Idea on how to add plate type detection on the A1 series

Hello @BambuLab,

My idea is to add holes on the back of the plate. In the area where the plate detection probing is done.

When doing the plate detection the toolhead will probe the yellow point to confirm that a plate is present and then will probe the two other positions (red and blue).

This technique could detect up to 4 plate types.

  • no holes
  • red hole
  • blue hole
  • red and blues holes

I think it would be a real plus for A1 Mini and it can certainly be done on the A1 as well.

Of course this requires to update the production line of the build plates. But I guess adding two holes on a plate will be easy to do.

And a 3d printable JIG could be done to help users with “older” plates to drill holes at the right locations.

If the toolhead is probing positions without holes, it will damage the nozzle and the plate over time.

Another idea. Why not OCR the plate text with the included camera, if possible?
On my plates it says: Bambu textured plate, Bambu Engineering plate, Bambu high temperature plate. On the last one the text is a bit less visible but I guess they can see the color of the plate, as each one is different.

That shouldn’t be an issue. The nozzle is already probing and touching the build plate every time it’s homing z, doing a bedmesh, …

I think the only bambulab printers that have the plate type detection is the X1 series and uses the “LiDAR” camera to read a small qr code on the front of the build plate. Same idea than OCR but easier to implement I guess.

All these ideas are controlled by the main board. Can the current MC handle this operation? I know it’s why the P1 series can’t do this yet.

If you look on bambu labs store they state “spaghetti detection coming later” on the P1P camera.

I see that build plate detection is already added now in the new firmware for the a1 mini… how does it work ? there are no qr codes on the plates so does it just probe and then touches the surface to identify the thickness of the build plates ?

I believe it is build plate detection not build plate identification. The test is to see if a build plate is installed, not to determine the plate type.

1 Like

hmmm makes sense
meaby your right :stuck_out_tongue:

Probing two more points on the bed shouldn’t be an issue for the MCU :slight_smile:

It’s only detecting of there is a plate on the heating bed by probing the yellow point at the back of the build plate. This section of the plate is outside the heating bed aera. So if there is no plate when it probes at this location there is no force applied to the sensor since it’s trying to probe mid air. This is how the plate detection works. Simple and clever :slight_smile:

Well since there is no camera on the print head to read barcodes or any information, perhaps using the nozzle sensor to measure thickness would be a possibility. It wouldn’t have to be the entire plate just an area where the first probe would be a hole where there was no bed material that would set a baseline. then the nozzle could lift and move to a spot where a designated thickness would be. This would tell the software and printer what bed you have, and then it could set itself appropriately. Then a complete bed level with the new plate settings.
Just a thought
M1

Thanks for sharing your thoughts :slight_smile:

Unfortunately that will only detect the plate thickness not really the type. And this requires that every type of plate must have a different thickness which is not the case at the moment if I don’t mistaken. And their thickness will have to be consistent between all production batches. Which can be a hassle.

Not necessarily, Let’s use this for an example only. The plate thickness is not relevant because by making let’s call it a “calibration square that sits on top of the sheet in the upper region of the build sheet perhaps where the nozzle cleaning takes place” a PEI sheet could have a 1mm thick square that would be probed. Cold sheet could be a 1.5mm square etc. First the hot end probes the bed to determine a sheet 0 reference point, then it would raise up move to the detection block area. This would give you a differential between the two points. Then a software component would calculate based on thickness of the block exactly what type of sheet you have. That in turn would tell the machine this is a PEI sheet etc. Next the machine could do a complete bed level. During that phase it sets the Z offset much like it does now.

The only thing that would be required is that those calibrations blocks were all made to a standard. Because they sit on top of the build sheet the thickness of that sheet is pretty much irreverent. Your telling the machines software what type of sheet based the thickness the probe detects. Would it be perfect? That would yet to be tested but in theory it would work.
Have a great day
M1

1 Like

Ah yes, sorry I haven’t fully understood on your first post.
And it would be more efficient since you only have to probe one point against two in my proposition.

But it’s maybe a bit harder to manufacture. And since the plates are double sided, I’m wondering if that could not be annoying when putting and removing the plate on the bed.

Honestly I don’t know, but I really like the idea, super smart :wink: