"Trad Rack" with "Belay" in RRF

[Exerqtor]

I'm in the start phases of builing a Annex Engineering "Trad Rack" for my RRF driven Voron using a Fysetc StrideMax Dual as the controller for the MMU system (so I've got that part of the equation sorted so far).

I've how ever also thought about trying to impliment the Annex Engineering "Belay" extruder sync sensor into this setup. But i'm a little bit worried the constant monitoring and adjustement needed with such a compensation system (which i suspect will have to be done through daemon.g) will put to much strain on RRF.

Am I overthinking this or is this a shared concern by others here as well, might as well just be me having an unjust feeling that a heavy daemon.g is a somewhat ticking time bomb in RRF's current state (since it's often the refered to as a potential cause if someone has issues).

---

Also i didn't know what forum category was best suited for this discussion, so please move it if "General Discussion" wasn't the correcnt one 😅

[oliof]

Push/Pull setups have been done with RRF before, without something like the Belay sensor. I really think it's a bit of a weird concern with modern industrially produced filament feeders and well dimensioned filament.

I also don't see how Belay will work well with retracts/unretracts and filament swapping, but maybe I am overly sceptical. On the other hand, it's an alpha device.

Personally, I would first check whether there is an actual issue with slight differences between push and pull feeders, or whether the inherent springiness of the bowden system is enough to dampen the effect. As they say in the Belay description, in the usual range of 0.95 to 1.05 esteps factor adjustment, there is barely any visible motion.

[oliof]

And answering your question about daemon.g: Yes, additional checks in a more or less right loop will cause an overhead. Depending on the MCU in use and the free overhead this may or may not be an issue. We know that with an STM32H7 board, movement at 2100mm/sec and acceleration of 100,000mm/sec^2 is possible. So you may just need to pick the right hardware for your use case.

[oliof]

also, maybe this would be a case for an external board querying the Belay measurement and just sending adjustments to the RRF board via serial or http.

[T3P3Tony]

@Exerqtor we have seen people using various systems with push pull dual extruders in the past and they are relatively straightforward to implement.

@oliof they are generally needed, if you have such a long filament path or a heavy spool that you need this sort of mechanism then you are probably aiming to do large prints, use a lot of a filament and any difference in the two extruders will eventually become apparent.

The specific device referenced I have not heard of before. Looking briefly at the video on the GitHub you could do this with two triggers at each end of the belay range of motion. one trigger would set extrusion rate of the push/helper extruder to the max of the window (e.g. 1.05) the other to the min (e.g. 0.95).

You could also have a look for other systems that achieve the same. I have seen simpler solutions (@oozeBot do i remember you having a solution for this or am i imagining it?)

[T3P3Tony]

@Exerqtor that's not to say you can't use daemon.g to monitor the inputs if you wanted to and adjust it that way. a loop every couple of seconds should be sufficient so its not going to be a big additional load.