I am new to 3D Printing and the X1-Carbon is my first printer. Had it for about 2 weeks, please bear with me if I say nonsense.
While I was printing a large piece using Bambu PLA Sparkling, the nozzle got clogged after 2 hours printing and I didn’t notice it for at least another 30 minutes. There was no spaghetti to be detected and stop the print, it was just not extruding, the yellow gear visible on the outside was spinning and when I unload the filament, I noticed it was grinded by the gears slipping on it. I guess all of this is expected, I believe nozzle clogging may happen from time to time given that the printer comes with a needle to unclog it.
But then I was thinking: “this thing has a precise LIDAR!” there was a large gap between the nozzle and the print without material, why not using the LIDAR to detect this gap and stop the print automatically?
I agree to your argument, but only assuming that LIDAR cannot be used while printing. I noticed that the printer does first layer entirely then uses LIDAR but that may be because it needs to read first layer entirely to assess quality and not necessarily because it cannot do both at once.
If it can do both at same time, then it could account for some tolerance. For example, if layer height is 0.2mm, if LIDAR doesn’t detect material - say - 5x greater than this then stop the print. That would account for current layer + 4 other missing layers.
The print would still be lost after skipping 5 layers, but at least would reduce the waste of time not knowing that print failed.
Yes, but the printer knows what is the offset between the LIDAR position and nozzle, and instead of checking whole layer it could check the region it passes through to avoid stopping print - if the nozzle gets clogged it won’t print entire layers, so even if it checks part of the layer, it could still find the problem.
I was thinking further about this today, and even if the printer can’t use LIDAR while printing, it could still do a “check layer height every X-minutes” so it wouldn’t impact printing performance as much.
It could print for 5 minutes, then stop and check, print another 5 minutes and check again. For those very long prints that you leave running over night, it could be useful.
I would imagine they could use the load cells in the bed to find these issues too. The load cells are generally capable of detecting the pressure of printing, so it’s just up to Bambu to utilize the data in the software. It wouldn’t be a quick coding change and would likely cost a bit of money to implement, but that’s the next step.
If the printer had a shaft-encoder wheel that was driven by friction with the filament being pulled in to the toolhead, the printer could detect if the filament was advancing when it was supposed to and detect when the filament is stripped and/or the nozzle is clogged.
But how valuable is something like this? If the print is running overnight, or when you’re away, whether it’s “air printing” or not doesn’t matter until you wake up or come back from wherever you were. So, so what if the printer has continued to print nothing while you slept, or were away? Until you get back in front of the printer there’s nothing you can do about it except stop the print job. But letting it continue to run doesn’t really hurt anything. Stop it when you finally realize it’s not printing anything and the only difference relative to the printer detecting that itself is the cost of a KWH or so of electricity.
Spaghetti detection is a completely different story. There, you definitely want the printer to stop so it doesn’t waste material and create a big mess you have to clean up…
In these 2 situations it doesn’t make much sense, but when it happened to me I was sitting right besides the printer doing other work on PC and I lost a good 30 minutes maybe more that I could have stopped the print and restarted it. It was just an annoyance.
I saw on YT a guy that had a farm of P1Ps printing 24x7 as means of his income, I guess someone using the printer professionally for profit would like to avoid wasting 30 minutes, that would mean less throughput and less money in his pocket.
A shaft-encoder wheel would also work, but then another component that would make the printer more expensive. I was wondering if this is something doable with the current hardware and just a software update.
@just4memike idea above using the load cells, if it is doable, it would just be a software update too.
It would be a firmware update but that’s assuming the load sensors they are using are sensitive enough. In general… yeah they should be but seeing that they failed to consider people were going to add enclosure for a printer, it’s possible the sensors are bargain basement and could have issues with such low loads.
Either way, if the hardware is up to the task, it would still mean Bambu Lab would be spending money updating a product that’s probably seen most of its sales when they could be working on the next product. So, I wouldn’t hold my breath.
A person could probably retrofit their own alarm with an encoder and a simple circuit. Outside my scope of abilities.
Programming an acknowledgement between the extruder and buffer/hub may be possible in the printer firmware, then you are only getting warning if the AMS is utilized.
I would like to see a exposure silder for the cam. Many of us mod the lighting. But it doesnt matter how bright the lights are, the camera adjust the exposure too much and the picture is too dark. This is a problem when I print with darker filament.
I’ve thought that the AI approach to spaghetti detection was iffy at best. I’ve had prints fail that it didn’t detect, and multiple times I’ve had false positives. Plus, if a part is hidden behind another part the camera is worthless.
I agree the lidar scan would be a better solution, and that it would be slower, and implementing it would be non-trivial. But it would be nice to have the option.
