seriph LED

I think the EE will just have to trust me on the lamp’s design.

Let’s hope to build something this weekend.

Burn Baby, Burn!

Last night was the first time the Atlas LED was powered for an extended period of time.  The daylight white LED was powered for 65 minutes without changing colors or burning out. The same heatsink was used; however, the warm white LED was removed. The heatsink slowly increases in temperature, but remains touchable for the entire 65 minutes. The heat seems to cap at around 30 minutes. After 30 minutes, you can touch the heatsink for about 3 seconds before it starts to be too hot for comfort. A small low-speed fan would probably enable us to use that one heatsink for both LEDs.

So the heatsink we have now for testing could be said to be sufficient for one LED (though I suspect the warm white produces slightly less heat…if not, the same) if the heatsink is not enclosed, i.e., if it’s air-cooled.

I thought the light output from one daylight white LED was sufficient to be used as a task lamp at 12 inches high with ambient lighting.

O, and I was afraid the ceramic-encased resistor was going to fry so I used a medicine dropper to apply some cold water every now and then. You can see if fizz on the resistor, but not on the heatsink.

I hope to start building some actual prototypes from wood and/or cardboard to test out the joints mainly.

Sourcing Components: Delays!

I received the package of samples from ON Semiconductor. FedEx Overnight! Much to my dismay they shipped everything BUT the LED drivers. Oh no! Those were the only reason why I sourced samples from ON. Whatever they are creating, I hope it’s quick.

Then I found out from DigiKey that the microcontroller is backordered until February 15th. So much for spending this week getting familiar with it.

Sourcing Power Components

The LED lamp’s electronics can be divided into three sections:

• Control System
  • To accept user input, manipulate input data, send commands to LED drivers.
• LED Driver System
  • To drive the LED at a constant current.
• Power System
  • To provide adequate and stable power to the Control and LED Driver Systems.

ON Semiconductor was chosen to provide most of power circuitry, mainly because they were the only ones to offer a 1.5A LED driver in PDIP form. PDIP can be easily hand soldered, as they are through hole components.

PDIP

SOIC are much harder and TSOP nearly impossible to hand solder. The latter two are known as surface mount technology devices. These types of components are used in large volume production runs in consumer electronics.

SOIC

TSOP

The following parts have been sourced from ON Semiconductor through their samples program. Unfortunately their sample program requires a $11 shipping and handling charge. While it isn’t much, one better load up on samples to make the charge worthwhile. There is a maximum of 25 samples per component. It seems that DigiKey is the fulfillment operator, and the samples were shipped FedEx Overnight.

LED driver 1.5A NCP3065PG
Voltage regulator 5v 1A MC7805ABTG
Voltage regulator 3.3v 800mA MC33269T-3.3G
Voltage regulator 1.2-37v 1.5A LM317BTG
Bipolar junction transistor 2N3904G

Color Temperature

So…I mechanically attached the Lamina Atlases to the heatsink and made a makeshift “lamp”. I could see the difference in colors between the daylight (cool) and warm LEDs as well as both of them together. Then I tested them against a 15w halogen – the camera does exaggerate the halogen test shot though.

Also, the heatsink gets hella hot and is definitely insufficient.

Cool, Warm, Both, Both Wide, and Halogen:

I definitely like the Warm White better and wouldn’t mind getting 3 (or 1 cool – 2 warm), but heat will probably be a big issue…

—

Update: More Pictures!

SMD Fiasco/SideProj

Ok, so the SMD LEDs I got were totally useless. So I sorta converted it to a light table – but even 6 sets of 3 SMD LEDs are not enough (as you can see). Might make for a nice ambient light esp if I can control each cluster to pulsate for some cool patterns…like a light show!

So we’re definitely going with the Lumina Atlas, but we/I need to figure out how many to install per lamp (2 or 3).

Phil’s Basement Pt. 1

We powered the Lamina Atlas light engines for the first time on the 19th. We had one warm white and one daylight white light engine.

We concluded that one light engine is NOT bright enough, while two SHOULD be sufficient. There wasn’t a real discernible difference in the color of the projected light, but you can see the difference on the light engine itself – especially when it was powered on 5v. The daylight white is also much brighter than the warm white. The light engines produced an unforeseen amount of heat that melted the electrical tape in one test, which also caused the LED to turn blue!!! (because the LED was inverted and lost contact with the heatsink)

Criteria

Function Criteria:

1.	Height Adjustable (24″ minimum)
	Ideally 10″ to 24″ of clearance for detail and general work (and to clear monitor height)
2.	Swivel
	Lamp head needs to adjust to spatial changes in environment, e.g., placing foreign objects on table that may interfere with light
3.	Minimal Footprint
	Limited working space
4.	Effortless On/Off Control
	Lamp will be toggled on and off frequently, so it needs to be easy to find and activate & shouldn’t be placed on the base as the base can become covered with paper
5.	Dimming (Hi/Lo)
	Controllable light intensity for detail, general, and ambient purposes

Engineering Criteria:

1.	Buildable
	Task lamps needs to be buildable with available materials, tools, and methods
2.	Light Output
	Lux (SI unit of Illuminance) needs to be suitable for a task lamp
3.	Heat Dissipation
	High-output LED engines require additional heat dissipation *Important from tests
4.	Light Spread
	Light engines emit light from a small area and may need a diffuser
5.	Lens Protection
	Prohibit direct contact with the lenses of the light engines

Materials & Manufacturing

An initial list of major materials and tools we will be using.
Materials:

1.  Acrylic sheet (aka acrylic glass/pmma)
2. PET plastic sheets (?)
3. Diffusers/Polarizers/Lenses
4. Heatsink
5. Primer/Spraypaint/Varnish
6. Electrical Materials (circuits) to be populated by Mr. Ray

Tools:

1. Band Saw
2. Heat Gun
3. Orbital Sander
4. Power Drill
5. Coping Saw (?)
6. Electrical Tools to be Populated by Mr. Ray

Research/Development

1. The Problem: Design & Engineer an affordable High-Output LED Task Lamp, that can be made with available resources
2. The Criteria: See Posting
   
3. Initial Designs: Currently working on sketches and diagrams
4. Initial Bill of Materials: Under Construction