Closed azerimaker closed 3 years ago
First test is conducted with two nodes (Serial no. 79 (unsealed) & 80 (VE9)). Step 1: Programmed the nodes and placed them side by side for T-30 minutes to allow their temperature to balance.
Step 2: After running them at an ambient temperature for T+10 minutes the sensor readings were in their acceptable error margin (typ. ±0.2 C, ±2.0%).
Temperature accuracy is really good with respect to the reference thermometer.
Step 3: At T+15 minutes, placed the nodes inside a fridge.
At T+30 minutes the readings are pretty close. Node 80 (sealed one) lags about ~2 minutes behind, which is negligible.
After T+55 minutes. Both temperature and humidity readings are relatively close.
Reference thermometer shows ~1 °C difference, this is most probably due to the internal heating of the PCB inside the enclosure.
Small correction: There was a bit of confusion between the membranes. Actually, VE7 is the black one with lover airflow value (290 ml/min/cm²), VE9 is the white one with much larger air-flow of value (1150 ml/min/cm²). So the above test was actually conducted with VE9 membranes and I already updated the above labels accordingly.
Starting the second round of test with the same nodes. This time the node no. 80 has VE7 (black) membrane seal, which has lower airflow value.
T+15: Sensor readings are bang-on.
Temperature readings can be confirmed with the reference thermometer. Its humidity values are way off and not reliable, so only using T values.
At T+45 minutes we can observe that the sealed sensor follows the unsealed one pretty well. Observed (but not precise) delay between the nodes was around ~3 mins with unsealed one leading ahead.
T+55 minutes: blue trace (sealed node) clearly shows the approx. delay.
Verdict: After extensive tests we can confirm that both seals are working well and depending on the use case and region we can chose either one of them. For rather harsh environments with a lot of rain, VE7 would be a better option. For milder climates and faster reaction times VE9 could be preferred.
(more findings will follow)
T+150mins:
Summary:
As part of the #36 testing I'll carry out testing different waterproof seals. Current version (v1.1) of the nodes have openings (⌀ 2mm) on the front panel to allow pressure compensation and reduce the thermal delay for each measurement. Temperature & Humidity sensor (SHTC3) sits right under the opening. Normally this opening compromise the IP65 rating of the enclosure, however, we've decided to use an air-permeable mesh stickers from GORE® and they were kind enough to supply two circular variants. Part numbers are as follows:
Why do we need this?
In order not to compromise the IP rating of the enclosure.
What is already there? What do you see now?
BOPLA offers rubber seals for pressure compensation.
What is missing? What do you want to see?
However, they're not air-permeable and therefore not suitable for temperature & humidity readings.
How do you propose to implement this?
I've prepared 2 nodes, one without a seal, other with VE7/VE9 seals. I'll carry out several tests at an ambient temperature and inside a fridge, to observe the delay between nodes.
Environment:
basic_lorawan_th
app and TTNV3 with Datacake integration.Acceptance Criteria:
Low delay and low (will be quantified later) offset between inside and outside of the enclosure, with respect to a reference thermometer.
What can you do yourself and what do you need help with?
All