Help Me Hack A Bike Light

[SuzyC]

I am souping up my bicycle lights to  maximize using all the juice in a rechargeable Li-Ion cell circuit  to blink a white LED (front light) or two ultra-efficient red LEDs (rear).

I want to make thus my first SMD PCB  project to satisfy this goal.

I want use a always keeps its charge Li-Ion battery to power it, to make a front very bright white LED blinker/ high power flashlight.

I already have a bunch of almost new Li-Ion cylindrical batteries taken out of a laptop battery pack. Even one is quite big, but not too big, to fit into the bike light, but then leaving little room for circuitry..so I go SMD to get all the stuff to fit.

I am thinking of using a rim-mounted pickup coil magnet-generator for an output, salvaged from a wheel blinker to run up the front light to charge the battery while I ride!

I want to be using a touch control to turn it on/off so I can make the light enclosure airtight, so there is no moisture creeping in  corroding everything when it rains a lot..which is does here.

If I use a MCU I can get the battery voltage to blink out serially to show the Li-Ion battery charge state.

I've already got two large blinkers on my bike that are alkaline AA/AAA battery powered, but every winter, the extreme humidity rots the on/off switch and battery connections.  These are are also using "dumb" circuits, just a limiting resistor to the LED's, so things just get dimmer as the battery wears, but quit altogether before the batteries are fully exhausted.

I also don't like the one's that mount way down on the wheel, they use magnets to generate power, but it makes changing wheels to fix flats and repairs more complicated and time-consuming and i don't get to have a bright front headlight/flashlight

What's so cool about the blinker IC's is that they can work down to .8V and so fully exhaust a pair of batteries and they can be turned on/off with a single spst momentary switch.  I can also use a boost converter to get my wheel magnet-generator output to give a steady 5V to a PICMCU, but I don't quite know how to get a touch switch to work through the plastic lens of the bikelight enclosure.

But I googled and couldn't find an IC part number for a blink'n chip!

I have taken apart a 16-wat LED replacement for incandescent light bulb and I have salvaged over 27 very high efficiency white SMD LED's, eager and ready to make my front light a powerful flashlight as well as a flasher.

I know I can buy some kinds of  non-chargeable blinky bike things pre-fabbed at the bike ship, but they will not work as a bright front light, so what's the fun in that?

Any of you geniuses out there in Nerdland know how to do get this idea to work?

[Kalvin]

Nice project. However, you don't want to discharge your Li-ion battery below its designed cut-off voltage: "Avoid very deep discharges (below 2 V or 2.5 V). Very
deep discharges will quickly, permanently damage a Li-ion battery." from the article "Proper Care Extends Li-Ion Battery Life"
http://powerelectronics.com/site-files/powerelectronics.com/files/archive/powerelectronics.com/portable_power_management/battery_charger_ics/804PET22li-ion-battery-life.pdf

[SuzyC]

Thanks Calvin, I must then also figure out how to get my circuit to shut off automatically when the battery gets too discharged or else maybe switch to rechargeable AA batteries, but they don't have soldering tabs and I don't want to use or have room for a battery holder.

If I have to use NiMH batteries, then my next project must be to make a spot-welder, that's going a bit off track!

 I have several 2450mAH AA cells, but there is room for just two inside the bikelight enclosure.  I think the rechargeable NIMH can take total discharge without damage, but the lower voltage won't drive the white LEDs, I have 15 of them on using just one series resistor, and they all glow equally bright, but also found that they run full  bright with at least 3.6V.

I just made a quick SMD PCB to try hooking up in parallel 15 white SMD LEDs salvaged from a lightbulb. The LED's are really bright with just 40mA of current shared by all 15. They are so bright they hurt the eyes looking straight them!

[amyk]

You can get 18650s with protection circuits built-in or design your own (usually based around one of the popular ICs like the S8205A or DW01), but it is very important that lion cells do not get discharged completely as that damages them.

Do be careful with those types of batteries, the amount of energy contained in them can do some serious damage - just look up "li-ion fire" on YouTube for some examples.

[SuzyC]

Thanks Amyk, but I think there is a little more hysteria then necessary about the danger of  Li-Ion batteries.

I have several times shorted them out by accident wiring errors but these accidents just burnt out the Li-Ion batteries and melted the insulation off  even 18 gauge wires attache to them. The batteries became quite warm to the touch and there was some steam vented by the Li-ion cells. I understand that hydrogen is released, so there is a danger of fire  in unventilated enclosures.

After being abused by a sudden short circuit their charge acceptance dropped precipitously, and internal resistance increased so much that  they were unusable to deliver any large current.

So, now I am a little more careful but not at all afraid of using them in my circuits. No Shorts Allowed!

[TopLoser]

I think there is a little more hysteria then necessary about the danger of  Li-Ion batteries.

http://www.candlepowerforums.com/vb/showthread.php?141137-Inhaled-vapors-from-battery!!!

[SuzyC]

The CR-123 primary battery is not a rechargeable chemistry Li-Ion battery like used in laptops.

[TopLoser]

The CR-123 primary battery is not a rechargeable chemistry Li-Ion battery like used in laptops.

Good point well made, you are of course right.

But after I read that thread I cleared a load of those CR-123's out that had been rattling about loose in my office. And I've seen people trying to 'recharge' 2 CR-123's in series in a 18650 charger.... I'm always a bit concerned when people start fiddling about with any kind of lithium batteries now.

[Yago]

Aren't the 18650s used in laptop batteries of the more volatile ICR, as opposed to IMR?

Using a knackered unknown battery is not a great start, it may have already have been abused and over discharged.

Buy a good quality cell or battery, and the old adage of "use protection"!


How about a system that requires no batteries, a cap running the lights, scavenging energy from bike/rider?

[SuzyC]

The Li-Ion batteries came from an almost new battery pack on a laptop, I know they are good.

Using a Li-Ion battery in a circuit doesn't need to be more hazardous  than using a cellphone or a laptop PC.

If some types of Li-Ion  batteries overheat and leak electrolyte, don't clean up the electrolyte released with water, some Li-Ion electrolytes will release hydrogen fluoride gas when coming in contact with water.

I understand that if I am careful with the battery, especially with soldering to the terminal tabs, careful handling, insulating  the battery contacts/connections  to protect it from shorts, watch out for over charging,

If I am careful, I don't have to have any more worry than using my laptop. The very low power generated by the bike wheel magnet-coil generator is easy enough to keep within safe voltage limits.

If I use a thin piece of hookup wire (#30) to connect the battery to my PCB, the wire itself will serve as a protective fuse in the case of a short.
 -------------------------
Thank you all for allowing me to remind myself of the dangers of this hobby.

There are many hazards with electronics and I am aware of them:

Soldering flux fumes, even in tiny amounts, especially with skin contact or inhalation, some water-soluble soldering flux fluids or its fumes,fluxes can cause violent allergic, or respiratory reactions, and  even a short exposure can cause lasting damage the eyes, skin, lungs,  cause cataracts, blurring of vision, from temporary blurring of vision to prolonged exposure causing changes to the eye, permanent astigmatism, non-water cleanup fluxes always can cause eye and lung irritation, prolonged exposure can cause asthma, etc.

Lead ingestion from soldering fumes damages the nervous system and other parts of the body, but soldering flux can be even more dangerous.

Contact with soldering irons can cause serious burns.

Fumes from charging or shorted lead acid batteries can create hydrogen gas and risk of explosions.

Fumes form burning plastic wire insulation can be very toxic and release fluorine(Teflon) or chlorine gas(vinyl insulation).

Cleaning fluids like MEK or TCE etc. can over time turn your liver into a cancerous sponge.

Contact cleaning foaming agents can cause lung irritation, asthma.

Glues and spray paints contain dangerous solvents, i.e. toluene, that can be inhaled and sicken and intoxicate, cause  hallucinations.

Epoxies can cause a life-long skin or eye sensitization leading to life-long episodes of painful inflammation over the years to come.

Superglue can glue your fingers together.

Burning selenium rectifiers can give you an overdose of selenium fumes.

Electrolytic capacitors when subjected to over-voltage or excess AC current,  especially tantalum or liquid electrolyte aluminum types,  or when somehow reversed in polarity in a powered circuit can explode like a firecracker or take off like a rocket and the fumes are no doubt also toxic. Fast traveling exploded electrolytic cap cases can hit you in the eye.

Burnt PCB fiberglass or pthalates as in perfboard, or burning resistor fumes, over-heated transformer impregnation fumes, smoked IC's or semiconductor cases..these also are no doubt toxic.

Going through a collection of screws and bolts to find the one you need: cadmium and other toxic heavy metals rub off onto the skin.

Grinding things, sanding, PCB drilling, sawing, working with materials can cause flying metal pieces,  can release inhalable micro-particulates that can enter and stay in your lungs, especially take care when working with fiberglass,  tungsten and beryllium or many other exotic metal containing parts.

Some batteries or lamps or fast switching relays contain mercury or cadmium.

Vacuum tubes contain extremely dangerous thorium in their heaters.

Neon bulbs use radioactive elements.

High voltages can generate X-rays.

Very high magnetic fields can cause changes in the brain.

Over-heated components on a circuit board, hot parts can cause burns.

Very strong magnets can cause pinching injuries.

Rare earth magnets can poison you when handled or drilled or grinded.

Hand tools can cause skin cuts and other bodily injuries.

Trying to place tiny SMD components on a PCB and getting them to stay in place before soldering can cause a nervous breakdown.

Electricity can cause dangerous shock, broken bones or other injury from involuntary muscle contractions, and even possibly death.

Even sitting down for extended periods without getting up for a break can acutely damage the legs, its nervous and vascular system, causes a deterioration, and even if a long sitting session is followed by a long workout at the gym, the damage remains.

Sipping aerosol whip cream while working at the test bench can expose one to nitrous oxide and bouts of hilarity.

Did I forget anything?

But if one exercises reasonable caution, isn't electronics fun?

[Yago]

I know you are making a point there, just not quite getting it!
*Runs away before gets shouted at again...*

[SuzyC]

The point is that it is prudent to always be careful, but it is worthwhile to take reasonable risks.

[Zero999]

Soldering flux fumes, even in tiny amounts, especially with skin contact or inhalation, some water-soluble soldering flux fluids or its fumes,fluxes can cause violent allergic, or respiratory reactions, and  even a short exposure can cause lasting damage the eyes, skin, lungs,  cause cataracts, blurring of vision, from temporary blurring of vision to prolonged exposure causing changes to the eye, permanent astigmatism, non-water cleanup fluxes always can cause eye and lung irritation, prolonged exposure can cause asthma, etc.

Few people have such a severe reaction to solder flux fumes. True you should avoid inhaling them and use a fan to blow them away if you must but it's not that dangerous.

Lead ingestion from soldering fumes damages the nervous system and other parts of the body, but soldering flux can be even more dangerous.

Impossible. You can't ingest fumes. You must mean inhale fumes but solder fumes do not contain lead because the temperate is too low to cause a significant amount of lead to evaporate. Ingesting small pieces of solder or chewing it will give you lead poisoning but handling it is unlikely to cause significant exposure to lead.

Generally lead free solder is often more dangerous than leaded solder because it requires higher temperatures and the flux is more aggressive. Leaded solder is safer, as long as it's handled correctly and disposed of in an environmentally friendly manner.

[SuzyC]

One less thing to worry about!

I know about lead, but what is this leading towards making this project?

Aren't we getting off topic???

[wraper]

The Li-Ion batteries came from an almost new battery pack on a laptop, I know they are good.
Using a Li-Ion battery in a circuit doesn't need to be more hazardous  than using a cellphone or a laptop PC.

There is protection circuit in all laptop and cellphone battery packs. Once you remove battery from the pack, there is no protection anymore. Didn't you notice multiple temperature sensors inside the pack where you removed batteries from? Also PCB with bunch of mosfets, ICs and current shunt resistors. Do not forget huge recalls of the laptop batteries because of the safety issues.

[SuzyC]

Wrapper,

The protection of a battery back has a circuit that protects the battery from overcharging, overheating, charging while in over or under heated environments, and a simple non-healing fuse prevents short-circuiting hazards.

This is not the case  here in my project, the tiny charging current from the wheel generator cannot ever create any heating, there is no chance of excess charging current, unlike charging the battery in the laptop using a 3.5+ Amp capable 18-21V charger bricks commonly used with a laptop.

It is easy to prevent overcharging, a LM431 shunt regulator or just adding a zener would easily accomplish that, but even a very slow exposure to over-voltage(very unlikely with the feeble generator output and use of the light at night) would not be a hazard. The very unlikely  overvoltage at a tiny charge current would just plate the electrodes inside the cell, decreasing charge acceptance and this slowly ruins the capacity of the battery, but it is without any danger of explosion.

A fuse(or even a thin piece of wire would work) would protect the Li-Ion cell against a short-circuit condition hazard.

[mark03]

Even if you are satisfied that the fire, etc. risks are low, you still have to worry about damaging your Li-ion pack and thereby having it stop working   That can happen pretty fast with overcharging and overdischarging.  At some point it's easier to turn to off-the-shelf battery-management and battery-protection ICs for this task.  Also beware that overheating is not just a too-much-heat-in-a-small-space risk.  If a cell catches fire, it will burn, vigorously.  Do you ever park your bicycle indoors?

The wheel generator idea is cool, but I doubt you will be able to get enough juice from a homebrewed solution with a magnet spinning past a coil.  I hear those generator hubs are really nice, albeit expensive.  There was also a modern version of the "bottle generators" which used to be standard on bikes.  It was called Lightspin and I understand it performed quite well, but now they are difficult to find, the generator hubs having taken over the market.

Anyway, what makes the generator side of this project interesting IMO is the opportunity to do MPPT (maximum power-point tracking) for optimal battery charging from the generator.

You might enjoy looking at my bike-light project for some ideas.  No generator though.
http://www.keteu.org/posts/bikelight.html

Mark

[SuzyC]

Mark03 many thanks, I will take a look at your link!

You seem to be the only one so giving me some advice directly relating to my bike circuit design.

I am not using a Li-Ion pack, just one 16850 Li-Ion cell of the 6 cells salvaged from the laptop battery pack. The battery pack was rated at 4400mAH, so the individual cells are therefore 2200mAH each.

I always park my bike outdoors.
 
The wheel generator on my bike, is just a hacked version of the those sold as front and rear no-battery-bikelight-blinkers, these are just wheel axle mounted LED light enclosures that are that have inside a pickup coil to generate electricity from the pair of flat magnets mounted and rotating on the bike wheel spokes.

A modified wheel generator has already shown to be able to weakly charge the Li-Ion cell and power the front blinker/headlight, which itself is my modified battery powered version of a front basket mounted bike blinker.

Due to weak charging ability, my idea is to use the bright headlight function only when needed for a short time at night, relying on the charge stored in the Li-Ion battery, while always having a white LED array blinking for safety at night.

I took apart the wheel LED light blinker pickups enclosures and disconnected the LEDs  and put instead inside the case a Schottky diode bridge and a 1200uF/6.3V cap. I dressed a twin pair cable along the frame to deliver the generator output in DC form to to the front basket-mounted blinker light enclosure where a PICC chip manages things.

During the day, as I take a ride, the LED's will be off, charging is continuous, and a PICC mcu will periodically awakens from sleep to monitor the charging.

So, during the day, the Li-Ion batter is charged, practically without any drain while I go for a ride.

At this time, while I working on a new design using two MCP1624 boost regulator ICs. The first 1624 charges the battery.  A second MCP1624 converter connected directly to the Li-Ion battery will be used to generate +5V to power the PICC mcu chip.

I have been researching LI-Ion cell charging on the web and also touch switch circuits. I have now decided to use a MCP1624 boost converter to continually charge the sing Li-Ion cell and its output is  set at 4.18V.
A 5V 1-W zener diode is to be connected across the DC input to the MCP1624 connected directly to the generator output  to make sure the generator voltage does not rise high enough to damage the MCP1624.

I have created a modified version of a touch control technique (Microchip AppNote 1298) that does not require built-in touch control hardware in the MCU but uses A2D. With touch control I now have a water-proof sealed front light/blinker without any mechanical switches allowing moisture to enter the bikelight enclosure.

At present,  I am using the mechanical original on/off switch on the front basket blinker light for on/off. The final version is still being worked  on my project bench and a PCB needs to be scaled down iwith SMD parts to fit inside the front light enclosure along with the battery.

I figured out how to use varying  generator output to feed a CMOS 555-based boost converter of my own design  to develop 6.8V from as little as .8volt output from the generator. It starts up around 1.8V Vin.  The 6.8V boosted output feeds a 78L05  to power the PICC mcu that controls charging and blinking functions. The PICC chip monitors the battery voltage and serially blinks out the battery voltage by selecting the "show battery voltage" function. Two momentary contact surfaces serve as  touch switches and are made from the coiled part of small bright brass safety pins. Touching one sw surface selects various blinking modes, including "always on" and "show battery voltage" and "sleep/off", etc. Touching both contacts awakens/turns on the circuit if it is asleep/off.

At present the generator output is so weak that the Li-Ion cell eventually needs to be removed and charged after about a month during the winter.  I must be careful to not use the flashlight function too long or too often. However, the generator output will keep the battery charged if I just fast-blink the 15-LED array with a 2mSec on time 1-Sec off duty cycle, and this is sufficient for a front bike blinker light. Because the front light has so many very efficient bright white LEDs, even this short fast blink is seen as very bright to anyone in front of the bike. Doubling the number of magnets on the spokes would double the charging current and allow the battery to be charged by the generator alone.

I took my DVM with me on the bike and I have found that the generator's output, even at the highest pedaling speed,  is  <100mA Max, and this is at short-circuit, and in any case, this current is pulsating and varying because of the spacing of the spoke magnets. I measured something <30mA average output while reaching about 3V with a slow pedaling speed. At these tiny currents, there is no danger of heating up the battery.

I am still working on the code to wake up the PICC mcu from SLEEP, (during the day) every several minutes to make an A2D measurement of the battery voltage, so if charging causes the Li-Ion cell voltage to exceed about 4.18V, the PICC chip will switch on the LED array full power until  the voltage drops again to about 4.1V, thus preventing LI-Ion over-voltage damage during charging.

In any case, I have searched the web and found the answers to the questions I've posted by my own work on this project and by getting the help I needed  from viewing/posting on other electronic/hacking forum websites.