Earth Notes: On Website Technicals (2026-05)
Updated 2026-05-21 10:03 GMT.By Damon Hart-Davis.
: Damping powermng
Currently there are oscillations as the battery is discharging towards the minimum state of charge (SoC) allowed for dumping.
This is in part because there seems to be not enough compensation for sag under load from internal battery impedance, and wiring impedance from battery terminals to the charge controller.
I have bumped that estimated impedance by ~50% from 85mΩ to 128mΩ (sagLAmVperADumpCorr). The new value is also possibly faster to calculate with. This value seems a little too high (at 18°C and near fully charged), but for now provides a little hysteresis.
I also switched to using battery voltage on the same ~1s filter as current for the SoC estimation to try to improve effective accuracy.
A powermng log sample though the oscillations. Observe the UC SoC value and B1 battery voltage:
2026/05/19T21:30:06Z AL 0 B1 12397 B2 -1 P 12348 BV 12073 ST OK - e A1P 0 B1T 18 UC 59 A1V 0 2026/05/19T21:40:06Z AL 0 B1 12580 B2 -1 P 919 BV 12380 ST OK - t A1P 0 B1T 18 UC 71 A1V 0 2026/05/19T21:50:06Z AL 0 B1 12589 B2 -1 P 894 BV 12459 ST OK D e A1P 0 B1T 18 UC 73 A1V 0 2026/05/19T22:00:06Z AL 0 B1 12394 B2 -1 P 12010 BV 12056 ST OK D T A1P 0 B1T 18 UC 58 A1V 0 2026/05/19T22:10:07Z AL 0 B1 12394 B2 -1 P 12308 BV 12073 ST OK D T A1P 0 B1T 18 UC 58 A1V 0 2026/05/19T22:20:06Z AL 0 B1 12378 B2 -1 P 11995 BV 11986 ST OK D T A1P 0 B1T 17 UC 51 A1V 0 2026/05/19T22:30:06Z AL 0 B1 12357 B2 -1 P 11974 BV 12012 ST OK - e A1P 0 B1T 17 UC 47 A1V 0 2026/05/19T22:40:06Z AL 0 B1 12540 B2 -1 P 916 BV 12363 ST OK - t A1P 0 B1T 17 UC 60 A1V 0 2026/05/19T22:50:06Z AL 0 B1 12549 B2 -1 P 917 BV 11564 ST OK D e A1P 0 B1T 17 UC 62 A1V 0 2026/05/19T23:00:06Z AL 0 B1 12357 B2 -1 P 11913 BV 12047 ST OK D T A1P 0 B1T 17 UC 47 A1V 0 2026/05/19T23:10:06Z AL 0 B1 12357 B2 -1 P 12098 BV 11222 ST OK D T A1P 0 B1T 17 UC 47 A1V 0 2026/05/19T23:20:06Z AL 0 B1 12354 B2 -1 P 11910 BV 11924 ST OK D T A1P 0 B1T 17 UC 47 A1V 0 2026/05/19T23:30:06Z AL 0 B1 12351 B2 -1 P 11969 BV 12003 ST OK - e A1P 0 B1T 17 UC 46 A1V 0 2026/05/19T23:40:06Z AL 0 B1 12510 B2 -1 P 889 BV 12398 ST OK - t A1P 0 B1T 17 UC 54 A1V 0 2026/05/19T23:50:06Z AL 0 B1 12510 B2 -1 P 889 BV 12380 ST OK - e A1P 0 B1T 17 UC 54 A1V 0
Seems better in size of SoC ripple than yesterday:
2026/05/18T22:30:06Z AL 0 B1 12433 B2 -1 P 12110 BV 12091 ST OK - e A1P 0 B1T 17 UC 55 A1V 0 2026/05/18T22:40:06Z AL 0 B1 12620 B2 -1 P 859 BV 12468 ST OK - t A1P 0 B1T 17 UC 76 A1V 0 2026/05/18T22:50:06Z AL 0 B1 12626 B2 -1 P 859 BV 12477 ST OK D e A1P 0 B1T 17 UC 77 A1V 0 2026/05/18T23:00:06Z AL 0 B1 12433 B2 -1 P 11961 BV 12100 ST OK D T A1P 0 B1T 17 UC 55 A1V 0 2026/05/18T23:10:06Z AL 0 B1 12430 B2 -1 P 12045 BV 11222 ST OK D T A1P 0 B1T 17 UC 54 A1V 0 2026/05/18T23:20:06Z AL 0 B1 12430 B2 -1 P 12082 BV 12047 ST OK D T A1P 0 B1T 17 UC 54 A1V 0 2026/05/18T23:30:06Z AL 0 B1 12430 B2 -1 P 12045 BV 12021 ST OK - e A1P 0 B1T 17 UC 54 A1V 0 2026/05/18T23:40:06Z AL 0 B1 12589 B2 -1 P 919 BV 12442 ST OK - t A1P 0 B1T 17 UC 70 A1V 0
I am taking the opportunity to reduce complexity elsewhere in the code.
: over-damped
The estimated impedance currently set as 256mΩ is likely far too high, but may be better matched when it matters more to avoid damage at low SoC. This value eliminated oscillations, and the dump load stayed on all night as intended.
Because of generation via the PWM solar controller (eg the 100Wp panel), and load direct from the battery such as the AA charger in the kitchen, there is a limit to the precision that this area of code can achieve anyway. For example, this huge jump in reported SoC is probably due to the former:
2026/05/21T05:40:06Z AL 264 B1 12388 B2 -1 P 12277 BV 12038 ST OK D e A1P 3141 B1T 18 UC 66 A1V 27567 2026/05/21T05:50:06Z AL 283 B1 12369 B2 -1 P 12172 BV 12003 ST OK D e A1P 3428 B1T 18 UC 61 A1V 28025 2026/05/21T06:00:06Z AL 305 B1 12360 B2 -1 P 11953 BV 12038 ST OK D e A1P 3790 B1T 18 UC 61 A1V 27878 2026/05/21T06:10:06Z AL 559 B1 12510 B2 -1 P 12185 BV 12126 ST OK D e A1P 6916 B1T 18 UC 77 A1V 28757 2026/05/21T06:20:06Z AL 793 B1 12635 B2 -1 P 12433 BV 12319 ST OK D e A1P 9996 B1T 18 UC 88 A1V 27149 2026/05/21T06:30:07Z AL 839 B1 12592 B2 -1 P 22150 BV 11257 ST OK D e A1P 10811 B1T 18 UC 100 A1V 29114 2026/05/21T06:40:06Z AL 759 B1 12500 B2 -1 P 22113 BV 11915 ST OK D e A1P 9482 B1T 18 UC 100 A1V 28568
I think that I shall leave this higher value in place for now, but keep watching.
: Podbean Vanished?
Since about the Podbean wild polling has evaporated, with the last couple of obvious entries:
3.228.X.X - - [22/Apr/2026:14:19:49 +0000] "GET /rss/podcast.rss HTTP/1.1" 200 150293 "-" "Podbean/5.6 FeedParser" 51.81.X.X - - [23/Apr/2026:14:28:47 +0000] "GET /rss/podcast.rss HTTP/2.0" 429 711 "-" "Podbean/FeedUpdate 2.1"
There has been only one obvious Podbean poll since the start of this month from
Podbean/FeedUpdate 2.1 so I wonder if something finally noticed no updates to the feed since mid-January?
Meanwhile the naughty list is:
# A non-empty txt map lookup of the %{md5:%{HTTP:User-Agent}} means bad!
# Built: 2026-05-03T08:14+00:00 (UTC)
# MAXHITSPERUAPERDAY: 10
# MAXUAS: 25
#----------------
# request-count User-Agent
# MD5hash approx-hits-per-day
#----------------
# 203 Spotify/1.0
4582d9bdbcef42af27d89da91c6eb804 203
# 138 Google-Podcast
8dea568b39db0451edd6b30f29238eaf 138
# 95 atomstr/0.9.13
fed9820e6b1b919d848e718e0bf49e41 95
# 75 Amazon Music Podc
d69be2563c9f1929edf2906d41809aea 75
# 62 iTMS
97f76eb7e02c5ff923e1198ff1c288cd 62
# 31 MuckRackFeedParse
62b46fff1cf5f8af7b4b37a2f783b57a 31
# 27 axios/1.6.8
b534882134248c9a5957e0c011a37037 27
# 19 itms
2e7f714a929b3f52f3c094710819a99a 19
# 17 Mozilla/5.0 (Wind
6b9a00393fb1607b0ada13520f814ab5 17
# 17 Mozilla/5.0 Apple
fac378ac3b3d3886829021b3309d4fd1 17
# 12 Mozilla/5.0 (Maci
c6ad665e86260e0f285ae36e4f62e527 12
uptime
15:30:00 up 244 days, 3:00, 1 user, load average: 0.53, 0.96, 0.76
Utility summary
At 16WW over 7 days from 2026-04-27 to 2026-05-04: electricity import 0.5kWh/d (net -9.8kWh/d).
At 16WW over 14 days from 2026-04-20 to 2026-05-04: electricity import 0.5kWh/d (net -11.6kWh/d).
At 16WW over 28 days from 2026-04-06 to 2026-05-04: electricity import 0.6kWh/d (net -9.5kWh/d).
At 16WW over 368 days from 2025-05-01 to 2026-05-04: electricity import 3.1kWh/d (net -1.6kWh/d).
