Welcome! Log In Create A New Profile

Advanced

Vagueness of G-Code Reference

Posted by av8r1 
Vagueness of G-Code Reference
March 12, 2017 01:38AM
Converted my problematic delta to Duet WiFi. There's a lot to like, the documentation isn't one of them.

Duet3D's wiki is redundant and difficult to search. The information you need for, say, setting up a Z-probe, might be in the Hardware > Attaching a Z Probe, Auto Calibration, Mesh Bed Leveling, or any other section. There's a Configuring for Delta and Calibrating a Delta. And you still have to read Configuring for Cartesian because humans I guess.

Also, I've been power cycling the printer every time I make a change to config.g or its brethren. Don't know if that's necessary; I can't find any information confirming or denying that. I don't know if it loads that file into memory at startup, if there's a G-Code to reset the board (searching the g-code reference for "reset", "reboot" or "restart" are fruitless) or if it continuously reads from that file and I've been wasting my time. All it says is "saved changes."

There's a fairly useful command, M579. It adjusts the cartesian scale of the model--so you don't have to go smurfing around with diagonal rod lengths or whatever, you can just say "it needs to be 0.2% bigger in Y" or something. You get no feedback when putting in these codes. The wiki (reprap.org or duet3d, they're verbatim copies) doesn't say if that setting will survive a power cycle, or if it needs to be put in config.g, or what. It also doesn't say what "1 to 1" is. What happens if I enter the same M579 code twice back to back? Is it making absolute, or incremental relative corrections?
Re: Vagueness of G-Code Reference
March 12, 2017 03:56AM
Why is it that I can spend ten trillion dollars on a circuit board and the automatic features it's supposed to have are just as worthless as a $12 RAMPS stack? Auto calibration, auto level, they're still more work than twiddling damned bed screws.
Re: Vagueness of G-Code Reference
March 12, 2017 03:56AM
Why is it that I can spend ten trillion dollars on a circuit board and the automatic features it's supposed to have are just as worthless as a $12 RAMPS stack? Auto calibration, auto level, they're still more work than twiddling damned bed screws.
Re: Vagueness of G-Code Reference
March 12, 2017 04:10AM
Anyone want a Duet Wi-Fi? $150.
Re: Vagueness of G-Code Reference
March 12, 2017 04:14AM
Remember kids, if it's more work to get an automatic system to work than doing it manually, it's not automatic.
Re: Vagueness of G-Code Reference
March 12, 2017 06:40AM
It sounds like you're having a bad day. A few comments:

1. We welcome suggestions for organising the wiki documentation better. What we do not want to do is to repeat the same information in several places, because that makes updating it more difficult when things change.

2. On a Cartesian or CoreXY printer it is always best to level the bed using adjusting screws as far as possible. On my 200 x 200mm Cartesian printer, I only use the Z probe for getting an accurate Z=0 height just before starting a print.

3. Bed probing can be really useful in other situations:

- For auto-calibrating a delta printer
- On a Cartesian printer (especially one with a large print area), to compensate for a bed that isn't quite flat (because you didn't use tooling plate) or for sag in the gantry (very common)
- For grid-probing the bed and viewing the height map in the web interface, so that you can see if there are any variations in the nozzle height that indicate mechanical problems (assuming you have an accurate Z probe)

Edited 1 time(s). Last edit at 03/12/2017 06:42AM by dc42.



Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: Vagueness of G-Code Reference
March 12, 2017 05:53PM
My cartesian machine has no Z-probe, a RAMPS stack, and it works just fine by adjusting bed screws. I tried using your IR z-probe and it did run the Z-screws up and down, just...it either wouldn't adjust enough, so it would print diagonal to the bed, or it would print too high. Never could get it to work, and I will never entertain the idea of a Z-probe on a cartesian printer ever again.

Come to think of it, that's the Z-probe I'm using now. Is that my problem, that I'm using an infrared Z-probe?
Re: Vagueness of G-Code Reference
March 12, 2017 06:23PM
For the record, do I have to power cycle the board to make changes to config.g take effect?
Re: Vagueness of G-Code Reference
March 12, 2017 07:02PM
Quote
av8r1
For the record, do I have to power cycle the board to make changes to config.g take effect?

The config.g file is read at startup. So you need to restart the machine to cause changes to config.g to take effect. That's why when you edit config.g in the web interface, it asks you if you want to restart the Duet after saving it. Other ways of restarting it are to power cycle the board, or press the reset button, or press the emergency stop button on the web interface, or send command M999.

If you just want to try out the effect of changing a parameter, you can just send the gcode command to the Duet instead. Then if you decide to keep the change, do the edit to config.g.

Regarding your Z probe issues, what bed surface are you using?

HTH David

Edited 1 time(s). Last edit at 03/12/2017 07:04PM by dc42.



Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: Vagueness of G-Code Reference
March 12, 2017 08:34PM
My web interface does not ask if I want to restart the Duet upon saving changes to config.g. I've upgraded the firmware that came on the board (to 1.17e) but not the web interface, I may need to do that. I figured the Reset button would do the job, but it's perilously close to the Erase button. I am one of Murphy's chosen people, so I dare not risk it. M999 it is.

Copy, G-codes that change settings do not survive a power cycle and must be in config.g to be made "permanent."

Bed surface is clear borosilicate glass b/w red PCB heater, topped with Elmer's glue stick. The probe is attached to the fan shroud using the bolt-on mount suggested on your blog, on the left side of the nozzle. Offset (as far as I can tell since I can't see where the beam is) X0, Y-18. Trigger height as near as I can measure it seems to be 1.6, but I'll recheck this.
Re: Vagueness of G-Code Reference
March 13, 2017 12:33AM
Okay. I've done Z-probe repeatability tests at all my calibration points. They're all within a few .001mm I've put these in as h values in bed.g per the duet3d wiki. It still prints wrong. I'm trying to print a large round disc to get it printing flat across the whole bed surface. If I see it's tight in one place, say near the back tower, I adjust the h value for that probe point, and it messes with the whole bed.

I don't think this is usable. I don't think this actually works.
Re: Vagueness of G-Code Reference
March 13, 2017 02:17AM
These products--Duet WiFi, RepRapFirmware and the Mini IR Sensor, were represented as able to automatically calibrate a Delta printer. I've spent a couple days trying to get it to do so. I feel I have been defrauded. This doesn't actually work. I'm going to contact Filastruder and ask for my money back.
Re: Vagueness of G-Code Reference
March 13, 2017 03:54AM
Quote
av8r1
Bed surface is clear borosilicate glass b/w red PCB heater, topped with Elmer's glue stick. The probe is attached to the fan shroud using the bolt-on mount suggested on your blog, on the left side of the nozzle. Offset (as far as I can tell since I can't see where the beam is) X0, Y-18. Trigger height as near as I can measure it seems to be 1.6, but I'll recheck this.

You may need a sheet of black paper between the PCB bed heater and the glass, to reduce variable amounts of reflection from the bed heater.



Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: Vagueness of G-Code Reference
March 13, 2017 04:58AM
I got it working. Here's what I did to solve the problem: I put the Duet back in its bubble wrap and reinstalled my RAMPS stack. I've emailed Filastruder for a refund, I give 47:1 odds they don't take it back.

If they refuse a refund, I will post it for sale for $150 obo including the IR probe.
Re: Vagueness of G-Code Reference
March 13, 2017 01:00PM
Well, it's a shame you gave up so quickly. I've been using the DuetWiFi for several months now, and I have to say, it's the best board I've used.

Unfortunately, 3d printing as I'm sure you know is not yet at the level where just anyone can buy into it like a paper printer. Especially at the hobbyist level, we spend many hundreds of hours modifying, tuning, tweaking, and working on our printers. You seem to think that you can buy the DuetWifi, slap it on, and have it just work. I hate to break it to you, but that's just not the way it goes. There's a reason why dc42 is constantly pushing out new software versions: because it is still imperfect.

You obviously didn't bother to learn how to use it properly. Many people have already successfully used features like auto bed levelling, so how can you say that it doesn't work? And if it doesn't work for you, then take more than two whole days to figure out why. dc42 has been nothing but accommodating to you, despite how rude you've been. He's under no obligation to continuously update the software. What you payed for was the board, not the software that came with it. The software is entirely free and open source.

We of course can't stop you from returning it or selling it off, but you should keep at it for a while longer to figure out why it isn't working for you, rather than acting like a child, crying when something doesn't go your way immediately.
Re: Vagueness of G-Code Reference
March 13, 2017 01:18PM
A wife phones up her husband:
- Dear, there is this crazy person on the highway, going the wrong way, be careful!
- What?! A single person?! ...Are you kidding?! There are hundreds!!!!!

I am eager to buy off your Duet with IR Sensor for $150 including shipment to Germany smiling smiley.
Re: Vagueness of G-Code Reference
March 13, 2017 09:41PM
@masterjuggler I'm happy for your success and hope you enjoy it very much.

I am aware of the user unfriendliness of 3D printers. I've built seven of them, four of which are in public use at a local maker space. I've spent those hundreds of hours, often with the goal of making the machines accurate and reliable even for those who aren't intimately familiar with how they work.

There has been much swearing.

So how in your mind does following the manufacturer's instructions constitute "not bothering to learn to use it properly?" I'll grant you, my build isn't NASA perfect. If it was, I wouldn't need auto calibration, would I? It's printing well enough in Marlin that I'm not willing to adjust the frame or gantry in any way other than adjusting the end stop screws and bed level screws. Replacing the plastic corners with metal--which would require pretty much complete disassembly--is out of the question. If it can't compensate for what I have, it's no use to me, I'm better off running Marlin.

I input the dimensions of the printer as described on duet3d's wiki, entering the m665 and m666 commands into config.g. I set the steps/mm and driver currents for all four motors, the thermistor settings, the z-probe offset. I used DC42's bed.g file generator to create a probe pattern that fits on my build plate. I checked Z-probe repeatability and set the trigger height corrections in bed.g for each of the probe points. Running G32 at the beginning of a print is no help, neither is G29, or both. By the way, if you find you have to put in trigger height corrections in bed.g for autocalibration, do those corrections also apply to the G29 mesh leveling? They don't run the same pattern. So if you do a G28, G32 and then G29, are you probing with corrections, then probing without corrections?

I'm at the point where I'd have to start throwing parts at the problem to either make my bed better suited to an IR probe, to get a different probe, or to dick with the frame. None of these things are easier than going back to Marlin. I have no obligation to you or anyone else to spend any more of my time trying to convince it to print a flat first layer.

@Teilchen; Filastruder won't take it back, PM me.
Re: Vagueness of G-Code Reference
March 13, 2017 10:15PM
Be that as it may, it still doesn't justify berating someone who was legitimately trying to work with you, which is the thing that is pushing my buttons. I checked out your other posts, and I'm not questioning your experience or intelligence. If you tried the board and it doesn't work for your situation, then that's totally understandable. What's happening here though, whether you mean for it to come across this way or not, is straight up childish whining.

Edited 1 time(s). Last edit at 03/13/2017 10:17PM by Masterjuggler.
Re: Vagueness of G-Code Reference
March 13, 2017 11:01PM
I'll grant you that. I'll be the first to admit that I am a festering asshole, one with nearly no patience. Some of it's learned behavior, some of it is lack of mental health, most of it is just plain old being a bad person. I know it and I'll own it.
Sorry, only registered users may post in this forum.

Click here to login