Thermo Regulator fails with bridge bypass?

[Sevio]

I decided to finally get around to making some better food for my poor duplicants who have been living on meal lice for a long time, so I wanted to make a big, walk-in fridge cooled with a Thermo Regulator. I built a standard hydrogen radiator setup with a packet combiner to reduce the Thermo Regulator's throughput to half and keep it from overheating. A valve splits the packets into 500 g to ensure the radiator is always full. The problem is, my gas bridge bypass that ensures the radiator is always circulating is making the Regulator fail to run at all.

Every time a packet passes by the Regulator's input, it does its little animation and the info window says "Cooling", but it doesn't actually heat itself up or take heat from the packet. Instead the packet continues on to the bridge. Why doesn't the Regulator try to absorb the packet? If I remove the gas bridge, the regulator works fine.

[Saturnus]

This might sound strange but have you tried moving the bridge one step right?

[Sevio]

1 minute ago, Saturnus said:

This might sound strange but have you tried moving the bridge one step right?

I have, unfortunately it gave the same behavior.

[ImperialWolf]

I found the same issue and my fix was to remove the bridge and put in a valve - When I want the regulator running I close the valve, when I want it to bypass I open the valve. Interestingly enough I have seen it still work for a short while and then after a few cycles it stops if the valve is open. Also I have been playing with packet totals as the behavior changes on 90% full lines and 75% full.

[Reaniel]

23 minutes ago, Saturnus said:

This might sound strange but have you tried moving the bridge one step right?

In my experience, this would help a bit, but would still stall it from time to time for some reason.

I've seen this happen in my Oxygen liquefier setup when I have 3 thermal regulators hooked up together.  For whatever reason, the first machine would go off and on while the rest are running smoothly, despite having the exact same bypass mechanism using bridges when powered off. 

It eventually settled off and ran much more smoothly after a few deconstruct/rebuild + refill of hydrogen, but I'm still seeing occasional power-off of the first machine.

EDIT:  I've just found an old AU save with the 3rd machine having this problem... lol...

[Sevio]

It sounds like this bypass bug might have been introduced in the Oil Upgrade if it's happening to old saves that previously worked fine.

Edit: Or perhaps this could be related to a non-full throughput radiator like I am using here... I distinctly recall a working bypass in this upgrade when I was setting up a demo oxygen liquefier.

[manu_x32]

@Sevio I loaded and old save with lots of thermo regulators and aqua tuner bypasses, and they still all work.  I think it's your non-full throughput and/or your packet re-combiner causing the issue. I also typically put a valve before the machine.  I think there needs to be a valve somewhere on the pipes but not sure if putting it before or after the device are both ok.

[Reaniel]

1 hour ago, manu_x32 said:

@Sevio I loaded and old save with lots of thermo regulators and aqua tuner bypasses, and they still all work.  I think it's your non-full throughput and/or your packet re-combiner causing the issue. I also typically put a valve before the machine.  I think there needs to be a valve somewhere on the pipes but not sure if putting it before or after the device are both ok.

Erm, you might want to check the packet temperatures, because your screen shot clearly shows a packet of hydrogen not going into the machine but into the bypass bridge.

[MidnightSteam]

Hmm, I don't know about the Thermo Regulator bypass valve but here's my version. In the red square is where the bypass happens. When the regulator is backed-up the packets will switch to the horizontal bridge. Yes, you'll have 3 packets sitting in the other pipe before the radiator; well, you can shorten it so it'll be 2. Ahh I see why you guys designed it that way. I tried replicating your design but only some packets were being cooled; usually the second packets.

Picture:

• Two gas bridges
• Packet combiner is by the gas valve(set to 1000g)
• Long pipe right after the regulator's exit so packets don't cause a jam and slow the regulator down
• Turn thermo switch off to see gas bypass the regulator

Might not be exactly what you wanted but here's the save files. Maybe it'll give you some ideas.

• Before hydrogen enters the system so you can make changes - Hydrogen Entering.sav
• After - Hydrogen Entered.sav

 

[Kiako]

I had the same problem and it seems that the hydrogen can't "see" an exit of the cooler because of the bridge's exit. The hydrogen is trapped in the cooler.

This works better :

[Saturnus]

41 minutes ago, Kiako said:

I had the same problem and it seems that the hydrogen can't "see" an exit of the cooler because of the bridge's exit. The hydrogen is trapped in the cooler.

This works better :

This is effectively the same I suggested above that he said he tried.

[Whispershade]

It is my experience that it takes a second for a thermo regulator to cycle up from a power down state and for runthrough pipes the first packet in a series will keep going.

I have noticed the behavior since AU and is one of the reasons my early designs used switches and filters to designate flow.

I think you are better of just taking the energy hit and process the halfsized packet unless you to do something more for the regulator cooling. Though, once the room being cooled stabilizes. The regular will operate only intermittently, particularly with a full pipe of full packets. So it will have down time to cool anyway.

[MidnightSteam]

Ahhh, here we go! Combination of the two. Setting gas valve to 500g and combining right before the regulator.

 

Here's the save as usual: Hydrogen Entered Two.sav

[manu_x32]

7 hours ago, Reaniel said:

Erm, you might want to check the packet temperatures, because your screen shot clearly shows a packet of hydrogen not going into the machine but into the bypass bridge.

ah ah, yes, sorry for the confusion, the thermo regulator is off in the snapshot, but it does work properly, tested multiple times.

[Sevio]

@MidnightSteam I tried your version and while it sorta worked, when you try to fill the radiator using a pipe bridge it fills the pipes up so much that everything grinds to a near halt and it's only processing a packet once every 4 seconds or so. I tried filling it at different points but because the packets get confused once in a while it allows the packets to bunch up in the radiator and then it gets clogged up.

I settled on this version which does use one extra bridge but avoids packet confusion and can be filled with the indicated ghost gas bridge:

[MidnightSteam]

@Sevio HAHA yes! I love it So much faster!