Project

General

Profile

LED Node v2,v2a - LN

The LED Node v2 combines a JeeNode with the functionality of the MOSFET Plug - its narrow (18 x 132 mm) form factor is a good in-line fit to standard RGB LED strips.

The LED Node uses mostly through-hole parts, with the SMD STN4NF03L MOSFET’s already soldered to the board, so it’s easy to assemble. It includes the ATmega328, the RFM12B wireless module, 3 sensitive MOSFETs able to drive easily lengths of RGB LED strips (even types with 30 or 60 LEDs per meter).The MOSFET drive circuits are tied to I/O pins which support hardware PWM, providing independent, flicker-free dimming of each color. A 4-position heavy duty PCB termination matches the connections of standard RGB common-anode strips.

Note that the LED Node does not have the full set of 4 JeeNode’s “ports”. Two standard-wiring 6pin headers (Ports 1,4) are provided in non-standard positions (board end and side).

Power issues

This design pays particular attention to handling relatively large currents effectively. There is provision for adding an optional thermistor for active monitoring of the MOSFET temperature, important when driving strips near the 2A maximum recommended with high duty cycles. With more modest loading (lower maximum currents and/or duty cycle), the MOSFET’s typically stay just hand warm.

There are two useful power modes:

(A). While programming, the power can come from the FTDI connector in 5V power mode (diode D1 is required). There is insufficient voltage to turn on the LED strip (which protects the limited current output from a USB BUB or similar FTDI adapter), but a test LED + limit resistor can be probed between any colour (RGB) pin and GND for verification.
(B). Attaching the + 12 V feed will energise the LED string and the logic (see below). The FTDI supply is safely isolated by diode D1 and can be disconnected.

Note that in power mode (B), the onboard LDO is working quite hard, removing the excess (~9 V) voltage as heat. If your sketch has extended transmit times/duty cycle, monitor the LDO case temperature and fit a clip-on heat sink if required. The LDO is safely temperature-limited internally, but operation of this protection collapses the 3.3V rail, forcing the ATmega into reset until the LDO internal temperature drops lower again.

Remember that the PWR pin can be at 5V or 12V, depending what configuration choice you made. If you are adding a plug in module, check if the module uses the PWR pin and for any limitations. For example, the Relay board requires 4-6 V on its PWR pin. In the 12V case, cut the etch described above and use a suitable step down method (zener, regulator chip) to get to the required range.

For custom LED strips driven by greater than 12 V, disconnect the power/logic path by cutting the thin trace that links the + supply wide trace to the LDO then power the logic separately through a header PWR pin as described in the second section of the ‘Oops’ Weblog.

Read the ‘analog’ LED strip specification carefully. The series resistors embedded in the strip are there to limit the current draw of the individual series/parallel LED diodes. If in doubt, connect the strip directly to the power supply (that’s a 100% duty cycle) and measure the current per color. The sum of these three currents must be less than the rating of the power supply and less than 3*2A (this last restriction is relaxed if you add in the optional thermistor to detect and then limit excessive MOSFET temperatures). For extra long, single colour strips, the MOSFET outputs can safely wire in parallel.

Note that standard header pins are suitable for a total average current of only ~1A, so the commoned return pin may exceed this limit first. Use direct soldering to the oversized output pads for higher currents/duty cycles.

Errata v2,v2a

Please note the pilot production batch (v2) correctly marks the FTDI orientation on the TOP (component side) silk screen by showing the position of the GND pin. Ignore the incorrect GND marking on the BOTTOM silk screen.

Here are the correct board markings:

In the RFM12B zone, the bottom copper flood area is not grounded. For the best range from the radio module, ground this shielding device. See the ‘oops’ weblog post for the details.

Version 2a corrected this FTDI marking error and improved the clarity of the silk screen. There are no changes to the circuit and component layout in version 2a from version 2.

Related weblog posts

jlpcb-146.brd - v2 (62.4 KB) jcw, 2012-12-13 20:09

jlpcb-146.pdf - v2 (52.5 KB) jcw, 2012-12-13 20:09

jlpcb-146.png - v2 (43.7 KB) jcw, 2012-12-13 20:09

jlpcb-146.sch - v2 (269 KB) jcw, 2012-12-13 20:09

DSC_4339.jpg (114 KB) jcw, 2012-12-27 21:38

top.png (18.8 KB) jcw, 2012-12-27 21:40

jlpcb-146b.pdf (266 KB) martynj, 2013-04-13 17:25