BicycleLighting: Difference between revisions

From Noisebridge
Jump to navigation Jump to search
No edit summary
 
(11 intermediate revisions by 2 users not shown)
Line 44: Line 44:
=== parts ===
=== parts ===
* 1x or 2x 555 timer chips
* 1x or 2x 555 timer chips
* 1x LM393 (op amp) analog comparator for voltages
* 1x photoresistor
* 1x photoresistor
* 1x pot
* various resistors & capacitors to produce the right 555 outputs
* various resistors & capacitors to produce the right 555 outputs
* 1x fabric pressure sensor
* 1x fabric pressure sensor
Line 54: Line 56:


The photoresistor registers about 140k ohms in inside natural light, held under the bike seat.  In the bathroom with the door mostly shut, simulating nighttime, it registers 22M.  Tentative initial ON value: 20Mohm.
The photoresistor registers about 140k ohms in inside natural light, held under the bike seat.  In the bathroom with the door mostly shut, simulating nighttime, it registers 22M.  Tentative initial ON value: 20Mohm.
Learned a bit of comparator chip circuit theory from David... Have the op amp circuit breadboarded out such that it turns on one LED when the photoresistor is occluded.  Circuit includes a pot for calibration, forming the second voltage divider.
Notes on my pots:
blue pot ranges from 0 ohms to about 2.35k ohms.  the pair of leads swap resistance levels compared to the single lead, such that the total resistance is always about 2.35k ohms.  When the pot is turned all the way to one side, resistance between the single lead and one of the doubles is 0, and from the single lead to the other of the doubles is 2.35.  When it's turned all the way to the other side, the resistances on the double leads, against the single lead, is swapped.
Have the 2 circuit bits combined with a transistor, fully functional.  Drew circuit diagram & breadboarded additional version from diagram to test diagram; made a couple small corrections.  Diagram is correct (possible resistor values s/b changed for optimality as NB supply hadn't anything smaller than 100 ohm).


=== next step ===
=== next step ===
Figure out how to translate the photoresistor's output into an ON/OFF signalSurely I don't need a microcontroller for this!
Big(?) capacitor behind seat switch s/b sufficient to keep lights on for some time after seat is vacatedHow big must it be?  Test.
 
Alternately, I have a scribbled diagram that uses a Schmitt trigger which is claimed to be usable for this purpose... research & poke at this more.


I can create the seat cover with embedded pressure switch whenever, it's just a switch, so it can wait until last.
I can create the seat cover with embedded pressure switch whenever, it's just a switch, so it can wait until last.

Latest revision as of 22:10, 15 January 2009

I know rachel is attempting her own bicycle lighting project, so I am hoping that by putting notes on the wiki we can share ideas. Rigel

rigel's dynamo lighting system[edit]

This will be based around the Shimano DH-3N71 Dynamo hub. It outputs 3W at 6V. I'm not sure how the voltage or current changes as speed increases, but I am under the impression that it is fairly well-regulated internally.

Currently, I have it wired directly to a rectifier, whose output is a 220uF cap and a Cree XR-E LED. This produces light adequate to ride in the dark and avoid most obstacles. None of this is regulated in any way, so at low (slightly faster than walking) speeds it begins to flicker. This might be a problem for long steep climbs in the dark.

What I would like, ideally, is to have the ability to draw current from a li-ion battery pack when it begins to flicker. In addition, I would like to be able to charge the batteries, and additionally a cellphone if feasible, from the dynohub when i am riding in daylight. I have not found the drag to be noticeable from one LED, but presumably if drawing more current this could become a problem.

Since I am a n00b, I dont know the first thing about implementing this, but i have been led (ha!) to believe that i need the following modules:

  • voltage regulator
  • battery charging circuit
  • led driver circuit
  • 5V usb output for alternative uses

The cree LEDs are brightest, i'm told, when drawing 600-700mA. The voltage drop, i believe, is 3.3v per LED, and they are best wired in series to avoid thermal runaway (not sure why this happens, but it's been talked about on several forums i've seen, so i think it best to avoid it altogether) I would probably start with 2, but may use as many as 4 li-ion cells for this purpose.

Additional tasks[edit]

  • reflector design - will likely try to use the reprap for this. are there software packages for this sort of thing?

questions[edit]

  • What voltage regulator should i use?

- for a circuit that's powered by a battery only, no voltage regulator is required, at least not in general. if one uses a battery that provides too much voltage, then some means of dropping the voltage to the load circuit is needed (this can be diodes in series to the load). in the case of a battery and load being refreshed by a charger (e.g. something connected to pedals), then it's probably a good idea to protect from over-voltage (spikes) and to protect from possible reverse currents. - if power drain is a problem and you've determined you should use a voltage regulator, look at specs for switching regulators rather than linear regulators. --Jstockford 11:17, 26 November 2008 (PST)

  • do i need a multiple output voltage regulator or should i step up the voltage in the battery charger module itself?
  • what battery-charger IC should i use?

- maybe none, possibly just passive components and diodes, maybe an active circuit using transistors or op amps. --Jstockford 11:17, 26 November 2008 (PST)

  • what kinds of interconnects (i would like this to be modular for the working design) are waterproof (and cheap!), as i will be using this in potentially foul weather?
  • where is this circuit physically (might be no where at the moment, might be bits and pieces on a bike or on a shelf...). --Jstockford 11:21, 26 November 2008 (PST)

Rachel's automated lighting[edit]

The point of my system is that I don't have to remember to turn it off (or on). The first rev will be battery powered; a later revision may take advantage of Rigel's work and connect to a dynamo.

Rev one will contain a couple bright white LEDs in front and some red LEDs behind. The front will be steady, the rear blink, as is often seen on bicycles. The base is a seat cover, in which is embedded a pressure switch such that the lights are only on when the seat is occupied (a later rev will change this to leave the lights on for 2 minutes after the seat becomes unoccupied). An additional switch, to prevent unnecessary daylight use, is a photoresistor which will shut off power to the lights if it sees more than X amount of light.

parts[edit]

  • 1x or 2x 555 timer chips
  • 1x LM393 (op amp) analog comparator for voltages
  • 1x photoresistor
  • 1x pot
  • various resistors & capacitors to produce the right 555 outputs
  • 1x fabric pressure sensor
  • fabric for seat cover
  • batteries as needed

current state[edit]

I have the 555 circuit modelled for a nice rate of flash for the rear LEDs, using 3V.

The photoresistor registers about 140k ohms in inside natural light, held under the bike seat. In the bathroom with the door mostly shut, simulating nighttime, it registers 22M. Tentative initial ON value: 20Mohm.

Learned a bit of comparator chip circuit theory from David... Have the op amp circuit breadboarded out such that it turns on one LED when the photoresistor is occluded. Circuit includes a pot for calibration, forming the second voltage divider.

Notes on my pots:

blue pot ranges from 0 ohms to about 2.35k ohms. the pair of leads swap resistance levels compared to the single lead, such that the total resistance is always about 2.35k ohms. When the pot is turned all the way to one side, resistance between the single lead and one of the doubles is 0, and from the single lead to the other of the doubles is 2.35. When it's turned all the way to the other side, the resistances on the double leads, against the single lead, is swapped.

Have the 2 circuit bits combined with a transistor, fully functional. Drew circuit diagram & breadboarded additional version from diagram to test diagram; made a couple small corrections. Diagram is correct (possible resistor values s/b changed for optimality as NB supply hadn't anything smaller than 100 ohm).

next step[edit]

Big(?) capacitor behind seat switch s/b sufficient to keep lights on for some time after seat is vacated. How big must it be? Test.

Alternately, I have a scribbled diagram that uses a Schmitt trigger which is claimed to be usable for this purpose... research & poke at this more.

I can create the seat cover with embedded pressure switch whenever, it's just a switch, so it can wait until last.