Regulator Woes
MuteTiefling opened this issue ยท 13 comments
Minecraft Version
1.15.2
Forge Version
forge-31.1.74
Mod Version
pneumaticcraft-repressurized-1.15.2-1.1.1-11
Describe your problem, including steps to reproduce it
I've got a small setup at this point with a Flux Compressor feeding all of my primary PNC machines. Half way down the line I regulate down to 4.5 bar with a regulator module (advanced PCB installed). The last three times I've restarted the game, that section of tube simply disappears. I don't think it's exploding as none of the pieces are laying about. All tubes are advanced too, and the section before the regulator is limited to 15 at the compressor.
Any other comments?
Nothing in the logs, then. I've tried and failed to reproduce this one... my regulators stay exactly where I expect them to on restart:
I suspect this might be the ghost air source issue I was seeing with the Flux compressor. I've since moved the regulator further along the pipe and everything is fine.
I had this happen again when updating to 21, but not in the same place. Actually, it happened in a completely different line of tubes. Baffling. I had set up a straight run of tube with the regulator in the middle so all lines in/out were straight. Logged in after dropping in 21 to find that it had poofed out of existence.
No clue how to reliably recreate this, but one thing that I've noticed is that I have a tendency to put on the regulator after air is running in the pipe, so I'm effectively using it to plug a leak. I believe the last time I repaired one that had broken, that line had effectively drained to nothing (or nearly nothing, the compressor was still running but obviously everything was venting). That one has not needed replacement since. I'll see if I can reproduce that in my testing world.
Aaannd, nope. Working fine in test.
I put down two identical compressor setups feeding into a regulator, stepping down to 4.9 to power a universal sensor. On one, I let the pressure in the line build to 19, broke the pipe off and plugged it up with the regulator. On the other I placed the regulator before starting the compressor.
After a restart they are both still present. I'll leave them in world and labelled to watch in the future for now.
Ok. So, while I could not reproduce this in test, I did just have it blow up in my face again when I loaded back to my play world. So I sealed off the tube a few tiles back then broke the tubes around where the regulator should be so they'd be empty. After placing them in and putting the regulator down on empty tubes, I then re-connected to the main line. Pressure built back up and I logged out. Upon logging back in, it did not blow up. So that's progress, even if I can't figure out what's different between what I'm doing here and in my test world.
Hrm. It's gone back to poofing out of existence the last few reloads, so I guess I wasn't on to anything in particular with the previous tests.
I've now physically split that section of tube off from the main branch by putting a charger/chest/charger section in line.
I don't know if this is because the sub section is now 'stabilized' and not really doing anything or what. The main branch is a logistics network that's constantly pushing items/charging drones, so air is constantly fluctuating and those tubes are constantly updating.
Give build 23 a go. I've made a few tweaks which might help with this problem (I think it's related to the tube being over-pressure on startup and exploding, though not sure why). This does the air dispersal logic (i.e. pushing air to lower pressure neighbours) before the overpressure checks on each tick, as well as disabling explosions in the first 20 ticks of server startup (give the system a chance to "settle down").
Not sure if it'll help, but would be interested to know...
(This build also fixes the drone external program issue, btw)
So far so good. Only one restart performed, but nothing has blown up yet. Awesome job!
I think I may have found the real cause of the problem (see also #500 where duncanwebb found the same issue in 1.14.4). Calculation error in the air handler which allowed air to "backflow" into the tube the regulator was on, causing that one tube section's pressure to rise incorrectly high - often high enough to blow up the tube, especially right after server startup.
Try build 24 and see if that helps...