Garden environmental telemetry project – part 2

Garden environmental telemetry project – part 1 laid out plans for a simple maker / DIY IoT garden environmental telemetry system.

This article reports tests on three sensor configurations, all using a Pt100 sensor with 4-20mA converter.

This article reports

  • Pt100 remote from a 4-20mA converter (ie long Pt100 3 wire wiring);
  • Pt100 with co-located 4-20mA converter (ie long 4-20mA loop wiring);

Pt100 remote from a 4-20mA converter (ie long Pt100 3 wire wiring)

The 4-20mA sensor is 8m of CAT5 cable from the Pt100 element. The Pt100 signal sensitivity is about 400µV/° @ 15° (25000°/V).

Above, some experiments in AC powering and earthing shows the analog sensor chain prone to noise. The first half of the display is powered from a small power bank (18650 cell with switched mode boost converter). The second half is with the 5V derived from a DIN rail mount Chinese 230VAC power supply (copy of Meanwell MDR-2–5), not the quietest of output.

The wiring was swapped for a shielded three core cable, and it roughly halved the jitter, still not good enough.

The configuration is very prone to noise due to the long Pt100 connection which is quite sensitive.

Pt100 with co-located 4-20mA converter (ie long 4-20mA loop wiring)

The next experiment replaced the sensor with a Pt100 element with -50°-50° 4-20mA converter in the probe head (Pt100 three wire connection about 200mm inside the stainless steel probe. The cheap Chinese probes might look like they are all metal and so shielded, but the cap on this one is plastic.

4-20mA loop wiring is one pair of a CAT5 cable of 8m length.

Above, the left hand half is with 8m of CAT5 Pt100 wiring to the 4-20mA converter, the right hand half is with the 4-20mA converter in the probe head and 8m of CAT5 for the 4-20mA loop. The RMS jitter in the second case is 0.12°.

With the burden resistor used, the sensitivity of the 4-20mA loop is approximately 10mV/° (100°/V), which makes the peak quantisation jitter 0.1°.

The measurement jitter is only a little worse greater than the quantisation jitter, might be improved with shielding and a lower noise 5V supply, but the continued efforts will be put into a digital sensor connection.

The next trial will be of a cheap Chinese soil temperature and humidity sensor (~$30 inc shipping) with MODBUS digital connection… probably using a LoRaWAN backhaul.

To be continued…