the goal, you should consider doing a professional run of professionally populated and built PC boards... As soon as you introduce inexperienced builders (solderers) and prototype build supplies (perfboards, point to point wire) your failure rate skyrockets and you are back at square one... Thank You. I have no goal of any production of this effort beyond my personal benefit. I am making that very clear. I'm not asking for free engineering of a complex system from what I can tell. I assumed, with the YouTube video, that the circuit would be rather elementary ( for those smart enough ) and that it would be rather generic. I trust this assumption is true. Again I am not in any way utilizing your aid to facilitate commercial profit. We should endeavour to keep it basic and simple for the same reason. I would hope and believe that those who attempt this will have the interest and hopefully seek advice and guidance through the thread on the Motorized Bicycle site I have started once we have a good project. Through that hopefully enough people and or posts will provide. Well, it is the safety of the riders that we can focus on if that helps us motivate with clear conscience. I mean that in a serious and yet a humorous way. I know there is a line being crossed of "doing someone's homework" so I welcome anyone to visit and see my bike. I can provide new photos and share any aspect of my bike's construction. It's the least I can do. I truly am dependant on help in this case. Also this may help others. NO DOUBT; that buying up these devices minus the main electronics and making them a "Hard-Wired" device might sell some thousand units but I figure the company that produced them believed Wireless would reach a larger market even if it wasn't good for Motorized Bicycles which they may see as a niche market. I'd be guessing to say exactly. #### Ernst Aug 18, 2012 13 Hi Ernst. I've designed something for you to get the detailed discussion started. This is a controller module that may do what you want. Please have a look at it, and my description, and tell me what changes, if any, you want to see. First off Thank You KrisBlueNZ! Actually, Simply having a switch controlled left or right blinking circuit is sufficient. That also bodes well in having others manufacture their own. Would cut down on project complexity. When it comes to "Hazard" I pack a neon green vest and "Glow Sticks" in case I am changing a tube on the side of the road at night. My Bike has a rear facing "head light" which I put in flash mode at night to aid me in alerting drunk drivers who can see flashing white light sooner than flashing red, in my opinion and to alert anyone else. Also with the distances between the orange "Turn Signals" on the unit it would make little difference if only one is flashing over two flashing except in battery consumption. With my motorized bike, I have a kit of tools and parts that allow me to fix most common break-downs such as flat tire and magneto fail which happens when the engine runs at higher than normal rpm. Personally I believe the magneto is designed to fail that way since it takes it 1/2 mile to fail after what we can call an over-speed even when the bike has sat for 8 hours in between covering the actual (1/2 mile) distance. Honestly; I found the complexity of the electronics of the Bicygnals unit to be over-kill to be honest especially in light of a lack of utility for Motorized Bicycles (MABs) I would assume we realize that this is battery based so 12V is awkward. I Saw in the YouTube video that a 9V battery works. I am open to buying a more advanced battery such as a rechargeable type which can supply the proper power. Perhaps a converted "Bike Light" battery system would be the best? On the other hand that schematic seems a logical starting place from my point of view yet, I cannot suggest modifications beyond the request of simple left or right flash and a thumb switch activation. Last edited: #### KrisBlueNZ ##### Sadly passed away in 2015 Nov 28, 2011 8,393 Hi Ernst, Putting the switch inside the controller, and using an internal battery, and removing the hazard flash option, all simplify the circuit a lot. This circuit will work from any voltage from 6V to 12V. The absolute maximum is 15V (exceed that and the 555 may be damaged). The only calculations you need to do relate to the LED panels you use for your indicators. I've drawn these showing three chains of three LEDs each, but that's just representative. The factors that come into play are: 1. Battery voltage. The full battery voltage will appear across each LED panel, but bear in mind that the battery voltage drops slightly as it discharges. 2. LED current. The brightness of each LED is roughly proportional to the current flowing through it. Different LEDs have different efficiencies, of course. Each LED also has a maximum current specification. 3. LED forward voltage. This is the voltage that the LED will drop when it's operating at your desired current. It is a characteristic of the LED you choose. Different types and colours of LEDs vary quite a lot in this respect. 4. How many LEDs you want in each panel. Each panel must have one or more series strings of LEDs. Each string must have one or more LEDs, and a series resistor. All strings should be the same. Each string will operate at your chosen current. The LEDs in the string will have a combined voltage drop equal to the individual LED forward voltage (at the operating current you have chosen) multiplied by the number of LEDs in the string. This voltage should be significantly less than the battery voltage; if it's too close, changes in battery voltage will noticeably affect the LED brightness. You choose the series resistor using Ohm's law rearranged to R = V / I where V is the voltage across the series resistor (the battery voltage minus the total series LED voltage) and I is the current (in amps) that you want to flow in the LEDs. For example say the LEDs are operating at 50 mA and at 50 mA they have a forward voltage of 2.6V each. Assume you're using a 12.0V battery. Three LEDs in series gives 7.8V. That's a long way from 12V so try one more LED in the string. Four LEDs in series gives 10.4V. That leaves 1.6V across the resistor, which is good. You can't use five LEDs in each string because the total forward voltage would be 13V and you only have 12V available. So with four LEDs the resistor has 1.6V across it, and you want the current to be 50 mA. R = V / I which is 1.6 / 0.05 which is 32 ohms. Closest preferred value is 33 ohms or you could use 27 ohms to get a current slightly higher than 50 mA. Each string has four LEDs and draws 50 mA when lit. So if each LED panel has three strings (12 LEDs total), it will draw 150 mA. If you have the same front and back the total load will be 300 mA when the LEDs are illuminated. Those figures are probably a bit low. You should get samples of several types of LEDs with suitable colours and see how visible they are at different operating currents. #### Attachments • Ernst.gif 11.8 KB · Views: 2,040 #### duke37 Jan 9, 2011 5,364 Here is an alternative using a 4093 CMOS chip. Note that neether of the light connections go to the frame. If you pot the circuit in resin, put tails on the wires where they enter the resin so that they do not break there. #### Attachments • LeftRight.png 8.8 KB · Views: 666 #### CDRIVE ##### Hauling 10' pipe on a Trek Shift3 May 8, 2012 4,960 Kris, nice circuit. Probably just an oversight but leaving the CTL pin hanging isn't SOP. It's a HiZ input that could pick up noise. When not used a small cap to gnd is recommended on the data sheets. Chris #### CDRIVE ##### Hauling 10' pipe on a Trek Shift3 May 8, 2012 4,960 Here is an alternative using a 4093 CMOS chip. Note that neether of the light connections go to the frame.. Nice Minimalism ! Chris #### KrisBlueNZ ##### Sadly passed away in 2015 Nov 28, 2011 8,393 Kris, nice circuit. Probably just an oversight but leaving the CTL pin hanging isn't SOP. It's a HiZ input that could pick up noise. When not used a small cap to gnd is recommended on the data sheets. I've seen circuits that do, and circuits that don't, decouple pin 5. It's not really high impedance; it connects part way down the three-resistor voltage divider that sets the upper and lower threshold voltages, and the resistors in the divider are nominally 5k each - at least, they are 5k each for a bipolar (ye olde style) 555. Personally I don't bother decoupling it. I might if I was concerned with frequency stability... but if I was concerned with frequency stability, I wouldn't use a 555 in the first place #### KrisBlueNZ ##### Sadly passed away in 2015 Nov 28, 2011 8,393 Here is an alternative using a 4093 CMOS chip. Note that neether of the light connections go to the frame. If you pot the circuit in resin, put tails on the wires where they enter the resin so that they do not break there. Good advice duke. Potting resin is probably a good way to construct it. You'd have to use a fully enclosed control switch. You could probably drill and tap some holes into it afterwards, and use a pipe clamp to fasten it onto the bike. Or would the resin crack from the strain? #### CDRIVE ##### Hauling 10' pipe on a Trek Shift3 May 8, 2012 4,960 I've seen circuits that do, and circuits that don't, decouple pin 5. Well, I've been decoupling it like forever and since my spice doesn't like unterminated pins ... I did think the impedance was higher than it is though. #### Ernst Aug 18, 2012 13 Good Morning Friends. Coffee in hand and rubbing eyes here. I have had a couple of long days in a row. We are looking at August 31st for when there will be funds to purchase parts. I just got back to work so there is some delay in discretionary spending around here Vibration can be isolated enough with padded mounts or attaching the works to the front unit perhaps. I get the feeling the actual circuit board size will be reasonably small. The basic need here is to provide housing that is reasonable for all to obtain. KrisBlueNZ, again thanks. For safety reasons there is a need to have a switch attached to the handle bars. The controls must be as functional as we expect the turn signals on a motorcycle. Would the distance to the project mount if it is 12 to 18 inches be a problem with having a wired switch? Resistance? Circuit design? Allowing for a wired switch is a reasonable feature. I have gone with 1/8 inch audio plugs and can buy "splitters" at the local$0.99 store for a Dollar So I get two males and one female and are the same as seen in the photo a few posts up.
For wire they sell small gauge "speaker wire" which is nice at , again, \$0.99 .

I thought to ask if I might post links to better photos of the LED aspects of these units. They are assembled already so voltage utilization and their wiring organization are fixed perimeters.
Thankfully there are ribbon cables that make powering the LEDs an easy attach point.

In reply to 555 or 4093 CMOS I cannot judge. Naturally fewer parts and smaller footprint seems better but I will yield the direction to you guys.
I assume either part is easy for the average person to find at Radio Shack? I'm guessing most will turn to them to build a flasher unit for their MAB.
It may be wise to have a parts list they can satisfy.

As for water proofing would silicone do the trick? Can the CB be placed into silicon and more silicone be applied on top?
Silicone also may aid in vibration reduction; perhaps.

This project is looking doable! I was unsure a solution could be found but I am much less worried now. Thanks.

So, I can do some show and tell probably by Thursday evening with photos and parts on hand. Perhaps this may provide a starting point. I believe I have a project box but that is conditional; if it's the right thing to use.

I have tasks I must get done tomorrow evening so I will be delayed until Thursday evening on posting more photos.
Let me know if there are any specific shots I can take.

This doesn't get old for me "Thank You Guys!"

#### KrisBlueNZ

##### Sadly passed away in 2015
Nov 28, 2011
8,393
Sure, you could mount the switch 12~18" away from the control box, as long as the wires are well insulated. But I wouldn't use 3.5" jacks - any outdoor exposure will quickly lead to corrosion. You really need a fully airtight and water-tight plug and socket arrangement for everything.

If you make the circuit board small enough, you might be able to mount the whole lot on the handlebars, with the switch built into it. You would need small airtight/water-tight 5-pin plug and socket, with a 2-core wire for the battery and a 3-core wire to the indicator panels.

Yes it would be useful to know the details of the LEDs you want to use, and the battery voltage.

Do you mean you want to use standard 0.05" grey ribbon cable for the wires to the LEDs? I think that would be too fragile in an outdoor environment.

The 555 and 4093 circuits are the same apart from the oscillator. The 555 is smaller, more robust, and possibly more widely available, so I would go with that.

I would think covering the whole board in silicone sealant would do the job nicely. You will have to seal the whole thing, right up to the socket, which also has to seal when the plug is present.

I'll be interested to see what you come up with

Replies
38
Views
6K
J
Replies
17
Views
1K
josephkk
J
Replies
10
Views
398
Replies
7
Views
709
Replies
16
Views
752