Topics created by SupraGuy

@SupraGuy said in Duet 3 Mini 5 + 1LC toolboard:

if I'm going to mount the accellerometer to match X and Y axes, the mount will need to be different from what looks like it's intended.

It doesn't matter which way you mount it, as long as the tool board XYZ axes match up with the machine XYZ axes in some order, with signs reversed if necessary. You can configure the mapping between accelerometer and machine XYZ axes. See https://docs.duet3d.com/en/User_manual/Connecting_hardware/Sensors_Accelerometer#orientation.

@SupraGuy Does the fan voltage jumper match the voltage of the fan itself? Other than that, some fans just don't play nicely with pwm, have you tried a different PWM value? I've found that sometimes a really low value like 10Hz can work.

Okay, I've got it sorted...

What was confusing me is that I was trying to use it the G30 P0 ... set the exact same way as I do on my printer that has the DC42 IR probe, which doesn't deploy. It does work the way that I was using it, though I can see from the "proper" use of it that I am doing it a bit weird, by manually accounting for the probe offset.

@droftarts said in BLTouch driving me bonkers...:

usually used in a macro to level a bed with multiple independent motors.

Which is exactly what I wanted to use it for, as I have 3 independent Z motors (On drives 2, 3, and 4)

Took me a while to get back to this, but it's all good now.

@supraguy said in Dead Duet 2 Wifi... Sort of.:

so I'll use an external 5V supply,

This is the way.

@supraguy I believe splitting the axis into Y and U is the only way.

@weed2all True... However if the DWC were fully adequate, I'd have never got started down this path, since what I started with when I installed the Duet on the CNC was a 7" tablet which I had a browser connected to the DWC. Some things with that were annoying me, so I stopmed off upstairs to grab the PanelDue from my 3D printer and plugged it into the CNC instead. (Mostly on the printer I use it to set temperatures before I start to print stuff, MAYBE adjust babystepping.)

I see the reasons why things are how they are in the DWC, but for what I need when I'm standing in front of the machine, it's just enough different. So far, I've programmed a set of macros that do most of the functions that I want, and I can call them from the PanelDue, but the interface is kind of ugly that way.

In short, this is a project that was born of frustration with the existing set of tools, and I'm still learning what things I actually want/need when I'm standing at the machine, so I was planning to adjust it over time.

@jay_s_uk Where homing at the top makes sure that I remain within the limits of the machine, it won't keep be from plunging into the T nuts in the spoilboard. Since there is a variety of lengths that the tools could be. I recognize that this makes sense from the machine point of view, and this is the way that the machine currently on the drawing board is going to operate, but I'd have liked to be able to use the firmware axis limits as a safety/sanity check.

@dc42 I'm not sure that it would help. The problem with swapping bits in a router is that the amount of bit sticking out from the collet is different every time I change it, and in fact the router position within the gantry has a certain amount of variability as well, so if I take the router out of the clamp to clean or service it and put it back in, even the collet is in a different position relative to the machine limits. I'm not sure that I want to spend time chasing the machine limits every time I change a bit or service the router.

I believe that in general use, it's safe to say that I am not running any jobs, certainly not creating any jobs that are pushing the envelope of the motion limits of my machine. While homing the tool to the spoilboard isn't optimal, it provided a way (or so I hoped) to use the axis limitations of the machine to give me a convenient way to do a sanity check on my CAM work.

I will get the same end results (or close enough) to homing hte machine at the top and seting axis limits to the machine motion limits if I simply allow the axis limits of the machine to plunge into the spoilboard approximately as much as the most commonly used tool length, which will (most probably) keep the router collet off of the spoilboard, and depend on reasonable CAM to keep from damaging a bit, or T nut. I can actually safely lift the Z axis well beyond its usable range. The bearings on the Z axis slide will lose contact, but they will just contact again as I lower it.

20210501_043727.jpg

Anyway, this is the first project I completed with the Duet connected to the CNC. It's carved from a piece of 3/4" pine, and houses my PanelDue 5i

with my pause, the first step I do is move z all the way to the top of the z travel to get it out the way and then move the spindle away from the workpiece. On resume, I do the opposite

Nice work. The sense of accomplishment is well worth the cost of the tools.

@SupraGuy great, glad that fixed it! I should probably test what M671 reports when set like this...

Ian

Well, I took the plunge and updated to firmware 3.1.1

A few things were a royal pain, things that I'd tuned that needed finding and re-entering. Like extruder steps that were entirely wrong. There didn't seem to be a spot in the configurator for using 2 motor drives for the Z axis, Pressure advance needed tuning, and it scared the crap outta me when it said that the heat bed was faulting. (PID tuning needed to happen again.) All that stuff I figured out 2 years ago and have since forgotten.

So far, so good.

Wow. It's probably been almost 2 years since my old printer has done anything, but I think I'll have it working as soon as my Amazon order gets here today.

The other order will be tomorrow. A new Duet 2 Wifi with external antenna. I wasn't having reception problems before, so probably not necessary, but it''s what was in stock.

My spare board is gone. So is the Pi that I had programmed to manage many of the PanelDue functions, so I caved and bought the PanelDue controller and a 5" touchscreen. No idea how I'm going to mount it right now, but I'll think of something.

More good to know, but I've also got a 1A fuse on the 12V supply, right at the buck/boost. The supply is supposed to be capable of supplying 5A, but I have the fuse at the supply. I'll add that i have a 15A fuse at the 24V PSU, since there's about 450mm of wire between it and the Duet where the input power fuse is.

I'm a car guy. We believe in fuses at the power source.

@danal said in To retrofit, or not to retrofit...:

@supraguy

Laser, and keep us posted.

Well, you'll have to be patient, since this has been on the "to do" list for a long time, and I'm still gathering ideas/parts. If it's a laser, then I'll probably go with a CoreXY mechanism. one of the problems with a laser though is getting one that can do more than just engrave/mark stuff. I've been following Viktor's (VDX) stuff on the RepRap forums. Most of the inexpensive lasers aren't powerful enough to do any real cutting of anything more substantial than paper. (Though I suppose with enough passes, anything is possible.) The advantage tot he LED lasers is that they're lightweight, and relatively easy to apply PWM control, it can probably be done directly from the Duet. (Haven't looked that closely at it yet.)

Anyway, nothing will be happening that way any time soon, which is one of the reasons why I was considering retrofitting the Duet into the MBot printer.

@tlas if you want to use multiple 1.6A fans then the 1A fan fuse needs to be increased. Most 3D printer Fans are substantiall less than that hence the 1A fuse. Also you need to check what the initial turn on current for those fans are. it may be significantly higher than 1.6A if they are 1.6A in normal operation.

regarding the Duex, yes its safe to increase the fuse as long as you don't go over the maximum ratings of the individual heater channels. Also look at the total input current to the Duex and the wiring between that and the Duet and it all needs to be within spec.

Have you also updated the DWC version? It may only be the 2.02 versions that have fixed the too many files in a folder bug.

Easiest way to resolve without updating would be to move the SD card to a PC and manaully moving the gcode files into folders.

Regarding the dialog that disappears by itself, in the DWC under "Settings" > "User Interface" in the "Notifications" block you will find an option "Automatically close user-generated firmware messages" - just check if perhaps this is enabled; also there is a "Default Notification Timeout" setting as well.

Okay, it's sorted out. Chalk it up to a weird browser issue. Updated browser, and it works as expected.

I needed to go to "machine status, then refresh the page. Set the tool to "active", wait a second, then refresh the page again, then once more set it to "active" for the LED to come on. Once it did, it heated up quickly and accurately.

Now I just need to print some replacement parts, so that I can make the Z axis work...

0_1536296723135_BedMap.png
Okay, that looks a lot better. (In that it's at least mostly a plane, and doesn't look like I'm trying to print on wet corrugated cardboard, not that this is a great result!) I have a little compliance in the front right of the bed when probing, so that corner shows a little lower than it should, the rest pretty much is a sloped plain, slightly lower to the front. I had plans to figure out how I was going to make adjustments, it seems that it's time to put those into action. The map looks terrible in that scale though.

https://www.youtube.com/watch?v=s8WD4v3DnZ8
You can't really see the compliance in the first couple of probes (Weird, I'd have thought that any would be on the other side, since that is further from the sensors there.) Apparently the autofocus hadn't quite figured it out yet.

tghere. This video the axes should look straighter, since I know that I had leveled the screws at this point.