I've attached a diagram of a circuit I want to build for my motorcycle tail lights. A similar product is commercially available, but I wanted some options that it doesn't include. I cannot use a PIC for a couple reasons, so I need to do this the "old fashioned way" .

I have the circuit for the 4093 flasher that I got from Tony van Roon's website. He credits the circuit to a Ken Moffett. It is perfect for what I want for the flash component & amplifier. The pulse unit is the problem, along with combining it all. For example - if I use a 555 & decade counter, once the wig-wag portion is finished, I figured I could use the next pulse to activate the flasher - but how do I reset the 555 so it stops trying to pulse the lights?

Comments? Suggestions?

 

Interesting! Do you have a schematic of your wig-wag circuit...and your motorcycle wiring...since I assume you want to use the turn signals for the wig-wag and the tail-light/brake light for the pulse and steady.

Where, on Tony's site, is my circuit? (just a vanity thing )

Ken Moffett

 

If drive pin 4 of the 555 IC to ground it will stop pulsing. If you drive it high again it will start pulsing.

 

Ken - I've attached my redraw of the circuit from Tony's site. I redrew it for a couple reasons, but mostly because I'm trying to become marginally proficient w/ Eagle for the final circuit draw. I did change the TIP# because Eagle did't have a 117 that I could find. Tony's site indicated suitable subs as NTE262, TIP125 or TIP127.

Here's the link (if that's okay with the moderators):

http://www.uoguelph.ca/~antoon/circ/motflash.html.

The wig-wag feature is for my dual brake lights (1157 bulbs x2 + LED sets x2 (SgtWookie helped me with that LED addition)). Adding the commercial unit is pretty common in the Goldwing community - pretty much unheard of for my style bike (a supersport touring model). Part of what makes your circuit so desirable is that the brake light wire to my separate brake bulbs splits back near the actual sockets: the plug that connects the wiring harness to the brake switch circuitry has a single wire exiting & is located in a more convienent area. Consequently I can place your module design in-line with the brake light circuit after the connection plug but before the split, then run separate output wires to the split.

mik3 - thanks. I was concerned that as long as the brake was engaged, the power would continue to run the 555, just from start again. Since pin 4 could drive the output to low regardless of the circuit state, I thought it would just start timing all over again as long as there was power to the circuit. Most schematics I've seen tie pin 4 to the main with pin 8.

Appreciate the inputs & any assistance.

 

I finally finished putting together the pulse schematic. It is a combination of what I thought was the best of several different circuits from various places on the internet. I think my math is correct for the changes I had to make. Appreciate any assistance in review or recommendations.

I'll be putting the amplifier circuit together next.

 

Just a quick observation...the outputs of the 4017 should be diode OR'ed. Put a diode in between each of the six outputs (Q1-Q5) and their summed points ("To Left..." and "To Right..."). As connected, high and low outputs are shorted together.

I'll look closer later...I need to go get coffee.

Ken

 

As Ken says, you need diodes on the 4017 outputs. The anodes go towards the 4017. You will need to use the outputs to switch a MOSFET on and off, because CMOS logic can only output a milliamp or two at best.

You need a 10k pullup resistor from the 555 pin 4 to +12v, or the timer won't start.

The 4017 won't reset unless the power to the circuit is turned off for a period of time, because you have the reset line being activated AFTER the 555 pin 4 is pulled low. Since there won't be any more clocks from the 555 after it's pin 4 is pulled low, your circuit will get "stuck" there until power is removed for a period of time. Since you have a 220uF cap in there that has a discharge path through R1, R2 and R3, that could take a while [eta] unless the brake light bulb is also being powered by the +12v line.

4017 pin 5 output looks like you have it connected to a connector labeled "To Flasher Circuit Enable." I don't know what you expect that to do.

 

An 1157 light bulb is 27W and might set your bike on fire!
If the voltage is 13.8V then its operating current is 27W/13.8V= 2A. The TIP125 has a max allowed pulse current rating of "only" 8A.
But the lightbulb is incandescent and draws 20A or more when it is cool. Then kiss the TIP125 goodbye.

 

Thanks for the reviews. I'll make the corrections & re-submit, but as Ken said, I gotta have some coffee, lol.

 

Ken/SgtWookie: The attached shows the corrections I made based on your recommendations. I chose the 1N4001 based on the specs from the 1N400x datasheet. I thought it would be sufficient since the amplifier circuit is diode protected.

I added the pin 4 power up even tho the 12v line does indeed go to the brake light.

I had intended to use the pulse from pin 5 to trigger the 4093. That will also give me a connection point into the 4093 that allows my desired by-pass of the wig-wag. I was going to connect it to a darlington to boost the current.

I'm finishing the amplifier component soon - it's pretty simple since it's taken from Ken's flasher circuit.

By the way, my solution for the reset problem was to place them on the same line. Hope that's okay.

 

On the 555, pin 4 - I don't know why you put diodes in there. Neither were necessary, and the lower one will prevent Q1 from pulling pin 4 low.

On the right side of the schematic, you have two connectors that are labeled "V-->".
This may cause problems, since they have the same name - particularly if you attempt to make a PCB out of the schematic.

The diodes on the 4017 look good. If you're going to use them to control a MOSFET gate, you should also use a 10k resistor to ground to turn the MOSFET off by default.

Use the Erc function (Tools/Erc) - you'll probably get complaints about a missing junction at the pot (just above R2).
It also looks like the wire is broken at D7. You may have had the grid turned off, or set really low.

It's a good idea to leave the grid set on 0.1 inches for the schematic. Otherwise, you may have a hard time connecting things.

Unless you run a wire all at once in Eagle, the sections of wire aren't joined. You have to manually place a junction. Alternatively, you could download "addjunction.ulp" from Cadsoft's Eagle website and run it.

You may also wish to download "autoplace_v3.ulp" for when you're getting ready to make a board. It simply places the components on the board in locations relative to where they are in the schematic. It's not ideal, but can help save you some time in figuring out what is what.
[eta]
The three "connectors" on the right side from the 4017 outputs look like they are from the supply or supply2 libraries, with their NAME fields blanked out. This will cause problems when you go to make a board.

Instead, use objects from the Wirepad library, and change their NAME fields to something meaningful.

 

Thanks for the in-put. Made the corrections & tried to clean it up a bit. Also included the amp protoype (which is nothing more than a slight modification of the amp from Ken's circuit that Tony modified).

The TIP125 is a place-holder since Eagle doesn't have the TIP117 or an NTE262.

I am thinking about the logic gate I need to make to connect the two switching circuits to the amp-ins. Have to consider it a bit.

Thanks for the help.

 

Thanks for the input. Made the corrections & tried to clean it up a bit.

It looks like you posted the same image again, because I don't see any changes. Could you update it? [eta] OK, I see you did make some changes - but you left the two diodes in by pin 4 of the 555 timer.

There's another problem that I forgot about. The 4000 series CMOS IC's in the Eagle library use VDD for V+ and VDD for GND.

Also included the amp protoype (which is nothing more than a slight modification of the amp from Ken's circuit that Tony modified).

The TIP125 is a place-holder since Eagle doesn't have the TIP117 or an NTE262.

That's OK. They have the same package and symbol type.
Click on "Value" on the toolbar to the left; looks like a Euro resistor with 10k in black under it. Then you can click on the TIP125 and change it.

You might as well go with a higher current version, like TIP135 to TIP137; they're about the same price anyway.

As far as junctions go; you don't have to use them when connecting a single wire to a component (like resistor, caps, IC's, etc.) if the wire connects to the component at an angle divisible by 90. Try to avoid running wires at odd angles, like those to X1-1 and X1-2. If you do an Erc, you'll find that it'll complain about wires being too close or needing a junction there.

You can check to see if a component is connected by just dragging it a little bit up/down/sideways. The wires will move with it if they're connected.

I am thinking about the logic gate I need to make to connect the two switching circuits to the amp-ins. Have to consider it a bit.

Well, it depends upon what you want it to do.

Are you going to use Ken's flasher circuit for your main brakelights, or just the wig-wag thing?

 

Looking at the circuit some more, R1, R2 and R3 just aren't right. R3 will keep C1 from charging up to 2/3 VCC, so the 555 timer will never cycle/output a clock pulse.

What frequency did you want to get out of the 555?

6Hz would probably work pretty well. See the attached schematic for a 6Hz 50% duty cycle astable 555 timer circuit. You can use a 1N4148 or 1N914 diode for the one shown. Using fixed resistors will make it more reliable than if you used a pot.
[eta]

OK, to make things a bit easier on you, I re-drew your two schematics as one, and incorporated the changes I mentioned.

The symbol for the 555 timer was changed to that of the Intersil ICM7555 timer, which is pin-compatible with the LM555N, but it's easier to use for this particular layout. You'd rotated the LM555N symbol upside down, which was going to make it awkward to wire up, and difficult to read.

I changed all the +12V/+V symbols to VDD, and all of the GND symbols to VSS. Otherwise, you'd have a problem with wiring to the 4017 counter. Eagle will complain about VDD being connected to IC1 V+ and VSS being connected to it's GND pin, but that's OK - as long as you know what's going on.

I'm using Eagle 4.16r2, so unless you're using that older version, I won't be able to edit your schematics/layouts once you've edited/saved them.

 

SgtWookie - thanks for catching the error in attachments - i edited the post to correct to version 1c (lol - errors happen as my document folder rapidly fills because I'm keeping all the development stages so I can review them as a teaching aid.)

I added R3 to the 555 timer circuit because a problem w/ the orignial was that C1 wouldn't discharge if the power to the circuit stopped (i.e., brakes were released). That meant the original had delays in start-up or erratic behaviors when the brake was rapidly pressed. The circuit from Ken eliminated that problem so it works as designed regardless of number of or length of brake signals coming in. The intent was to discharge C1 after the current stopped at pin 8 so it would also start clean everytime.

As far as cycles, I want to get as close to 50% as possible. Additionally, if I can eliminate the small start up delay so each pulse is equal, that would be nice too. I read that if R1 is 1K it gives close to 50%. Your example looks simpler & cleaner, so I will go with it.

The intent is to combine both circuits so the lights alternate, then flash together, then go solid. As I've previously written, a commercial unit is available & is quite popular. It does offer two flash modes, but I only want one. I'm far more interested in being able to disable the wig-wag & still run the Ken's flasher or being able to by-pass the circuit entirely for stock function. I figured if I'm gonna put a 3 pin DIP in to accomplish this, then I could also add a safety circuit to monitor the voltage at the out-put end. I figured it as a simple comparator: If voltage at brake line in, then voltage at brake line out or switch to by-pass circuit.

Thanks again for the help. I'm getting there...albeit slo-o-o-owly.

 

SgtWookie - thanks for catching the error in attachments - i edited the post to correct to version 1c (lol - errors happen as my document folder rapidly fills because I'm keeping all the development stages so I can review them as a teaching aid.)

It's a VERY good idea to save them as different versions as you're developing your schematic and board. It's easy to make a mistake that's difficult to recover from, like if your schematic and board get out of synch.

Erc is your friend. You must fix "ERROR" messages, and be aware of the implications of "WARNING" messages. It can be difficult to read sometimes. If the board doesn't match the schematic, your life will be very unpleasant until you can get them back in synch, which usually means deleting things. The quickest way to get them out of synch is to add something to the schematic without having the board file open. After that, the board won't get the same updates as the schematic, and the longer it goes on, the worse it gets.

I added R3 to the 555 timer circuit because a problem w/ the orignial was that C1 wouldn't discharge if the power to the circuit stopped (i.e., brakes were released). That meant the original had delays in start-up or erratic behaviors when the brake was rapidly pressed.

I see. Did you find this out by experimentation, or was this a reported problem?

555 timer circuits generally have a longer 1st pulse due to the extra time it takes to charge the capacitor from 0v to 2/3 V+ vs the time for a normal cycle charging from 1/3 V+ to 2/3 V+. However, once the capacitor has charged above 1/3 V+, the circuit I posted will give a 50% duty cycle.

The circuit from Ken eliminated that problem so it works as designed regardless of number of or length of brake signals coming in. The intent was to discharge C1 after the current stopped at pin 8 so it would also start clean everytime.

If the brakelight is ON when the brake pedal is released, the timing cap will be discharged to below 1/3 Vcc in roughly 1/3 second. If it is not on, the discharge time will vary considerably. This could be made much more consistent by adding a small 12v bulb, such as a side marker type bulb, across VDD/VSS.

As far as cycles, I want to get as close to 50% as possible. Additionally, if I can eliminate the small start up delay so each pulse is equal, that would be nice too. I read that if R1 is 1K it gives close to 50%. Your example looks simpler & cleaner, so I will go with it.

Without using diodes, using the standard 555 timer configuration means that you can't quite get to a 50% duty cycle. The diode enables the 50% duty cycle.

The intent is to combine both circuits so the lights alternate, then flash together, then go solid.

Well, the way things are wired right now, you'll get three wig-wags (with the current RC network on the 555, that'll take about one second), and then Ken's circuit will kick in.

I don't think you should have a 1-second delay before the main brake light starts flashing; were it me, I'd want it flashing from the moment I hit the brakes.

As I've previously written, a commercial unit is available & is quite popular. It does offer two flash modes, but I only want one. I'm far more interested in being able to disable the wig-wag & still run the Ken's flasher or being able to by-pass the circuit entirely for stock function. I figured if I'm gonna put a 3 pin DIP in to accomplish this, then I could also add a safety circuit to monitor the voltage at the out-put end. I figured it as a simple comparator: If voltage at brake line in, then voltage at brake line out or switch to by-pass circuit.

A N.C. relay might be the easiest and most reliable way around that, with it's coil energized by ORing diodes from the output of the driver section. That way, if there IS output from the driver, the relay energizes and the wig-wag is allowed to function normally. If there is no output from the wig-wag circuit, the de-energized relay's NC contacts will supply power directly to the lights.

 

SgtWookie - Super review & you've given me plenty to work on. Good thing I'm not on a deadline. Especially since I'm back to work the next couple nights. Some thoughts...

When I opened Eagle for this project, it offered an update version 5. I'll have to check the web-site to see if I can get the older version.

The capicitor discharge problem was identified by the designer at Tony's site. When Ken's circuit was posted, one of the benefits was it completely eliminated the issue by placing a discharge resistor on the capicitor, so I thought I'd try to duplicate it. You pointed out that my design was problematic so I'll need to develop another solution.

You're absolutely correct about the delay - I want immediate lights, every time. Off the top of my head, I'm thinking that the easiest way to keep that capicitor ready is to provide a separate power line to the unit & use the brake switch line as the switch only. But I'd really like to keep this "plug & play" since that's the way most commercial units are designed, i.e., the brake light wire is cut & the unit is wired in series (notice Ken's design where Tony modified it at the amplifier end).

Timing wise, the 6 pulses for the wig-wag should take no less than 2 sec to complete. I figure somewhere between 60Hz & 120Hz. Definately don't want to put anyone into seizures. The two trimmers in Ken's circuit permit the necessary adjustments for the flash component, so I was trying to incorporate that into the 555 so I could adjust the rate (duration obviously is set by the 4017). But as you pointed out, this too could be problematic, so I'll be better off with a resistor of appropriate size.

I'm trying to avoid the use of incandescents (even tiny ones) entirely. My alternator is small & I've been converting everything to LEDs where possible to allow the power I need for the other gadgets I've installed (i.e., driving lights, heated gear, ECM ). That's why I've been staying away from that topic - I can handle the LED end. I just want this unit to be able to activate a stock system.

Finally, I figure the logic I need to control the input to the amp component is along the lines of: (CxNOT D)+(DxNOT C) where is C=(A+B). If any input from the wig-wag (A+B), then flasher (D) is off, & vice versa. I'm still playing with that part though to determine a good solution using DIP gate packaging.

Part of the challenge I'm discovering in this project is trying to accomplish the goal while adhering to KISS for number of components & function. We had a saying while I was on active duty: College Education to Fly it; High School education to Fix it; Doctorate's in Electronic's & Engineering to Design it that way.

Thanks again.

 

When I opened Eagle for this project, it offered an update version 5. I'll have to check the web-site to see if I can get the older version.

4.16 is still available on their FTP site. However, if you install it, you will no longer be able to edit/view anything you've created with the newer version.

If you have a 2nd computer, go ahead and install the older version on it. That way you'll be able to post things that I can edit.

I have a registered version, but I'm not willing to spend yet more money to upgrade my version to help people using the free version. It doesn't take a degree in Economics to appreciate that.

The capicitor discharge problem was identified by the designer at Tony's site. When Ken's circuit was posted, one of the benefits was it completely eliminated the issue by placing a discharge resistor on the capicitor, so I thought I'd try to duplicate it. You pointed out that my design was problematic so I'll need to develop another solution.

Actually, I haven't simulated Ken's design. Yet.

You must know that this project could have dire safety issues associated with it. You should also know that this is the Internet, and you HAVE to take everything you read with a pound of salt. I will try very hard to not steer you wrong, but I'm human, and therefore are not infallible.

You're absolutely correct about the delay - I want immediate lights, every time. Off the top of my head, I'm thinking that the easiest way to keep that capicitor ready is to provide a separate power line to the unit & use the brake switch line as the switch only. But I'd really like to keep this "plug & play" since that's the way most commercial units are designed, i.e., the brake light wire is cut & the unit is wired in series (notice Ken's design where Tony modified it at the amplifier end).

Don't make it more complicated than it has to be. KISS rules. After all, you need to be able to fix the freaking thing.

Timing wise, the 6 pulses for the wig-wag should take no less than 2 sec to complete. I figure somewhere between 60Hz & 120Hz.

6Hz = 6 cycles per second; that's the 555 timer's speed. The wig-wag will complete in just over 1 second, then hold.

Definately don't want to put anyone into seizures.

I'm not a Doc, but it is my understanding that seizures in those whom may be prone to such can be triggered at around 14Hz; 6Hz should be quite safe.

The two trimmers in Ken's circuit permit the necessary adjustments for the flash component, so I was trying to incorporate that into the 555 so I could adjust the rate (duration obviously is set by the 4017). But as you pointed out, this too could be problematic, so I'll be better off with a resistor of appropriate size.

If you want to be able to adjust the timing of the wig-wag, that could be done for bench tests. However, I do not recommend using pots in your final design, as they are nowhere near as reliable as fixed resistors.

I'm trying to avoid the use of incandescents (even tiny ones) entirely. My alternator is small & I've been converting everything to LEDs where possible to allow the power I need for the other gadgets I've installed (i.e., driving lights, heated gear, ECM ). That's why I've been staying away from that topic - I can handle the LED end. I just want this unit to be able to activate a stock system.

OK. So use a very low-wattage 12v lamp. The side-marker lights I mentioned are very compact, and would last a heck of a long time in this application.

Finally, I figure the logic I need to control the input to the amp component is along the lines of: (CxNOT D)+(DxNOT C) where is C=(A+B). If any input from the wig-wag (A+B), then flasher (D) is off, & vice versa. I'm still playing with that part though to determine a good solution using DIP gate packaging.

I'm too tired to eval that at the moment. That's not a cop-out, it's a fact.

Part of the challenge I'm discovering in this project is trying to accomplish the goal while adhering to KISS for number of components & function. We had a saying while I was on active duty: College Education to Fly it; High School education to Fix it; Doctorate's in Electronic's & Engineering to Design it that way.

OT: What branch?
Thanks for your Service. As of yesterday, I'm 233 years old, and I feel every bit of it.

I have a somewhat unique experience in avionics, since I've worked on them in extremely adverse conditions, such as -50F on a flightline somewhere up north, and on the flight deck of an aircraft carrier under lights out conditions while there were 45kt winds over the deck.

But I've also worked on most of the US inventory's systems in laboratory settings.

You don't necessarily need a Doctorate's degree to design it - just a good solid background. You don't even need a HS diploma to fix it, if it's designed properly.

But if you want to be a pilot in the service nowadays, you must have that degree.

My Grandpa didn't have a degree in WWII, but became a pilot in the Marines.
He flew F-4U Corsairs, shot down 18 Zeroes and two Betty bombers, confirmed.

 

Appreciate the in-depth response. Just wanted to fire off a quick note before sleep (I work 14+ hour shifts counting travel time).

(I gotta learn how to include the quote blocks one of these days, lol.)

Ken's circuit has been in use on a couple different motorcycles now for a couple years w/ no complaints at all, per various motorcycle forums & Tony's site. Also, the original circuit it replaced as an upgrade (a straight 555 type flasher) is also in use. So I have some confidence in Ken's circuit. As you know, there are a ton of LED alternating flasher circuits out there, but little for 12V. If I could have gotten by w/ a simple transistor astable, etc. That said...

Safety is #1. Everytime I ride, I have a pre-flight routine. Whether I'm going 5 minutes up the street or an hour to work or a rest stop on a 12 hour cross-country - same routine, same way, everytime: tires, mechanical, electrical. I also ride what I refer to as European style - full armor (toes to neck), full face helmet - year round. That said, I know that if I can buy one of these flasher units for $x, I can probably build one that is better suited to my needs & has the features I want - and do it all for less $$. And learn something in the process (again, a stepping stone for what I really want to build).

Thanks for the service you performed also, as well as your grandfather (guess this is a good day for it too). 21 years USAF here. Weapons guy. Started on fighters, went into Air Rescue, then Special Ops. Finished in management. Complete change of careers when I retired. I never did an aircraft carrier in 45kt, but have spent an awful lot of time on flightlines or LZs all over the world in conditions that those who haven't been there can't appreciate. Obviously you know. BTW, the F4 is one of my favorite planes, although I'm really a WW I fanatic. When I was flying R/C, one of my last scale models was an F4U - what a bear to land! But that's another story & not for here.

Thanks so much for your time & in-put. I want to do this right. Even if I had all the knowledge to do it solo, I'd still want peer review and in-put. But for now, I'm off to grab some sleep.

 

SgtWookie - I loaded Eagle 4.16r2 & all up & running. ERC displays the 3 expected warnings from your redraw of my schematics, so we should be ready to go. Anyway, now that I understand what was meant regarding the layouts in Eagle, my posts should be of a more appropriate quality.

Apologies for the 60-120Hz comment in previous post: I knew it was 6 to 12Hz max, but fat fingered it. I'm gonna play with the astable & perhaps 4.5 or so would be closer to the rate I want.

If I get up in time, will try to add Ken's circuit into your redraw. Then we can go from there.