brake/tail lights

[charlie_r]

I am looking for a way to avoid having two microswitches on each of my bike brake levers.

What I have now is two sets of 3 leds in series as my tail lights. I am running them with separate LM317 voltage regulators in constant current mode, set for 100mA.

I would like to combine the brake lights with them, and run it all with only one LM317, with one microswitch on each brake lever.

Power is coming from a 12V 8Ah SLA.

3V0 gave me **broken link removed**.

Now I need a bit of help adding another level to it. Page 2 of the attached .pdf from ON semi shows a way to drive multiple series strings from one V_reg.

What I need is a way to combine them, so I will only need 1 switch on each handle.

View attachment 61451

 

[kinarfi]

I had a similar but different problem, several higher voltage LEDs on the varying power of my offroad vehicle, so I upped the voltage and put every thing in series and then limited the current. Looks like you just have too many LEDs to run them all in series at 14.2 Volts. You could up the voltage and then series them all and limit the current once and turn it on with one switch.
make another post if interested in how I did it.
Kinarfi

 

[charlie_r]

Thanks for the reply.

Upping the voltage is currently not an option, unless someone can point me to a boost converter. Not knowing how much amperage those will handle, and with the requirement of needing 340mA on the bright setting.

As stated, I have as a power source of one 12v 8Ah sealed lead acid battery. I can't afford going to LiPo or other chemistries, so I'm probably stuck with the series/parallel topology.

 

[kinarfi]

Reread your post, several times, do I understand you correctly?
You have one string of LEDs that you run at low current for a tail light and you up the current through those same LEDs for the brake lite.
This is on a bicycle, so you do not have a charging system.
I assume these are Red 1 Watt 3.3V 340ma LEDs
If So, what 3VO gave you is perfect.
I drew it up on spice as such, and this is what I came up with,
It looks like the over head of the lm317 may be a limiting factor, my spice simulations show you need a minimum of 13.2 volts to get the desired currents, assuming Spice got your LEDs right.
Kinarfi

 

[kinarfi]

Here's some other ideas for you to consider, boost circuits.
You can also go to **broken link removed** and put in the parameter of vin min =9
vin max =13 v out =14 and I out =.3
The LT1270 is about $10 and the inductor is around 2 and if you pin to pin assemble this and cover it with silicone, it's very small
Kinarfi

 

[charlie_r]

Spice has it almost right.

I have the spec sheet from the supplier:

View attachment 61486

I think it may be "surface charge" on the battery, but after 10 hours on the charger, I measure 14.2V on the battery, which quickly drops to 12.8 under a load of 780mA, then stays there for quite a while (haven't measured the time).

Odd thing about these LEDs, I have pulsed them with these specs:


On time (t1) is 0.00229 seconds.
Off time (t2) is 1.04 seconds.
Duty cycle is 0.22%.
Frequency is 0.96 Hz.

With no current limiter, just a high current transistor on the output of the 555. Ran the test for 48 hours continuous. No failures. Which calls into question the specs from the supplier.

Without a scope, I can only guess that the current has been exceeding 1A for that very short pulse time.

What I have installed on the trike:

View attachment 61487

View attachment 61488

At 100mA, the red LEDs without lenses can be seen in bright sunlight for several blocks (about 700 ft).

While testing the tail/brake circuit, there is a very noticeable difference between the 100mA and 340mA.

The reason I want to use only 1 switch for each brake lever is to reduce the fiddly adjustments for the switches.

 

[kinarfi]

Charlie, got some more suggestions and simplifications, the LM317 is only needed if you have varying, increasing, voltage or if you want to limit the current when voltage is high so current will remain constant as voltage drops, in my opinion, for your application, it is not needed, as the battery discharges, you probably won't be able to tell over the period of 12 hours. (You may want to wait a while after charging to allow the surface charge to dissipate before light up the brakes) I don't know if you pulse the light while in use or if that was just testing, but look at this wiring diagram and see what you think, notice that the switch goes to ground, so you only need one wire to each brake lever. You could even make it real simple by having the lever contact a piece of springy metal to complete the circuit, or some such set up.
Kinarfi
PS I put some of these lights on my vehicle and if I remember correctly, it only pulled a 1/4 amp, +or-
https://www.ebay.com/sch/i.html?_nk...ue&_nkwusc=pirahna+tail+lights&_sop=15&_rdc=1
Item number: 330692686828

 

[charlie_r]

My reasoning for the LM317 is constant current not dependent on voltage from the supply.

Yes, I thought about the resistor method. Problem with that is it can never give a constant current, it is entirely dependent on supplied voltage. As the battery voltage goes down, so does the current. I would also risk overcurrent from a fresh charge. Been there done that, and replaced many LEDs as a result.

Eventually, I will be removing the battery from the case (you can see it in the cargo box), and permanently attach it behind the front light mount. The charge controller, currently in the case, will be in a project box. So when I am done for the day, I just plug in the battery for an overnight charge. With the resistor method it would be too easy to overcurrent the LEDs if I accidentally turn them on while charging.

Grounding through the frame is also not reliable, due to sending current through the steering system. Tried that before. Resistance through bearings is not constant, and would have to be accounted for in the circuit.

This is being used as a test bed for a (for me) major project myself and a few others are working on.

I have volunteered to build the light system for this -- **broken link removed**

It all will have to work perfectly, and be extremely reliable. The person we are doing this for will not have access to get it fixed should something break.

Since I have limited funds, it has to be done a cheaply as possible, without anything more complex than a few cheap ICs.

Seriously, I do appreciate your efforts in helping me figure this out.

The other option I have is extra LEDs for dedicated brake lights, with the added parts count. Which is probably the way I'll have to go.