Mars Hotel Summer 26 under Croplock

[gullywompr]

Mars Hotel from Lovin In Her Eyes, germination started 7/9/26

 

[gullywompr]

Day 0–10 (7/9 – 7/19) — seed to seedling, catching the diary up from the device's log.
Germination started 7/9; the run was created on the Croplock hub on 7/11 with the seed already down. The camera pod was on the seedling from 7/13: 61 photo reads that week, every one of them called "seedling", nothing flagged. The air pod sitting beside it read 24–36 °C and 30–95 % RH — that is a tray on a mat, not the tent. No climate control yet; the tent comes later.
[YOU: soak / paper towel or straight to soil, the day it broke ground, the cup it was in, what light it sat under before the tent.]
Photo: week-2026-07-22_2028.jpg (first frame the pod took of it).

 

[gullywompr]

Day 11–17 (7/20 – 7/26) — the seedling week.
From the device's log: the camera called it "seedling" 23 times and "healthy" 7 that week. It also said "pest" 9 times, every one on an out-of-focus frame and every one with the caveat that it could not say what pest — nothing was on the plant. The air pod beside it ran 20–32 °C and 46–94 % RH, average 80 %: still the tray and the dome, not the tent. No climate control yet.
[YOU: whether you domed it, when you potted up into the fabric pot, the soil and amendments — organic, plain water only.]
Photo: week-2026-07-26_1239.jpg.

 

[gullywompr]

Day 18–24 (7/27 – 8/2) — into the tent; the yellow top that wasn't.
7/27 morning the run moved to Veg and into the tent, and the hub took over the exhaust fan and humidifier the same day. First week of a new room: 23.6–33.7 °C, 51–83 % RH (avg 72). The loop humidified 149 times and vented 87, held the humidifier back 7 times because the tent was already warm and humid, and ran the fan for over-temperature three times — it was chasing the band from both sides. The camera flagged "deficiency" on the pale new top on 7/27 and 7/31; the Diagnose interview landed on the mobile-nutrient family at pH 6.2, then on "pH unknown, lockout not ruled out". By 8/4 those same leaves had aged into the mid canopy dark green: a pale new flush, not a deficiency, and the log says so.
[YOU: what you did about the pale top, if anything (plain water); whether you fed; when the reservoir went under the pot.]
Photo: week-2026-08-02_1216.jpg.

 

[gullywompr]

Day 25–31 (8/3 – 8/9) — the humid week.
25.5–28.8 °C, 59–91 % RH (avg 70). A second air pod went into the lung room on 8/1, so from here the hub knows what air the exhaust fan is pulling in. 532 fan decisions for "humidity above the band" against 4 humidify — the room was wet and the fan did almost all the work. 527 "dehumidify" advisories, and the first "leaf-thermal deficit" line: the leaf running cooler than the air, so air VPD and leaf VPD disagree. The camera said "deficiency" 5 times, still on out-of-focus frames. Daily crops of the top of the plant started 8/4.
[YOU: reservoir top-ups, any training, plant height.]
Photo: week-2026-08-09_1901.jpg — stocky, nodes stacked, on mulch in the fabric pot.

 

[gullywompr]

Day 32–38 (8/10 – 8/16) — the camera stops crying wolf.
24.8–29.4 °C, 52–76 % RH (avg 70). The camera's health gate went live on the hub this week: from 8/10 its verdict is "no finding" 53 times and "deficiency" 4, and three of those four were dismissed in Diagnose as "likely a false alarm from the camera — carry on". The fungal-risk model came online too and sat between index 45 and 70 ("elevated" to "high") most of the week, on humidity alone. 321 leaf-thermal-deficit lines: the leaf ran about a degree under the air almost continuously.
[YOU: anything you changed in the room that week; whether you saw any mildew — the log says none was found.]
Photo: week-2026-08-16_1235.jpg.

 

[gullywompr]

Day 39–45 (8/17 – 8/23) — the light gets measured.
24.0–31.5 °C, 46–89 % RH (avg 70). A spectral pod started reading the light on 8/17. 51 "leaf temp above safe cap" lines and 19 over-temperature fan runs say the canopy got warm under the lamp mid-week. The loop otherwise: 323 vent decisions, 38 humidify, 353 "in band, nothing to do". Fungal index still 45–70. Camera: 86 reads, all "no finding". For the record, the tent is an AC Infinity 3x2 and the light a Mars Hydro TS-1000.
[YOU: lamp height and dimmer setting that week, and whether you moved the lamp after the warm days.]
Photo: week-2026-08-23_1212.jpg.

 

[gullywompr]

Day 46–52 (8/24 – 8/30) — dimmed to about 65 %, and the tips.
23.3–30.6 °C, 51–81 % RH (avg 67). Lights-on and lights-off became ledger rows on 8/24. Fewer fan decisions (82) and the first "circadian VPD pulse" advisories — the loop watching the day/night swing. Leaf VPD sat at or under the 0.80 kPa veg floor from 8/26 to 8/31. Camera: 90 reads, "no finding". The fixture has been at roughly 65 % since about 8/25 with plain water in the reservoir. On 9/2 the newest flush showed a pale yellow band on the tips and margins; the best reading from the record is a flush that expanded under low VPD, not a feeding problem.
[YOU: the dimmer change and why; the day you first saw the tips.]
Photo: week-2026-08-30_1248.jpg.

 

[gullywompr]

Day 53–59 (8/31 – 9/6) — the flip.
22.6–28.7 °C, 56–81 % RH (avg 67). The run's husbandry went on record 8/31: soil, wick-watered from the reservoir under the fabric pot. 9/2 10:30 a watering entry — the only grower-entered row in the run so far. 9/5 the camera boom went up about 3 in; canopy now roughly 40 cm from the lens. 28 "vent floor" lines — the loop saying the fan cannot dry the tent below the room it draws from. Camera: 77 reads, "no finding". 9/6 at 18:55 the run advanced to Flower, day 59.
[YOU: veg height at the flip, any defoliation or training before it, why 9/6.]
Photo: week-2026-09-06_1207.jpg.

 

[gullywompr]

Day 60–66 (9/7 – 9/13) — flower week one, and the fan.
Lights-on moved to 9:00, 12/12, from 9/7. 20.2–33.0 °C, 51–86 % RH (avg 67); the lung room 51–70 % (avg 62). The disease-pressure side of the loop went live 9/11: the fungal index reached 97–100 with the tent 4–6 points above the lung room, and the fan ran as a purge — 61 purge rows, 105 declines. 9/13 the purge rule was rewritten: a purge needs a real 3-point gradient to the room it pulls from, runs at least 10 minutes, then leaves a spent lever alone. The same afternoon its retry wait held the fan off for thirty minutes while the tent climbed from 64 % to 74 % against a 61 % room — the tent showed it before the log did, and the wait now binds only when there is no drier air to pull. 63 "vent floor" lines and 51 leaf-thermal-deficit lines this week; the leaf ran about 0.9 °C cooler than the air. Camera: 33 "no finding", the rest inconclusive as the canopy fills the frame. The fan episode is also the first Short on the Croplock channel.
[YOU: first pistils, any stretch you measured, what you think of the fan running more.]
Photo: week-2026-09-13_1233.jpg — full canopy, the pod's clip top right.

 

[gullywompr]

Day 67 (9/14) — the fan and the humidifier argue, and the air sensors change.
21.4–28.1 °C, 50–78 % RH (avg 63). The fungal index ran elevated to high all day (52 to 100). The loop answered the pressure with the fan — 38 purge rows — while the humidity band kept asking for water: 25 "humidify" lines against 9 "dehumidify" the same day, and the alert at the top of the dashboard disagreed with the VPD badge beside it, caught from two screenshots. Since the fix, fungal pressure holds the humidifier off, the tent has to stay clear for a full hour before it comes back, and that memory survives a restart. That evening both air pods changed from the AHT20 to the SHT45 humidity sensor: the tent pod reporting from 21:34, the lung-room pod from 22:17 after its board was replaced. Camera: 8 "no finding", 4 inconclusive.
Luckily I didn't notice any bad smells or damaged tissue while hte index was that high. The new sensors tighten the slack between lung room and tent conditions, affecting fan operation - too much variance was making it start too late and not run for the right amount of time, much much better now.
Photo: event-2026-09-14_1403.jpg — the day's sharpest lit frame, the PPFD pod's arm at right.

 

[gullywompr]

Week two of flower. Tent ran 21.4 to 29.8C and 50 to 78% RH, averaging 66, against a lung room at 44 to 74%. The leaf sat about 1.1C under the tent air all week.
Disease pressure was the week. The fungal index went elevated or high about two dozen separate times and touched 100. The controller answered it with the exhaust fan: 136 purge runs, and 153 times it declined to purge because the vent was already spent. That decline rule went in on 9/13, and it is now saying no slightly more often than it says yes.
Humidity got argued over from both sides, 35 dehumidify calls against 28 humidify, plus 89 vent floor lines and 70 leaf-thermal-deficit lines.
The camera flagged nothing on the plant all week. 32 frames came back no finding, 47 inconclusive.
Nothing entered by hand this week, no watering or maintenance logged.
Just starting to see pistils coming in, no smell yet.

 

[gullywompr]

Device: the tent ran 20.4–26.7 °C (median 23.1) and 60–78% (median 69), with lights-on medians of 23.5 °C / 69% and lights-off 21.9 °C / 69%. The lung room ran 56–71%. The leaf averaged 22.3 °C. Humidity was the week: 47 "dehumidify" lines, 19 "vent floor", 6 leaf-thermal-deficit. The fan's disease purge changed state twice, and the readings went stale once. The camera analysed 75 frames: 53 inconclusive, 22 no finding, none a finding.

By hand: first pistils seen by eye on 9/21 (it's a female); two notes on the late-September humidity (93% that morning, 86% at 9 pm on 9/23); a note on 9/25 that stretch was pushing the plant up into the camera and new fan leaves were blocking it; and a defoliation on 9/26 at 10:56. After the defoliation the camera's view of the tops came back, and pistils are visible at several bud sites in the 8:12 pm close-up.

There are SO many bud sites, although many are below the canopy and struggling to get light. I defoliate, but you can't remove all the sugar factories (leaves) - it needs power during flower. But it looks like I'll be happy with the yield in a 2x2 space. A faint smell is evident now, but I can't classify it yet.

It's been a really healthy plant so far, with the exception of one deficiency episode as it entered veg: https://croplock.com/blog/in-band-and-sick-anyway/.

 

[gullywompr]

Device: the tent ran 21.5–28.2 °C (average 24.7) and 62–72% (average 68), with lights-on medians of 25.2 °C / 68% and lights-off 24.0 °C / 67%. The lung room ran 61–69%. The leaf averaged 24.1 °C. Quieter than week three on advisories: 11 leaf-thermal-deficit lines, 8 "vent floor", 4 "dehumidify". The fan's disease purge was written 10 times and changed state once; the mold-pressure model fired on six of the seven days, at risk 71–97 and 66–70% humidity. The light went on and off on schedule every day (9:00 / 21:00). On 9/29 at 7:36 the hub re-adopted the humidifier plug (a plug-only row; the plant would not have noticed). The week's row is 10/1 at 19:05: SYSTEM_SEX_OBSERVED, FEMALE, from pistils, camera pod, seven sightings. The pistil check went live on the hub at 12:27 that day, saw tufts on the 13:05 close-up and on every hourly one after it to lights-off, each at confidence 1.0, and the record took the call after the seventh. Ten days behind the 9/21 sighting by eye and six behind the hand-written row (9/24); it looks for tufts, not the first pair, so it will always trail a person. Since then: 12 of 12 lit close-ups on 10/2, 11 of 12 on 10/3. The camera analysed 83 frames: 46 no finding, 37 inconclusive, none a finding. On 10/2 the hub's own white-balance step was found to be feeding the health check a corrected frame it was never trained on; with it removed the check certified 10 of 12 frames the next day against 1 of 10 the day before. By hand: nothing this week except the 9/28 "Posted Page" row.

Getting pretty close to the lights... Humidity in the mid-Atlantic still exrremely high, about 93% overnight all week, I don't have the dehumidifying equipment to deal with it, so VPD is below range all week. By November, it will flip and I'l be running three humidiers at one to try to deal with how dry it is. Maintaining an even environment in your own home is tricky slash expensive, I can only go so far. But the plant looks great.

I've been removing fan leaves near the top and larving those lower below. This is the opposite of most advice (which is remove lower leaves). My logic is the fan leaves at top have to come off because not all but sites are that high and the big fan leave block them, so away they go. Instead, I keep kind of a "lower canopy" below most of the bud sites where the fan leaves don't get picked (at least not as thoroughly). Of course the light level down there is less that 3 inches higher, but it would be nil if I left the fan leaf canopy higher. But even that lower canopy below the buds keeps filling in, blocking airflow (drawing from a vent at the bottom through a port in the top), so I got work to do later.

Photos: the 10/1 12:05 frame, the 12:05 close-up of one bud site at 3x, and the hub's own card for the day (it counts 7/9 as day 1, so its day 88 is day 87 here).

 

[gullywompr]

Flower week 5, 10/5 to 10/11 (day 88–94).

Device: the week ran 20.8–26.0 °C (average 23.3) at 46.0–71.6% (average 59.5), 9,804 readings; the leaf averaged 22.4 °C. The room the fan pulls from fell to 40% at one point, the driest of the run, and the tent followed: the mold-pressure purge fired on two days this week against six last week. The loop changed a plug state 89 times; the fan declined a purge 22 times ("vent spent"), once an hour in the dark for two or three minutes at a time, a rule misfiring that is now fixed on the bench and not yet released. The week's wrong minute, 10/9 at 16:12: the leaf sensor reported 26.05 °C where the reading seventy seconds earlier was 23.87; the loop read that as leaf VPD 1.54, above the band, switched the fan off and the humidifier on at 16:13:06, and four seconds later its own advisory said raising humidity was the wrong move. The tent went 56 to 67% in one reading; the next leaf reading was 24.16; the fan was back on 68 seconds after it stopped. Why the sensor read 26 is not in the record (its housing temperature did not move). Filed as a bug the next morning; a fix is on the bench. The camera's pistil check saw tufts on nearly every lit close-up (11 of 13 on 10/10); its health read was 39 no finding / 42 inconclusive for the week, and since the 10th the lens locks at 25 cm on every dark frame against 67 cm the night before. Thirteen version rows: the hub ran eight builds this week as fixes were tried on this tent. Advisories: 55 vent floor, 15 leaf-thermal deficit, 4 dehumidify, 2 light limit.

By hand: 10/6 a defoliation (recorded 10/7, "time approximate"). 10/9 at 13:04, typed into the record: "Plant growing into the light, so I took it off the pulleys and affixed it directly to the roof poles, and cranked it from 75% to 100%. I'll see how the DLI is now." The tent's light meter moved less than two percent at the sensor.

Photos: the 10/10 noon frame (one bud site dead centre, the camera that close), and the hub's own card for the day (its day 95 is day 94 here).

 

[gullywompr]

Flower week six (day 95 to 101, 10-12 to 10-18): the fan never stops, and the box says the room is the limit
Device: the week ran 20.3 to 29.7 °C (average 24.6) at 48 to 71 % (average 64), 11,712 readings; the leaf averaged 23.7 °C. Lights-on medians 25.6 °C / 65 %, lights-off 22.9 °C / 66 %. The fan has not stopped since 10-13 09:26. It switched off once that morning when the readings went stale, came back on two minutes later, and has run ever since. On 10-17 at 11:08 the loop wrote that it had driven the fan for 2,742 minutes (it counts from the hub's last restart, on 10-15; the real run by then was about four days) and the tent was still outside its band: "The actuator cannot reach the band in this room." Its vent-floor advisory says why: the room the fan pulls from read 59 % that morning (51 % today), so venting cannot take the tent much below that. It also turned down a mold purge on 10-17 for the same reason (vent spent).
The hottest minute was 10-14 at 20:10, 29.7 °C in the tent; at 18:56 that evening the leaf read 28.2 °C, over the pack's 28.0 ceiling, and the box said so. On the morning of 10-17 the leaf-temperature check refused the leaf reading 226 times in two hours (canopy "too sparse"), and the loop steered on air VPD instead. That check was meant for seedlings; a build went on the bench at 11:16 that asks it only in the seedling phases, and it has not refused since. That fix is on the bench and not released yet. Seventeen version rows this week: the hub ran that many builds as fixes were tried on this tent.
Advisories: 28 vent-floor, 13 leaf-thermal deficit, 7 dehumidify, 1 mold risk, 1 leaf too hot. The powdery-mildew index ran 53 to 98 most of the week and did not appear on 10-17, the coolest day (average 23.0 °C). The camera's pistil check saw tufts on most lit close-ups (9 sightings on 10-17); its pollen-sac check stayed at no on every read. Health read: 26 no finding, 57 inconclusive. One caution on the 17th: the leaf read 2.1 °C below the air while lit, but the IR sensor's own body read 1.1 °C below the air as well, so that larger gap is not trusted yet.
By hand: no grower entries in the record this week.
Photos: 10-17 12:07, lights on, white pistil tufts at many bud sites.