Adding led brake lights

[tourist]

I need to build a circuit to detect when the brake lights are on and the turn signals on.  The brake light and tail light are the same bulb.  The tail light uses 2.4 volts then when the brake is applied the voltage increases to 13.8 volts.  This circuit seems to work on my test board.  When the input is 2.4vdc the leds are on at about half the brightness and when the brake would be applied and the input is 13.8 volts the leds are on full.  The input only has two values: 2.4vdc and 13.8vdc.  These components are just what I had lying around and I think I would use a different op amp in the final circuit because I will need to monitor 4 different lights, (2 brake and 2 turn signals)  The turn signal circuit differs by reversing the 10k and the 100k resistors.  The turn signals input is either 0v or 13.8 vdc.
I read out here that the 741 isn't good to use in automotive applications, idk.  I'm open to better components or design but I can't draw too much current from the light circuit because it is monitored in some fashion by the onboard computer.  Any help would be appreciated Thanks.

[audioguru]

Your LEDs are upside-down.
The 741 opamp's inputs do not work if they are less than about 3V and you have one input at only 1.25V.

Here in Canada it is illegal to change lighting on vehicles because too many kids used extremely dim and narrow-angle LEDs that cannot be seen at night.

[tourist]

Your LEDs are upside-down.
The longer leg of these leds (not the one that are actually going to be used) I connected to ground.  I assume that the leg is the anode.  That is the way the circuit is wired and sort of working.  I figure if the led didn't light I would just reverse the lead wires.

The 741 op amp's inputs do not work if they are less than about 3V and you have one input at only 1.25V.
I have seen other posts where you have stated that also.  while the circuit appears to work, with no input on pin 3 (+) the leds are light dimly which may be wrong but with the actual "running" input of 2.4v the output of op amp was about 5v and the leds are on at about half brightness.

Here in Canada it is illegal to change lighting on vehicles because too many kids used extremely dim and narrow-angle LEDs that cannot be seen at night.
I don't know if its considered changing the lights.  I am not removing or modifying the factory lights.  I'm adding a set of supplemental lights to my saddle bags. 

So, is there a better op amp to use?  I thought since there were four different inputs, left turn, right turn, left brake light and right brake light I could use a quad op amp package in the circuit instead of the 741 but I'm still just seeing if it can work.  If there is a better suggestion I can try it but these are simply the components that I have at my disposal to test with.  So if the design is ok than I can move on but I don't think that I'm there.

A goal is to have this fit on a 1.5" by 2" circuit board and there is not alot of room where the circuit will be mounted.

[audioguru]

An LM324 quad opamp has inputs that work as low as 0V and has plenty of output current to drive transistors. An LM358 is a dual opamp with the same opmps.

[tourist]

Well, I thought about that.  Got a CA324E that I tried but didn't get any output.  So I went out and bought the lm324 just in case my ic was bad.  The new 324 didn't work either.  This is the circuit, for the most part.
Now I only hooked up one op amp from the package and left the other pins open.  Will that make a difference?

[audioguru]

Again, you drew your LEDs upside-down. The cathode (the line) is supposed to connect to the resistor to ground. The anode is supposed to connect to the transistor.

Your "VCC/2b" has its resistors backwards.

Unused opamps should be properly disabled so they don't oscillate and cause problems but maybe your backwards parts are the cause of your problems.

[tourist]

I think its getting there.  I'll fix the leds symbols in the next circuit diagram.

Are the 1k resistors on the base needed or should they be a different value?
Are the 1k resistors on pins 3,5,10 and 12 needed  or of a different value?
I drew them in there because I saw them in similar circuits.

I reversed the vcc/2a and /2b resistors because on the turn signal portion I didn't want output until a higher voltage.  Would higher value resistors be better?

Should I or do I need to add some pull down resistors to pins 3,5?

Since these are extra lights on the side bags.  When I remove the bags the leds will no longer be in the circuit.  Do I need to add anything to the circuit for when the leds are removed?

[audioguru]

Are the 1k resistors on the base needed or should they be a different value?

The transistors in your circuit are emitter-followers with a very low base current so the 1k resistors can be replaced with wires.

Are the 1k resistors on pins 3,5,10 and 12 needed  or of a different value?

They can be up to 33k ohms but each pin needs an additional resistor to ground as an attenuator so that the input voltage never goes higher than the supply voltage minus 1.5V.

I reversed the vcc/2a and /2b resistors because on the turn signal portion I didn't want output until a higher voltage.  Would higher value resistors be better?

You don't need opamps nor transistors for the turn signals. When the turn signal voltage is 13.8V then it can turn on the LED for that side.

Should I or do I need to add some pull down resistors to pins 3,5?

You don't need 4 opamps since only two opamps from an LM358 dual opamp are needed for the brake LEDs.

Since these are extra lights on the side bags.  When I remove the bags the leds will no longer be in the circuit.  Do I need to add anything to the circuit for when the leds are removed?

A weatherproof jack.

[tourist]

Got rid of the Vcc/2b.

Changed Vcc/2a resistors.
I put the pull down resistors on pins 3 and 5 for the turn signal side.  I think I still need to use an op amp and transistor because of the total current draw from LEDs (7-10).  Correct me if I'm wrong, I may not have mentioned the quantity of LEDs.

The brake side I didn't add the pull down resistors because the input will either 2.4vdc or 13.8vdc.  Didn't think they were needed because there is voltage there in the operating state.

I still need to choose the LEDs that I'm going to use so I don't know if the current limit resistors are the correct size.

Am I missing something or does it look adequate?

[audioguru]

Red LEDs are 1.7V to 2.0V each. Then 5 can be connected in series for 8.5V to 10.0V. A current-limiting resistor for the turn signals will be (13.8V - 8.5V)/25mA= 221 ohms which is not a standard value. Use 220 ohms/0.25W. This is one string of 5 LEDs that draws only 25mA. If you want 7 LEDs then make one string with 4 and another string with 3 and re-calculate the current-limiting resistors.

The brake lights are dimmed by the opamp for the tail lights. Since you are dimming them by reducing the voltage then LEDs in series cannot be used.

But I don't understand why the opamps are used for the turn signals since the flasher can easily drive only 50mA more for the LEDs without any circuitry except current-limiting resistors.

 

[tourist]

This is going on a BMW motorcycle.  I could be wrong but it is my understanding that the "computer" controls these lights and monitors over current and under current conditions.  I have seen an extra 20 milliamps drawn when somebody used a relay in the circuit for their lights.  Don't know if it caused any problems after that first day.  I just know if I took my voltmeter and monitor the lights it doesn't upset the system.  I thought it would be safer to use the op amp and transistor.  I am not 100% sure what would happen if I try to draw that much current from the turn signal side.  I am not sure if anybody knows the truth so I just tried to keep consistent with the circuit.

I have built an array of infrared leds that had 5 in series with a resistor but it seemed to have a cascading effect of taking out the other strings of led.  Maybe my poor design so I again I opted for consistency.  Is it a better design, more efficient to string multiple leds?  I salvaged a sign with a bunch of leds that lit up at the same time and each one had its own resistor so I copied.

I'm still not set on the amount of leds.  I see some units that have 50 LEDs in just one disc.  The next step is to test this with the lens cover I made to make sure it is bright enough for me.

[tourist]

I'm thinking it may be bright enough.  Hard to actually tell from picture but it does look good so far.
The leds are from that display.  I think they are 3mm high flux. (http://www.superbrightleds.com/cgi-bin/store/index.cgi?action=DispPage&Page2Disp=%2Fspecs%2Fhf3-y5570.htm)  It has a 12vdc and 330 ohm resistors.  While they are not orange, they are close in color to the orange on the bike.  I have similar red leds (not the one I'm testing with in the picture).

[audioguru]

I don't know the spec's for your LEDs. If they are 2.0V and the supply is 13.8V then the output high of the lm324 opamp will be 12.4V and the output of the transistor will be 11.6V. Then the 330 ohm current-limiting resistor limits the current to 29mA and dissipates 0.28W so it should be rated at 1/2W. You could probably connect 4 LEDs in series then re-calculate a new resistor.

LEDs in series all have exactly the same current and the LEDs use the power much more efficiently making light than wasting most of the power heating many resistors.

[tourist]

Here is what I built and tested and included some real voltages and currents.  I'm thinking this is the final version and going to start on the circuit board design.


audioguru,  Thanks for your help, I hope I interpreted correctly.

[audioguru]

It looks good but I think the trail-lights will be very dim. Try it to see.

[tourist]

I did try.  Even with the colored lens I am satisfied.  They should put a warning on every led to remind you not to look directly at them.

[tourist]

Heres the circuit board for now.  Don't know if I can make those tiny traces. 
The board is actually 2" by 1 5/8".

[tourist]

Heres the led board. The lens was my first attempt, wrong color.

[audioguru]

I made a night light similar to your tailight with 36 Luxeon SuperFlux 4-pins LEDs running at 53mA each.
Boy oh boy oh boy those LEDs get hot! Their max allowed current is 70mA.
I mounted them on stripboard inside a compact cassette plastic box with many cooling holes.

I made another nightlight with the LEDs but used only 16 of them. They are housed inside a cringled glass cover.

I added a white LED to each but the heat destroyed the yellow phosphor so I relaced the white LED with a blue LED.

[tourist]

Mine aren't running too hot.  We'll see after the project is completed if I need the led brighter.

Finally got the circuit board  done and tested.  After a couple of days and countless tries finally got the traces to transfer to copper.  Drilled holes, mounted stuff.  Just looking around the house for hook up wires.

[merold]

I need to build a circuit to detect when the brake lights are on and the turn signals on.  The brake light and tail light are the same bulb.  The tail led light uses 2.4 volts then when the brake is applied the voltage increases to 13.8 volts.  This circuit seems to work on my test board.  When the input is 2.4vdc the leds are on at about half the brightness and when the brake would be applied and the input is 13.8 volts the leds are on full.  The input only has two values: 2.4vdc and 13.8vdc.  These components are just what I had lying around and I think I would use a different op amp in the final circuit because I will need to monitor 4 different lights, (2 brake and 2 turn signals)  The turn signal circuit differs by reversing the 10k and the 100k resistors.  The turn signals input is either 0v or 13.8 vdc.
I read out here that the 741 isn't good to use in automotive applications, idk.  I'm open to better components or design but I can't draw too much current from the light circuit because it is monitored in some fashion by the onboard computer.  Any help would be appreciated Thanks.

that's so cool.. sir..
I wonder how can i easily troubleshoot the defected of that??

[audioguru]

I read out here that the 741 isn't good to use in automotive applications, idk.

It is an antique that is 44 years old and only used by old teachers today.

I'm open to better components or design but I can't draw too much current from the light circuit because it is monitored in some fashion by the onboard computer.

The lousy old 741 opamp has inputs that do not work when they are anywhere near 0V or the positive supply voltage. Its output does not go low enough to turn off the transistor. It is too slow to follow the PWM pulsing.
I would use a low power MC33171 opamp or normal power MC34071 opamp. Their inputs work perfectly at 0V and their outputs go down very close to 0V.

Your LEDs are shown backwards.
Since your transistor is an emitter-follower then the 1k resistor at its base should be replaced with a piece of wire.

[Hero999]

Forgive me for butting in, I've not had time to read the entire thread.

I notice there are lots of LEDs connected in parallel; why? It's more efficient to connect them as three strings of five or even five of three.

If all this does is buffer the PWM signal then why use an op-amp IC? A comparator such as the LM311 or LM393 would be a far better option.