Room Board

There are some references to v1 to document older boards, but please refer to the new version 2 files attached.

The Room Board is a small board with sensors for in-house monitoring of some basic parameters:

  • temperature - either an SHT11/HYT131 sensor or a DS18B20 1-wire sensor
  • humidity - via that same SHT11/HYT131 sensor
  • motion - one of two different types of passive infrared detector (PIR)
  • light - via an LDR

Different configurations are supported – from including the relatively expensive SHT11/HYT131 to omitting the motion sensor. Depending on configuration, 1 to 3 resistors have to be added.

The Room Board is placed over opposite ports on a JeeNode (1+4 or 2+3), like a mini-daughterboard.

Hardware Configuration

There are no switches or jumpers to configure on the Room Board PCB after construction, you must choose which components to add when building.

  • Either PIR or EPIR can be used, but not both, and either SHT11/HYT131 or DS18B20.
  • If a Zilog EPIR is used, the LDR must be connected to LDR2, else to LDR1.
  • If the EPIR sensor is used, then R1 and R2 must also be added (both 100 kΩ).
  • If the DS18B20 1-wire sensor is used, then R3 must be added (4.7 kΩ).

See this forum post for details about how to match up the sketch and room board orientation.

Driver Configuration

The original rooms.pde was replaced by a newer design, the roomNode.ino sketch.

The roomNode.ino sketch supports various configurations. (Un)comment the definitions at the top of the file and set your chosen port number. The sketch will then automatically load a properly configured driver for the various sensors/options selected.

N.B.: Thanks to Forum user pavinder95, support was recently enhanced (Feb2014). Be sure to pick up the latest version from GitHub

PIR Module - general issues

There are several PIR module variants with differences in power requirements and signal output configuration.
The following refers to all variants.

At power up, the PIR module can spend around 2 minutes adjusting to the background infrared (IR) radiation, with multiple false 'EVENT DETECTED' outputs.

The "warm up" behaviour is from the initialisation of the combined analogue/digital motion detection circuit. The IR detector itself is fast, but with a tiny output signal mixed with a lot of noise. The amplified sensor output is integrated with a long time constant to estimate the current background IR radiation, then this level is adjusted to fall roughly centred between an upper and lower limit. Event triggering then occurs on a rapid input change that goes outside that 'dead band'. Very slow changes (e.g. as the room temperature changes) are ignored and gradually re-adjust the position of the 'dead band'.

The design depends on the multi-faceted Fresnel lens - lots of small observation zones are interleaved, so IR radiated from a body in motion passes from 'odd' zones to 'even' zones and back in succession, generating an 'up' then a 'down' event at the threshold detector. Those get 'OR'd together and drive a re-triggerable monostable. The on-board variable resistor adjusts the 'EVENT DETECTED' (Tx) time period over a wide range from ~1second to minutes.

The two-position jumper directly controls Pin1 on the BISS0001 chip (H and L are marked on the silkscreen), centre pin goes to P1, H is Vcc, L is GND. This is to enable (H) or disable (L) re-triggering of the output signal, meaning if there is further motion detected during the EVENT DETECTED state time (length Tx), then extend the EVENT DETECTED state for at least another Tx.

The driver is only interested in events and applies its own time constant on how fast/often to report multiple events. Best settings for the driver are to minimise Tx (pot fully CCW) and enable re-triggering (Jumper H).

Positioning the sight line of the lens is important - e.g. pointing to an area that receives strong sunlight can generate a triggers from passing clouds.

Note: these detectors are NOT suitable for PRIMARY intrusion alerting. There are well known (to the dishonest community) tricks that can defeat most PIR designs when positioned outdoors.

PIR Module - current version

The PIR module is a constant drain on the battery. The Room Board has used various variants of the module in the search for better battery life. The ELV PIR sensor was replaced September, 2010 by one with better specifications (link below) and we continue to work with the vendor on improvements.
The current shipping module has achieved very low quiescent current. The latest change removed buffering and any pull-up resistor on the output signal - driving the output pin directly from the EVENT DETECTED signal. R1 (100KΩ) is no longer required.


If the DS18B20 sensor is used, it needs mounting differently from the labeling on the board!

Header Pinouts

Design Info

Dimensions: 21.1 x 21.6 mm
CadSoft EAGLE design files: see below.
This plug can be operated at 5V as well as 3.3V - but only without EPIR present.

This board uses the conventions of a JeePlug and can be used with the ports on a JeeNode. It can also be used with other boards and MCU's, if you hook it up properly and adapt the interface code for it.

How to Get It

Related Weblog posts

There's an excellent post on the forum about how to use the PIR with the Room Board (note that the output state descriptions there refer to one particular variant of the PIR module only).

rb1.DSC_0717.jpg (11.5 KB) myra, 2012-07-17 15:16

rb1._00000317.png (39.4 KB) myra, 2012-07-17 15:16

rb1.DSC_0716.jpg (8.42 KB) myra, 2012-07-17 15:16

rb1.DSC_0712.jpg (7.55 KB) myra, 2012-07-17 15:16

rb1.DSC_0711.jpg (11.7 KB) myra, 2012-07-17 15:16

rb1.DSC_0714.jpg (14.7 KB) myra, 2012-07-17 15:16

rb1.DSC_0715_large.jpg (11.8 KB) myra, 2012-07-17 15:16

PIR_specs.png (39.6 KB) myra, 2012-07-17 15:17

jlpcb-119.brd (12.9 KB) myra, 2012-07-17 16:34

jlpcb-119.pdf (11.3 KB) myra, 2012-07-17 16:34

jlpcb-119.png (80.7 KB) myra, 2012-07-17 16:34

jlpcb-119.sch (58 KB) myra, 2012-07-17 16:34

HYT-131.pdf - HYT131_spec_sheet (586 KB) martynj, 2014-02-12 14:01