Topics created by vistalert

@o_lampe said in Requirements to have a 1000x1000 aluminium build plate:

@phaedrux said in Requirements to have a 1000x1000 aluminium build plate:

At 1sq meter maybe plywood and a coating of self-leveling cement? Embed a floor heating coil in the cement?

I like the idea! How flat would that cement be?

So long as you use the right 'cement', good enough for NASA.

Seriously, self-leveling compounds are used to make large reference surfaces and can be flat to unreasonable numbers of decimal places.

There's a ton of information out there. Look at polymer concrete and epoxy granite. Both are cheap and if done for a self-leveling finish, can be very very flat.
Just needs to be the right material and finished in the right way (materials that shrink are not going to produce a level surface).

SOLVED: Issuing M18 to turn off motors results in probe responding to deploy/retract.

Recap of issue:

Heat nozzle Extrude BLTOUCH probe does not respond to deploy retract

Suspect cause of issue is incorrectly formed stop, sleep, cancel scripts.

Perhaps these scripts should contain M18.

@Phaedrux many thanks for your assistance and input.

FOLLOW-ON Note in case it helps anyone else. This is my end of job script:

G1 H2 Z460 F3000 ; Drop the bed
G1 H2 X1 Y1 F8000 ; Move head near home position
M18 ; Motors off
M0 ; STOP which turns off heaters

I'm not sure why, but whilst there were instructions in stop.g the bed heater was shutdown but not the nozzle. So I emptied stop.g and in my shutdown.g the above results in the bed being parked, all heaters off and motors turned off, which finally results in the probe responding again.

@vistalert said in Turn one of the three Z screws (steppers):

as opposed to relying on smarts to compensate for what could be something that's misaligned

If you can't make it perfect, make it adjustable. If you can't make it the same every time by hand, make it automated.

@vistalert
First off, a quick description of what a tool is, because that's what you'll have - 4 tools. Essentially a tool consists of an extruder, a heater and a temperature sensor. There is also usually a heat sink cooling fan but this is associated with a heater/temperature sensor so it's not strictly part of a tool. Optionally there will be a part cooling fan and you'll likely have one per tool.

So you need to connect 4 heaters, 4 temperature sensors, 4 hot end cooling fans, 4 extruder motors and optionally up to 4 part cooling fans (you could maybe use one fan per pair of hot ends (tools)). There are also things like end stop switches and/or Z probes to think about.

You've got a couple of options - either the expansion board or what might be a better option would be to use tool boards for each extruder. It all depends on where you plan to mount the extruders. If you use a single expansion board, then you'll need to connect one tool (heater, temp sensor, heat sink fan and extruder motor) to the main board and the other 3 tools to the expansion board. So all the wiring will go to either of two central points (the main board or the expansion board).

What might work better would be to use tool boards instead of the expansion board. That would allow you to mount each tool board close to each extruder for example. Of course you'll still need to run wires from each tool board to each heater/temp sensor/fan etc. Alternatively, you could mount tool boards on the carriage. This would make wiring the heaters, sensor,fans etc a lot easier and neater but of course, you then have longer runs of cable from the tool boards to the extruders motors. Having tool boards on the carriage might also simplify connecting a Z probe but check the list of firmware limitation here https://duet3d.dozuki.com/Wiki/Duet_3_firmware_configuration_limitations

Actually, there is a third option which is to use Duet 2. So you'd need the Duet Ethernet or Wi-Fi plus the Duex5. But in this case, all the wiring has to go back to a central point because the Duex5 has to be very close to the main board.

Edit. I've only recently started to use YouTube more. There is a lot more stuff in my blog if you haven't already seen that (link n my signature).

Thanks every one - dropping the Zoffset slightly, in config.g has brought the nozzle spot on to 1mm from the bed with Z=1.

Brilliant, thank you, it works 🙂

@Dizzwold

Remember that magnetic force decreases with the (IIRC) square of the distance, so the blind holes are cool, but decrease the force a little.
The beds I built I just used through holes and superglue, 305x305mm and 25 magnets of similar size to the ones you are thinking of using (the ones by the edge are set back ~12mm). The critical thing for any glue is getting the surfaces really clean.

The only issues were if the glue failed, but those ( a couple IIRC pulled out) I just scraped out the old glue and re-glued.

The beds are now several years old and still work just fine.

Well one thing I hate about forums is when you see these big problems discussed, and advice and solutions offered, but then there's an abrupt stop in the conversation, and you never know what happened.

Several weeks ago, as you may have read here, I was really struggling with my supplied build plate, as it's 500x500 and only 4mm thick. Also, I struggled to find anybody in Western Australia, that could supply an aluminium plate that was guaranteed to be "flat", and when I enquired in Sydney, there were limited options for tool plate.

Long story short, thanks to the collective experience of everybody in this forum, I learned how I should go out and source a new plate, or at least, work with the constrains arising due to my location, and get a plate produced.

I found an engineering company here in Perth that could start with a 10mm plate, and machine it. Their advice was it would have to be machined on both sides, because of surface tension. Turns out, to get the plate "flat" it had to be machined 5 times. The resulting thickness was 7.7mm, which thanks to the advice above I know would mean there is minimum sag, due to own weight.

So I got the plate installed, and added my 6mm glass plate and finally got by bed mesh done.

I think it's safe to say the result is amazing.

e3e458e2-303e-4e46-95c1-163be1ba4301-image.png

049279f3-497a-4777-b54c-a3c5107d1657-image.png

I have a few more things todo before I can actually print, but I have my "shutdown" and "startup sequence", which puts me in a great place with consistent bed mesh results.

I have mechanical endstops for X and Y, so my homing here is pretty accurate I use stalling on my Z axis, and send the plate to max-z until it touches the couplings. And I call this my "parked" position" I now have a macros that raises the bed 480mm from the parked position, then I home all. So X then Y then Z all home. When I execute G29 now, I get an amazing height map.

Now I have to say the bed has been an investment, but clearly a worthwhile one. The plate cost me $280 AUD. The glass was about $25.

I've been cautioned about my bolt mountings, where the plate rests on the springs/bolts. The concern being, will the aluminium expand when heated (of course it will, by about 0.7mm), so the question comes will the plate be able to expand with these bolts in place without being completely constrained, therefore warping. Some thoughts on this would be useful.

Largely though, my issues seem to all be resolved.

As per the Duet documentation - one way to test the probe is to deploy it and press it momentarily. Remember to release it quickly. You will see the Z Sensor in the DWC interface change to indicate the probe has sensed contact.

bb35080d-eb57-4bf7-9810-44ca76b2bef4-image.png