MFP. Design Project - need advice for variable intensity, dual color lighting source!

[tedward]

Hi Everyone,

This is my first post on the forum and it come due to my need for some advice!

I am a university student, and am working on a project that reduces sleep inertia (groggyness) when you wake up from a nap.

One concept that I am working on will look essentially a modified set of sunglasses, with a light emission source coming from the top of the glasses, on the face side.

The light ideally would have two colors that would be shone onto the users eyelides; a blue, and a yellow/orange/amber that would emulate a sunrise.

The light source must:

Allow a gradual change of intensity
Not use too much power (the unit will be battery powered)
Be as small as possible.

I feel like LED's might be the right route to go down, but after that I don't know. Maybe you guys could suggest something?
I have looked at this, Link removed. See red banner at top of page is this any good? I don't see any mention of it allowing a change in current?

Thanks in advance!

 

[AnAppleSnail]

DFM involves some things that I'll recap for CPFers.

To be successful in Design for Manufacturing, consider:
1) Product scope, intent and complexity
2) Time to market
3) Cost
4) Product Competitive Environment
5) Organization infrastructure
6) Design, engineering and manufacturing tools
7) Staff experience

[List from Engineer's Edge

You've mentioned the scope and your hopes for cost. What is your experience here? If someone suggests a way to use a microcontroller running ARM instructions, wired as such, with this or that power infrastructure (LiPo with a micro-USB port to power and set timing), could you?

What similar products exist? One I know of is the sunrise clock, which simulates a sunrise, which can be sped up to different degrees, before the alarm time.

What are your production means? My biggest limit on homebrewed projects is that I can't shape metal and plastic beyond scraping, cutting, and grinding with hand tools. At a university, you hopefully have access to neat tools. What are they, and who will use them? Do you communicate requests to them with autocad files? Part drawings?

Consider users. What about people with large or small faces? How dark are the lenses of the frames?

Finally, what exactly is your scope? There are a few ways I can think of off-hand (And without caffeine!) to program such a device.

1-button interface: Select 5, 10, 20, 30, hour, and 2 hour cycles. Each press 'beeps' and steps forward. A 3-second pause in beeping counts out the time, (5 beeps for 1 hour, 3 for 20 minutes), and at whatever time is appropriate the lights do their thing to simulate sunrise, or even accelerated sunrise. I suggest a red/green LED indicator for battery life.

LCD interface: A 4-button keypad selects times to nap and wakeup speeds. It should also probably show battery life.

Bluetooth interface. Integrated with a smartphone/computer app (Blackberry/Android/iOS), this lets you leverage the device your target market has to set the device. The base app sets time to nap, sunrise speed, etc. Bluetooth sends the 'wake up' signal to the glasses.

To actually generate such a sunrise is easy. I suggest that you use some form of light tube in your lighting. A few SMD LEDs shining into an acrylic frame (with clear parts towards the user) would make an effective glow on the eyelids without being blindingly bright. Controlling the LEDs can be done with PWM (This can cause annoying strobe headaches) or 'current-controlled' driving, which would be preferable for this use. If you involve any widget that can control some output, you can dim or brighten LEDs on a timer. Tell us more about this, and we'll talk more circuits.

Edit: All LEDs can be dimmed, but not all drivers are friendly to dimming. You could have a dead-simple stepper motor slowly rotate a potentiometer to dim things instead of a circuit, but mechanical linkages are large and buggy. Most dimmer circuits will send PWM signals of varying intensity to control brightness. In theory, the dozen-kHz range is not seeable, but I would suggest GOOD research to back that up. Strobes in the eyes is painful, even on the edge of perception (60Hz for most people, except with moving objects).

 

[tedward]

Thank you for getting back to me so quickly, and also thanks for outlining the process, it might make things easier to understand for the other members who have better things to do than pour over engineering diagrams like me!

There are a number of similar products, that use a similar principle, the differing element being that it would be the only light based system designed around a busy commuter with limited access to sustained sleep. Philips for example do a alarm clock which emulates a sunrise and has an audio alarm as a backup.

I know extensively about materials and production processes, however I do not know much about electronics past a Physics A-level, and less about practical circuitry - that would be where your expertise would be so very useful!

I will be using rapid prototyping machines - an FDM 3D Printer, and I will be printing parts made using SolidWorks CAD software.

I would like the product to have a rechargeable battery, using a micro or mini usb to usb for the juice!

I have everything outlined that you mentioned, anthropometrics etc except for the actual method for delivering the light - which is again where your help would be so useful!

This is the area I don't know much about:

Do you suggest having on circuit that controls the light emission for the eyes, and another for the battery status LED?
What are the best places (ideally in the UK) to source these sort of components (small rechargeable batteries, these types of LED's) etc.
What type of LED's would you suggest?
What is a 'driver' - is it almost the current control unit for the LED?
I expect it would be possible to get around the strobe effect with a method of dampening the light - perhaps through a semi opaque diffuser?

Thanks in advance!

Ted

 

[Deeksie]

Hi Tedward,

-Standard 5mm leds would do, and can be easily powered, just make sure to put an appropriate resistor before them to limit how much current they draw.
-A micro controller is basically a very small, low power, low cost, computer complete with memory etc. Such as an ATTiny85.
-Arduino is an open source prototyping using micro-controllers with loads of on-line support.

I suggest you order an Arduino of some form, the Duevilanove one is probably a good choice. Learn how to use it and write basic programs (it really isn't hard).
-Write a program that controls the PWM output of the Arduino to create the sunrise effect.
-Find out how to burn the Arduino bootloader onto an ATTiny85 so it uses it's internal clock source.
-Use the boot loaded ATTiny85 to run the code that makes the sunrise effect, perhaps add more code to allow the user to vary the wake up time etc.
-Solder it all together on some pad per hole board, and spec the led resistor so that the led does not draw more than 20mA from the ATTinys output pin, or use a simple transistor to allow greater current.

I think you could probably do what you've stated with some analog electronics...but I think learning the Arduino stuff would be something you could use again and again in other projects, and is really not to difficult. I've certainly not given you lots of information, but I think it might be a good starting point for you.

There are plenty of parts distributors, Farnell, RS, and if you need little bits quick, bitsbox. I'm sure there are plenty more. eBay is often worth a look to.

I'm currently in my 2nd year doing an MEng degree, and specialising in electronics next year. I've found the Arduino stuff extremely useful in helping me get started in electronics. You'd be amazed at what is possible with a £2 micro!

I think battery charging depends entirely on what battery you choose.

I've done a fair bit with PWM controlled leds recently, I don't find the PWM to be a problem unless it's very slow. I'd say your safe over 500Hz, especially considering the application. I think it's only when you are viewing objects lit up with PWM light sources that it can become a issue.

I hope that's at least of some help...

Kind regards,
Deeksie

 

[CarpentryHero]

I was going to mention the RGB MCE emitter but that's overkill, 5mm led make more sense that close to your face. ;)

 

[MikeAusC]

The whole system can be controlled by an 8 pin chip that can be programmed in BASIC but has the power to drive the LEDs using PWM. The 08M2 SMD version is about 3mm wide.
http://www.picaxe.com/What-Is-PICAXE

You can get tiny LiIon batteries that would power the LEDs without needing a driver.

I would use SMD diodes as they're a lot shallower than 3mm or 5mm.

element14 has all the parts you'd need.

 

[DM51]

Welcome to CPF, tedward. :)

This is an interesting project, although I'm not sure the "Recommend me a Light for..." forum is the best place for your thread. There are various places it might fit well, but for the time being, as you are initially trying to decide which LED would suit best, it probably belongs in the the LED (emitters) section, so I'm moving it there now.

If it needs moving on again from there at a later stage, just post here with a suggestion and it can be done.

 

[SemiMan]

Just a guess, but expect you may achieve the best and easiest results using a blue LED and a warm white LED, or possibly neutral, but you can cool up the LED with the blue. Watch the blue wavelength as there are better colors for sleep cycles. Ditto to the post on SMD LEds for packaging, but efficiency is likely to be everything, as is ease of use. If I was designing from scratch, I would consider perhaps a single AAA with a boost circuit to run everything. Light requires power and how often do you want to be changing the battery. Alternate would be the 2032.

Newark (i.e. Element14, their branding program for engineers) or Digikey or Mouser all are suitable. You will tend to find Digikey is better at stocking things. Can you hook up with an elec eng student to do the heavy lifting on the electronics? It may be easier as I expect you are going to have your hand full with light control such that it does what you want without it being intrusive.

Semiman

 

[tedward]

Thanks you all so much for this.

I think I may steer away from Arduino, as potentially programming a chip in BASIC may well be easier in the short term as I have someone who can help with that.

The way I see it at the moment:

I will have 4 LED's per lens (8 overall) which will consist of:

2 x "sunrise" (can anyone help spec an LED for me on this one? A LED that emulates the sun would be ideal!)
1 x Blue
1 x Warm White (Help moderate the blue)

What type of Millicandela rating would suit?

 

[AnAppleSnail]

What type of Millicandela rating would suit?

I'm glad you asked! The millicandela rating is kind of a crap way to rate lights. Honestly, most small LEDs (Even sub-rice-grain-size ones) "can" produce more light than you'll want. My experience with subtle lighting is that adjusting to present lighting conditions is the way to go. I've made line-following robots with BASIC chips that self-calibrate to present lighting conditions at turn-on. I think you'll want to research using PWM with the BASIC chip (And a 555 timer?) to make dimming happen. Use FAST PWM or you'll annoy people.

 

[tedward]

Any of you guys happen to be in Bristol, UK? Would love to sit down, buy you a coffee/beer etc and talk through some of this!

 

[MikeAusC]

I'm in the Antipodes... but if you asked on the Picaxe Forum I'm sure there's a good chance that you'll get some good help.
Picaxe is a UK product and a lot of members are in the UK.

 

[carl20]

How much do you think second hand FDM 3D Printer would cost? would it be advisable to buy it? no idea on its maintenance cost either..
http://www.inkjetsuperstore.com