The solar cap, third version
The third solar cap changes the thread, seal and assembly after two earlier enclosures failed outdoors.
We have the third version of the solar cap on the bench. It still screws onto the top of Stake, but the parts below the solar surface have changed. The first cap cracked in frost. The second enclosure showed condensation inside. We are treating both failures as unfinished work, even when a new sample looks fine after a day outside.
The cap charges a small rechargeable cell. We are designing around about three weeks of operation without sun, with readings taken every 10 minutes. That reserve matters when a bed sits under cloud or the weather keeps us indoors. It is an approximate duration, and cold temperatures and the cell’s charge at the start will affect it.
A cap we can assemble consistently
We revised the wall around the thread and the way the seal sits against the body. The cap has to close properly without asking someone to judge a tiny gap. We are fitting it with dry hands, wet hands and garden gloves, then opening it to inspect where the seal has landed.
We also changed our assembly notes. A clean, dry seal and dry parts before closing are part of the design’s practical requirements. We cannot solve condensation by writing a reassuring sentence about the enclosure. We need a repeatable process and evidence from units that have experienced temperature changes.
The body remains glass-filled nylon, with the two moisture pads at 5 cm and 15 cm and a soil temperature sensor below the cap. Keeping those positions fixed lets us compare enclosure versions without changing the sensing arrangement at the same time. The intended overall height is still 28 cm.
We are checking the radio connection with the cap fitted. The antenna changes after the garden trials need to work inside the complete body. A board lying uncovered beside a phone does not tell us what happens behind a shed.
Sunlight is a variable
We are recording charging alongside the readings. A vegetable bed can have good morning light and afternoon shade, and leaves grow around a probe during the season. Dirt on the solar surface is another ordinary condition we have to account for. We are keeping those placement notes with each record.
Our next checks will combine charging, cold cycles and water exposure. We will inspect the cap and enclosure afterward, even if the electronics still report readings. The IP67 target needs evidence from the assembled product.
This version gets us closer to ordering tooling, but it does not close the weather testing. Once we have moulded parts, we will repeat the checks. Differences between a prototype and a production part can be small enough to miss on a drawing and large enough to admit water.