Aluminum surface treatment options?
I bought the IR sensor for my Duet in part because it's what best fits the carriage. I've got it installed and running but the mesh data returned seems to be a bit noisy. I didn't read the file data yet, but the graphic suggests the reading might be noisy by as much as ±0.1mm.
I suspect it might be my bed and I'd like to know what options there are.
The bed is an aluminum plate. The IR probe's support web page says anodized matte surface works but really want to avoid anodizing it because I have already stuck a heater to it. I've sanded the top surface in steps down to 500 grit (US standard) with an orbital sander to knock down the shine but I have no idea if it's good enough. The sanding did help a lot though. Before, the readings were very unworkable and looked like a reading could be off by over half a millimeter.
I'm curious if there's something else I can try. If there aren't any suggestions, I might try a surface etchant meant to prepare aluminum for painting. I recall it tends to whiten and dull the surface.
I've been using Wolfbite as an adhesion promoter, and I've forgotten to test the bed before applying the material to see if that's throwing anything off. I will need to remove that before doing any more treatments anyway.
Willy510 last edited by
What surface are you going to print on? Just the bare aluminum or glass or PEI?
Willy510 last edited by
Just run a bunch of G30 S-1 commands where you plan on probing the bed and see if they come back consistently the same or very close from each probe like within 0.01 of each other. Wile I don't print on my aluminum bed bare I did probe it a lot when building my printer and the ir seemed to work fine with it
OK that's a very helpful command to know. My results seem very predictable, it looks like the noise range is less than 0.005mm at any given point at that point, and I move away and go back and it still reads within the same tight range.
Now I'm trying to run tests to see if there's something in the mechanism or about the bed that's causing these variations.
Try probing at several points close to each other (e.g. 1mm apart), to see whether surface scratches etc. are still causing the trigger height to vary.
I did as 5mm and even 1mm pitches tonight. I'm still getting erratic results, with and without the Wolfbite, with machine lights on or off. Scanning a flat piece of paper on the bed shows a nice and flat scan so the probe is behaving well with a different surface.
I expect it is probably some interaction with the probe and aluminum surface. It may well be simply not matte enough in a way that anodizing would fix but that's just not feasible without risking damaging the heater pad.
The only suggestions I have are either to sand it with a superfine grit and see whether that helps sufficiently, or to paint it black and use glass on top.
eric-s last edited by
Would sanded kapton perhaps work? I've had great success with (swirled w/400grit) sticking to kapton with a variety of filament types, and it's my 'go to' bed surface. It's defintately not shiny like bare kapton (or raw aluminium).
I'm considering using the IR sensor for the machine i'm currently building which will have a Mic6 aluminium build plate. Do you think this might work for your situation, since anodizing it seems like it's too risky.
FWIW, The kapton is crazy durable when used this way. I've have close to 100 prints on the first application of it…
number40fan last edited by
I have tried a few different things with the IR sensor and haven't found anything better than glass painted black on the bottom side. There is a test posted on this site for checking Z probes and the IR with painted glass was the only one that I ever had a perfect 30 probe score on. A mildly slow first layer, you can even print on bare glass without any adhesive.
The sensor works with a target of unsanded Kapton tape, and sanded Kapton should work too.
One of my printers uses the IR sensor with a glass bed on top of a black aluminium sheet. I print PLA on it without a coating, i just wipe it with vinegar before printing. As number40 says, painting the underside of the glass black would also work. Another of my printers uses the IR sensor with a bed of blue tape. The third printer used the IR sensor with a PEI sheet painted black on the underside, until I fitted the Smart Effector.
deckingman last edited by
Plain (6mm thick) glass on aluminium painted with black stove paint worked well for me on my CoreXY, as did Kapton tape on my little Mendel. But the IR sensor wasn't too happy when I started using 3DLac on the glass.
DjDemonD last edited by
If you're using plain aluminium why not use electrical contact to probe just connect bed to GND and hotend to SIG on z probe connector. Yes nozzle and bed have to be clean but then the bed has to be clean to get good performance from the ir sensor.
Of course their are other sensor solutions (see below).
vbtalent last edited by
I assume most people don't operate this way because they don't print directly on aluminum is that correct?
This is essentially how my CNC works though.. accuracy is very repeatable.