Is there a trick or workaround for this?

[Xuhybrid]

I'm not sure if this is a new thing but i don't remember it happening in QoL3, The first packet in a series going into an Aquatuner is always ignored. So if you were to have a packet and a space repeatedly, your Aquatuner will turn on, consume power (and generate heat?) and the packet will keep moving into the overflow without being affected.

Is the only way to avoid this, to remove the overflow? It's only the first packet and it feels like some kind of timing issue. I'm not using any automation on the Aquatuner and it's fully powered.

[Gus Smedstad]

This clearly deserves a bug report.

In my applications, I typically have full pipes going into an aquatuner, though I do use shutoff valve / thermal sensors to control input. In such a system, an individual packet with an unusually high temperature can go through by itself. In practice ut's pretty unusual in such a setup for two water packets to be have significantly different temperatures, so generally this bug should be a problem.

For your particular application, your best workaround is to set up some sort of buffering system so that water doesn't enter the aquatuner in single packets.

[DonDegow]

Isn't the output of the aquatuner obstructed when this happens? (either full pipe, different liquid or weird piping without multiple outputs connected to multiple inputs)

A screenshot of your piping would be helpful

[MorsDux]

27 minutes ago, Gus Smedstad said:

This clearly deserves a bug report.

In my applications, I typically have full pipes going into an aquatuner, though I do use shutoff valve / thermal sensors to control input. In such a system, an individual packet with an unusually high temperature can go through by itself. In practice ut's pretty unusual in such a setup for two water packets to be have significantly different temperatures, so generally this bug should be a problem.

For your particular application, your best workaround is to set up some sort of buffering system so that water doesn't enter the aquatuner in single packets.

Yes if you use automation to toggle aquatuner on and off, every first packet will slip (hate it!)

[DonDegow]

It feels like a buffer/filter gate can solve this issue.

[Xuhybrid]

1 hour ago, Gus Smedstad said:

This clearly deserves a bug report.

In my applications, I typically have full pipes going into an aquatuner, though I do use shutoff valve / thermal sensors to control input. In such a system, an individual packet with an unusually high temperature can go through by itself. In practice ut's pretty unusual in such a setup for two water packets to be have significantly different temperatures, so generally this bug should be a problem.

For your particular application, your best workaround is to set up some sort of buffering system so that water doesn't enter the aquatuner in single packets.

Yeah i went away to watch something while thinking about it and reached a similar conclusion. I use a similar system with a liquid shut off, so a buffer would likely solve the issue, but the main problem comes when my reservoir are full and i toggle off the input. I've been doing that with a shut off as well and it leaves sporadic packets in the loop which never leave. And it's now i realise that i should toggle the aquatuner when the reservoir is full instead, and leave the loop full.

Thanks for the feedback. I'll definitely file a bug report, though i can't have been the first to notice.

1 hour ago, MorsDux said:

Yes if you use automation to toggle aquatuner on and off, every first packet will slip (hate it!)

See, this is where i used to experience the bug. It's why i started using the liquid shutoff in the first place, but this is with the aquatuner always on. It seems to miss the first packet no matter what. Making the shutoff necessary for temperature. Perhaps i just didn't notice this "feature" before.

1 hour ago, DonDegow said:

Isn't the output of the aquatuner obstructed when this happens? (either full pipe, different liquid or weird piping without multiple outputs connected to multiple inputs)

A screenshot of your piping would be helpful

It can be just a single packet all on it's own in a loop and still not go in.

[nakomaru]

I believe the issue is that aquatuners take one second to turn on, and they probably consume power but I don't imagine they generate heat during that time.

You can design a non bypass tuner, or do a variety of other alternatives to limit the losses from this. I like non bypass tuners. The coolant that enters is always intended for it.

[MorsDux]

1 hour ago, nakomaru said:

I believe the issue is that aquatuners take one second to turn on, and they probably consume power but I don't imagine they generate heat during that time.

You can design a non bypass tuner, or do a variety of other alternatives to limit the losses from this. I like non bypass tuners. The coolant that enters is always intended for it.

Of course solutions exist they are just less elegant as simple automation would provide if it worked as intended / should. 

[Xuhybrid]

8 hours ago, nakomaru said:

I believe the issue is that aquatuners take one second to turn on, and they probably consume power but I don't imagine they generate heat during that time.

You can design a non bypass tuner, or do a variety of other alternatives to limit the losses from this. I like non bypass tuners. The coolant that enters is always intended for it.

If the loop stops then the thermo sensor can let through a bad packet and break the aquatuner or on the flip side, allow a lot more uncooled liquid out of the loop. I wouldn't need anywhere near as much automation or piping workarounds if it functioned as expected

[Xuhybrid]

So i've spent some time today pondering the responses and managed to reach a conclusion i'm happy with. By putting a buffer inside the loop, removing the aquatuner overflow and moving the shutoff inside the loop as well, everything in the loop is stored in the buffer, no packets get missed or blocked and i'm bypassing the bug entirely. Thanks for the feedback.

Spoiler

For anyone curious about what i did. Built in sandbox world but i emulated my survival build with a refinery loop on the left.

Found a bug report here from February not confirmed by Klei. Be sure to drop by.

 

[nakomaru]

Yep, that's basically what I meant. You might consider connecting the temp sensor to the shutoff and having no automation going to the tuner. If you don't have perfectly insulated pipes and the aquatuner needs to stay off for a long time due to high temp, you can have coolant waiting inside the chamber and boil.

This will also give a delay between the heat generation and the on signal, and should act to reduce thrashing on/off.

[Xuhybrid]

The regular temp sensor is just to stop the aquatuner going over safe limits and the steam turbine keeps the temperature well below that safe limit. It's just there as a fail safe if something goes wrong with the turbine cooling.

Edit: Just realised what you were suggesting. I see no drawbacks to making this change. Thanks!

[WanderingKid]

I use a twin bridge setup to deal with the spinup delay on the aquatuner.  It's been there for a very long time, btw.  This lets the 1 second delay wait while it spins up and it'll pass through when it's done.