@baesjerker said in Ender 3 Pro configuration:

My Ender 3 pro board failed on me for some strange reason. I think the problem was that the extruder motor driver died. 2 of the 4 pins went straight to ground.

So now I have a Duet board. I have no problem understanding electronics and so forth, but G-Codes and such is way beyond my knownledge.

So far I have managed to home x and y without any problems. Both hit the endstops and work as expected.

Z is another story. Z has no endstop, but use a BLTouch. It works but I have to stop it with my finger because it tries to probe outside of the bed.

I can manually change it so it probes on the bed via homez and homeall, but when I press the mesh bed compensation it will probe outside of the bed again.

Do anyone have a working configuration for Ender 3 Pro + Duet + BLTouch? That would help a lot.

I am actually working on an Ender 3 for a friend of mine. I cannot imagine the differences in machine configuration being drastically different. I would suggest using the Z end stop on the top, that is what I am doing for now and up until this morning when I accidentally corrupted the sd card by trying to write to the bedmesh.g file.....while it was actively reading the bedmesh.g file, it was working extremely well and it is extremely precise.

Much of your results will depend on (1) How well you assembled it (2) how precisely you assembled it (3) how good you are at troubleshooting, calibrating, and tuning it.

My configuration will probably be about the best out of anyone using an Ender because I am a machine builder, I am a CNC machinist, I am a Perfectionist, and I have access to the measurement equipment to calibrate it extremely precisely.

Getting dimensional accuracy with a low-end 3d printer is not within most people's realm. In general, it is not easy. Luckily, Creality did a phenomenal job on the design, relatively speaking. It is, however, time-consuming. I calibrate extrusion with a micrometer and I check perpendicularity with an indicator that measures in ten-thousandths of an inch. I assemble it on granite because of the way granite countertops are manufactured (they are CNC machined, then precision "lapped" which is like wet sanding but more similar to grinding and polishing to get a shiny finish). Granite is VERY precise, the flatness is amazing, we use higher precision granite at work for inspection tables. If you don't have one already, get a 123 block (machined, then ground) from amazon or something and a Fowler, SPI, or Mitutoyo caliper. Amazon or MSC direct will do.

Assuming your machine is similar in the way it is assembled (your Z height will likely vary due to the change in the Y extrusion) then you could pretty much use my config and you'll only need to do a could things like calibrate your z endstop height from the bed, calibrate your bltouch trigger distance, and maybe your steps per mm may be off a tiny amount if the deviation in the steppers is drastic, but I doubt it'll be all that much.

The duet hardware alone just STOMPS the sh*tty ender board, there is literally no comparison between the two. The firmware....good god, you have no idea how good it is, how far ahead it truly is. Switching to the Duet is as good of a decision as buying the ender pro, but better because once you get to know the duet and RRF....you'll begin to understand just how much you can do with the boards and how much development actually went into and is still going into the duet ecosystem.

Since I corrupted my SD card this morning (luckily I had a backup from the night before) I lost a lot of the final config numbers, like the numbers you are asking for to avoid it going off the bed.

So you aren't super frustrated doing it yourself until I am finished, this should solve some of your problems:

G31 P500 X-41.5 Y-10.265 Z0.9                                ; Set Z probe trigger value, offset and trigger height

This is the offset value in the X and the Y (both negative, as cnc offsets are ALWAYS relative to the spindle/tool head)
This is assuming you have the same BLTouch mount that I have, I can only go by what I have in front of me.

Where this file is located depends on your configuration. If it is a standard config file from the configurator tool, then the G31 value will be by the rest of your Z probe stuff about 3/4 of the way down.

I have a separate "bedmesh.g" file and call up bed.g file like this:

; bed.g
; called to perform automatic bed compensation via G32
;
; generated by RepRapFirmware Configuration Tool on Tue Jun 26 2018 13:03:45 GMT-0400 (EDT)

M561 ; clear any bed transform

M98 Pdeployprobe.g  ; deploy mechanical Z probe


M98 Pmachine_bedmesh.g ; configure the bed mesh

G29                 ; probe the bed and enable compensation

; Probe the bed at 4 points
G30 P0 X10Y30 H0 Z-99999
G30 P1 X10 Y220 H0 Z-99999
G30 P2 X175 Y220 H0 Z-99999
G30 P3 X175 Y30 H0 Z-99999 S4

M98 Pretractprobe.g ; retract mechanical Z probe

My bedmesh file just looks like this:

; machine_bedmesh.g
; February 18, 2019

; This file defines the extent and density of the bed mesh.
; The way this is calculated is by taking the max X, subtract the Z probe X offset
; from it and use that as the second of the X values.
; For example, if max X is 383 and the Z probe X offset is -43, the parameter would be X0:340
; For the Y parameter, make sure you don't start too far to the front or you will hit
; the bed clips with the hot end.  Anywhere between 35 and 40 is probably a good start.
; The S parameter defines the distance between two mesh points in mm.

M557 X0:175 Y20:220 S40      					   ; Define mesh grid
M376 H15 ; Define height(mm) over which to taper off heightmap compensation

This should get you past the issue you are having now, is you are still going off the bed after this, bump up the X value in the M557 command found in my bedmesh file, or measure the offset in the X and Y value for your BLTouch mount and alter it in the G31 command, obviously left of the nozzle is negative, right is positive as it is relative to the spindle/tool head.

I'll upload the rest to my github when I am finished.

You should have a topic/sub-category specifically for the wiki.

When I go to this page: https://duet3d.dozuki.com/#Section_Compatible_add_on_Hardware

Which obviously is accessed from the "table of contents" on the homepage of the wiki/documentation.

I do not see links for the Duex5 or the Duet Expansion Breakout Board, like the pinouts that I am looking for now that just received them in the mail.

I really should see them there, the smart effector is there....but the expansion boards aren't?

Could these links be added? I can't edit and add to the wiki, can I? I know the wiki still relatively young, I'd like to add documentation for what I'll be doing in the upcoming months

Too bad the duet forum can't be merged with like a Discord server, that would be pretty cool and pretty convenient with how you can like tag or pin people or everyone in that particular thread, I'd be way more active on a Discord server, just saying, phone mobile notifications for the discord app are kinda nice, I have it notify my only when someone specifically tags me.

@dc42

Not sure if you're interested in this at all, but I have have access to a few industrial machines, I found a way to get the list of the supported G-Codes and M-Codes in Fanuc's controllers. I have access to a 2014 Doosan Puma GT2100 CNC lathe and idk what year Doosan DNM 5700 mill and a very large and rather new Haas GR712 and two ProtoTrak controllers.

Perhaps I'm unfamiliar with the specific intent of the Duet 3.0, but I vaguely remember a conversation with you about standardizing the G-Codes at least, as M-Codes tend to be manufacturer and/or machine application specific. I'll get lists/pictures of all the G-Codes and M-Codes if you want and if there's any other information you want/need from the Fanuc controllers (kinda wanna reverse engineer their controller myself cuz they are the standard for industrial CNC controllers). It's for several networking and interface ports, can't imagine it would be that hard to get code from the controllers:)

Anything you wanna know about some of the industrial controllers we have in our shop??

Oh, and this is something cool I discovered is Macro B Programming on Fanuc controllers. Click here for a PDF on my Google drive about Macro B Programming inside Fanuc controllers

It allows parametric hand programming for canned cycles like G71 on a CNC lathe which is a roughing operation.

link to pictures of my program I created in a couple minutes

@baesjerker

You are welcome! I know what it is like to switch to a superior....thing...whatever that may be, software, hardware, programming languages, manufacturing techniques, ect. Only to be limited by the genuine lack of exposure to it and the first couple hours, days, weeks, or months are just deeply disappointing, discouraging, frustrating, and you end up questioning if it was the right decision if it is bad enough.

That is not what people need for additive manufacturing, better OSH, better FOSS, or just in regard to learning in general.

I am sure if you are less busy than I am the next few days, you might get it printing fairly quickly (it is not hard once it homes correctly)

You can try to calibrate it yourself to learn some of the code and whatnot. if you do, keep this code in the back of your mind when you are calibrating extrusion and when you are printing test cubes to verify dimensional accuracy (don't forget there is a big difference between high-resolution and high-precision/accuracy).

https://duet3d.dozuki.com/Wiki/Gcode#Section_M579_Scale_Cartesian_axes

That along with

https://duet3d.dozuki.com/Wiki/Gcode#Section_M572_Set_or_report_extruder_pressure_advance

These are two vital codes that make tuning SO MUCH EASIER!

This may very well change but as of right now I have M572 set to

M572 D0 S0.074										; pressure advance

in the bottom of my config.g a few lines above my M501 command.

You want a 123 block is because it is a precise way to set the nozzle height without touching the bed (I do not heat the nozzle, but i clean it thoroughly before I calibrate so I know I am touching metal, not plastic).

No damage to the bed and the calibration is perfect. Just make sure you swap out the springs for solid bed mounts, I got mine at Ace and I would have to get a part number to tell you exactly what it is, but now my bed remains the same height. In the CNC realm, consistency is what you need to achieve repeatable precision.

https://youtu.be/nIjM3FZahjE?t=115

I just use 25.4 mm for my z height because 1.0000" is exactly equal to 25.4 mm, no trailing numbers, no nothing.

If you do this though, add a mm or two or three to either the first layer or subtract a couple from your overall Z height (distance from high-end Z endstop) because it is too close to the bed print normally!

If you do this, incrementally lower the nozzle slowly because that will be 1.0000" to touching the bed, not the few thou high to allow the plastic to come out of the nozzle. If you forget to add a few thousandths of an inch to this, you will damage your bed.

I will keep this post updated but I will be a bit busy over the next three days with the ABB Yumi getting delivered at work tomorrow morning.

I didn't mean to say lower the overall power consuption by using higher voltage, I meant lower the amperage in the circuit, sometimes I hear a whine in my motors which drives me nuts, and I have heard that running the motors and heaters at higher voltages is just overall better by cutting down in heating times, thinner gauge wire, a smaller form factor.

I know a little (next to nothing, but I'm learning) about 3 phase power. That is why I said 3 phase power (which is what our Haas CNC machines operate off of) and a 3 phase servo motor.

What confuses me is that I was under the impression that a multi-phase servo motor technically operates off of alternating current (pulses) being delivered through varying voltages or at full voltage in short bursts through PID tuned PWM tuned by rotational error feedback which monitored the back emf.

Which I find confusing because how is that not a sensor? How is it a sensorless motor when it is using Hall Effect [Sensors] to measure. Google defines a sensor as "a device that detects or measures a physical property and records, indicates, or otherwise responds to it."

Is a lot of this counterintuitive or is it just me?

If what you are telling me is true, then this is what I need to be reading: https://www.digikey.com/en/articles/techzone/2013/jun/controlling-sensorless-bldc-motors-via-back-emf

Right?
We also have a few of these:

http://www.micromo.com/4490h048bs.html

So what about the spindle motor in a CNC router where the force being applied across the shaft is not actually across the shaft, it is part of a drive system that…still required torque...hmm a $700 computer fan, a $700 motor in a Dremel, $2,800 in motors for a quad copter? I mean...this seems over-engineered for anything I am capable of, I'd like to be able to put some of these motors to good use.

So small robotic arms, or the end servos on the end of the robotic arm?

We have so many high torque, or high rpm two wire motors...it's a little ridiculous. I am learning a lot from trying to learn how to use them, schools near me are theoretical education. I need to learn by application or assosiation, which is why I am struggling with this.

I must sound like an ignorant child in comparison. I spend so much time reading and I still know nothing. I have indirect and at times direct access to the German Faulhaber Motor company, direct access to a Haas VF2 CNC Mill that I can personally get parts to within 0.0002 of an inch from tolerance and a 3 year educational license to Autodesk Inventor HSM Ultimate. Is it too much to ask for a custom built CNC machine? Haha

@baesjerker Make sure filament is Dry, filament is extremely hygroscopic, it absorbs moisture from the air EASILY. Best to make a dry box for it, I use I think tupperware containers, large ones (10 or 20 gallon) They are air tight, then I throw an Evadry (Amazon or Costco I think also sells them) reusable humidity absorber in there will my filament. If there is any moisture in your filament, due to the temperature of the nozzle during extrusion, the water instantly boils and creates bubbles, blobs, air gaps and all sorts of seemingly impossible and inconsistent problems that are a friggin nightmare to try to diagnose/troubleshoot. As a direct result of that, some of the things you could be changing could be bringing you further away from dimensional precision and worsen your print quality. A word of advise: Keep a ledger, document EVERY change, no matter how small. If you do this, reverting to old settings is effortless and there is no possible room for error. I make this easier by simply commenting out code and adding lines, instead of changing values. Just....don't forget to add comments so YOU know which settings are for what.

Assuming your filament has been properly dried (24 hours at the very least) then recalibrate your extruder. Don't even attempt this without a quality caliper, at the very least.

What are your Travel Speeds, your non-printing speeds, set to?

What is your print temperature?

DC42 is right, retraction settings are important because pla is known to ooze and you must relieve the back pressure on the nozzle and get it away from the heat of the nozzle enough for it not to come out while moving to the next print coordinate.

Temperature is equally important, however, if the temperature is too high, the filament is more of a liquid and retraction won't do a damn thing. Travel speed (non-printing moves) should also be as high as you can without creating skipping or overtravel because the the faster the nozzle gets to the target coordinate, the less time the nozzle has to ooze. Change temperature first, then travel (non-printing) speed. I usually go overboard and go with like 300-800 but make sure it's north of 250. THEN retraction speed, then retraction distance. You want retraction speed somewhere around 30-60 depending on how dialed in you are (usually), if I'm wrong about this, someone correct me.

You want retraction distance as small as possible, but with bowden I've found 4-10mm the typical range (but depends entirely how long the tube is, or even the type of filament).

Do you have

https://duet3d.dozuki.com/Wiki/Gcode#Section_M572_Set_or_report_extruder_pressure_advance

In your config? I don't recall what mine is set to, there is a thread here about an in-depth analysis of pressure advance.

What filament are you using out of curiosity? What brand?

Everything comes into play and every variable must be flawless in order to get the results you're looking for. Start eliminating variables. carefully and methodically. It's time-consuming, yes, but we're working on automating additive manufacturing, standardizing 3D printing.

Received my Duet3 today......looking at this up close for the first time, as couple thoughts come to mind.

(1) it's bigger than I expected. I actually prefer that it is a bit bigger because it gives the impression of more of a "motherboard" or "mainboard" which is what I wanted out of the Duet2 initially.

(2) there are two different type of headers on this, the motor headers are significantly "beefier"
(Kinda wish they were the same like tan-ish color though.....first world problems )

(3) lotta components on this board....getting me excited and I haven't even plugged it in yet!!!

(4) there's two RJ45 ports on it?? Hmmm, okay, cool.

(5) kinda wish someone would hit these with a fine file, knife, or like red or blue scotchbright before they got sent out, I have it here, I just wish I didn't have to, again, first world problems.

Excited to use it!!!

Thank you!

@t3p3tony Hey, so how close are you guys to that Q2 2019 goal for early production boards? Can you take my money for it yet?

Was it an RJ11 port??? Whoops! Clearly I didn't look at it particularly closely

My bad!!!

@bearer said in Duet 3 Mainboard 6HC - initial production run.:

@Dougal1957 said in Duet 3 Mainboard 6HC - initial production run.:

Mine certainly had an image on else I would have been lost as I know very little about the way Linux interacts (I know a lot more now since I got this thing on Saturday)

Doug

Did the Duet 3 just get classified as a "gateway drug"?

It should have if it didn't

Received my Duet3 today......looking at this up close for the first time, as couple thoughts come to mind.

(1) it's bigger than I expected. I actually prefer that it is a bit bigger because it gives the impression of more of a "motherboard" or "mainboard" which is what I wanted out of the Duet2 initially.

(2) there are two different type of headers on this, the motor headers are significantly "beefier"
(Kinda wish they were the same like tan-ish color though.....first world problems )

(3) lotta components on this board....getting me excited and I haven't even plugged it in yet!!!

(4) there's two RJ45 ports on it?? Hmmm, okay, cool.

(5) kinda wish someone would hit these with a fine file, knife, or like red or blue scotchbright before they got sent out, I have it here, I just wish I didn't have to, again, first world problems.

Excited to use it!!!

Thank you!

@eumldeuml I work at a Robotics Integrator in Sarasota, FL that has a fully equipped machine shop, I'm actually designing and building a CoreXY, Kossel/Delta, and a...Slightly different machine as well. Literally every component apart from (1) controller (2) power supply and (3) motors (and filament and tubing) will be machined in-house, with linear rails and ball screws or belts from suppliers through work. Gates for belt and either Hiwin, THK, SKF linear rails and ball screw depends on the cost, as those will be expensive.

I'm spending WAY too much on this. I have dumped....idk I've lost count at this point, $5,000 maybe? I'm over $3,700 for sure in 3D printing and I'm kinda just hitting the "F*ck it!" Button with this build. I'm trying to shoot for <$3,500 for everything...but not necessarily keeping track as of what I'm spending so far. Everything will either be machined to within a few tenths or precision ground.

I will post pictures when I start actually manufacturing it, design isn't finished yet.

@grizewald Hmmm, I didn't know that about the comments and content. I also was not considering the retrieval of information or the storing of it, I was thinking purely communication specific. I use Swype for my phone and there is no way I can type faster on a keyboard, no way. A phone, specifically an android one, also utilizes Google's Talk-To-Text which you can more or less speak normally and it'll beat most far typers with minimal errors in most dialects, with exceptions of course. Just a quick thought, I don't care to go into chrome and logon each time and go to the website specifically to go into the forum. Their own app (built with flutter) could also be an alternative. ¯_(ツ)_/¯

Too bad the duet forum can't be merged with like a Discord server, that would be pretty cool and pretty convenient with how you can like tag or pin people or everyone in that particular thread, I'd be way more active on a Discord server, just saying, phone mobile notifications for the discord app are kinda nice, I have it notify my only when someone specifically tags me.

@dc42

Not sure if you're interested in this at all, but I have have access to a few industrial machines, I found a way to get the list of the supported G-Codes and M-Codes in Fanuc's controllers. I have access to a 2014 Doosan Puma GT2100 CNC lathe and idk what year Doosan DNM 5700 mill and a very large and rather new Haas GR712 and two ProtoTrak controllers.

Perhaps I'm unfamiliar with the specific intent of the Duet 3.0, but I vaguely remember a conversation with you about standardizing the G-Codes at least, as M-Codes tend to be manufacturer and/or machine application specific. I'll get lists/pictures of all the G-Codes and M-Codes if you want and if there's any other information you want/need from the Fanuc controllers (kinda wanna reverse engineer their controller myself cuz they are the standard for industrial CNC controllers). It's for several networking and interface ports, can't imagine it would be that hard to get code from the controllers:)

Anything you wanna know about some of the industrial controllers we have in our shop??

Oh, and this is something cool I discovered is Macro B Programming on Fanuc controllers. Click here for a PDF on my Google drive about Macro B Programming inside Fanuc controllers

It allows parametric hand programming for canned cycles like G71 on a CNC lathe which is a roughing operation.

link to pictures of my program I created in a couple minutes

@t3p3tony said in News on Duet 3 release?:

Tim beat me to it!

https://forum.duet3d.com/topic/11540/duet-3-mainboard-6hc-initial-production-run

Hmmm that's a lot of information to digest....ima have some questions in that thread soon.

@elmoret said in News on Duet 3 release?:

@noskillzengineer said in News on Duet 3 release?:

I want one of the early prototypes/early releases.

Do you have an estimate as when we might see them? I am extremely interested.

https://www.duet3d.com/Duet3Mainboard6HC

Thank you so much! I was unaware preorders started! I need this in my life

@bearer said in News on Duet 3 release?:

There really isn't any information on that wall of text to suggest which of the Duet boards would be the best fit for your needs. Maybe if you describe what size and number of motors you need, and other IO it'll be easier for the Duet guys to help you evaluate which board to use.

To me its not entirely clear if you want to make 3d printer kits, cnc machine kits or convert your existing cnc machines to be controlled by a Duet.

The number of motors will vary as I want to do all 3 as well as design and create new machine tools. The motors, or rather axis count, will vary from 3 to probably about 20, as of right now. There is a wall of text because I am not like any other company in this space, I want to elevate hobbyist and small business machines on an unimaginable scale. I see great potential in this project and I think it's time the industrial realm caught up with the 21st century.

Most that know me already know where my priorities lie, the Duet2 is a powerful mainboard, if I can do what I suspect I can with the Duet3, the next iteration of the Duet3 (or a customized variant of it) will come very shortly after, even if I have to employ my own Hardware and Firmware engineers to do it (nothing personal to the team here, it would likely just be a more premium option that could more competitively compete in the industrial space, if I run into any limitations, so the overall cost to entry for the Duet3 remains competitive, then if price is no concern there's also an option for that crowd).

I don't necessarily need someone to tell me what is most appropriate for my applications, I am familiar with the Duet and RRF (relatively speaking). I just want to start playing with the Duet3 already

@dc42 said in News on Duet 3 release?:

We're planning a pre-production run in August. The price will be higher than the Duet2 and you will also need a Raspberry Pi 3B+ or similar to manage the communications, storage and user interface. But you will get 6 stepper drivers capable of handling somewhere between 4A and 5A peak current, 10 heater/fan drivers, 9 GPIO channels for endstops, filament monitors, Z probes etc., plus CAN bus expansion if you need more of anything.

You also talk about OEM support. I'm filing the paperwork for my own company this week or next, much of my plan was to either build a high-quality kit and sell just the kit and have a config for a duet or sell with one in it. Each and every component will be CNC machined, precision ground, or precision lapped so that the hardware will never be a problem for this community again, as I really have only been hardware limited as far as precision goes. Typically, CNC machined brackets, components, or kits are unnecessarily expensive and from my perspective...I can't really understand why. More people need access to this technology and the less people are fiddling with calibration and hardware, the more innovation and entrepreneurship will thrive. That's why I'll do it for nearly what it will cost me to make it.

I run industrial CNC equipment and play with ABB and Kawasaki robots at work and the duet is about the closest I've seen to the industrial Fanuc controllers that control....well, most CNC machines. I would bet it wouldn't take all that much to replace one of the controllers for one of these machines (as I look at the CNC mills and CNC lathes) with a Duet and have it work...pretty darn seamlessly, with the exception of the whole AC motor DC stepper thing. There's probably a bit more inputs and outputs on these machines, maybe some beefier components, the principle really isn't all that different though...the hardware is.

The big difference that I see is AC motors, the SIGNIFICANTLY higher power requirements (this Doosan CNC Lathe says up to 76 Amps @ 220v on the back of the machine), many more pump connections (oil pump, low pressure and high pressure coolant pump, something for pneumatics actuation for turret movement for tool changes, oil scavenging in the coolant, etc), the HMI (the Human to Machine Interface, or the physical controller interface), the massive footprint, castings, and motors....and some convenient safety features like an enclosure.

I want a duet to run the majority of my machines, I'll buy a few new machines likely with Fanuc controllers (because I know it so well) and make all the machines I'll design with those to a few tenths or less. I'd prefer to have a duet run the machines I design and sell.

This industry would not be where it is today, nor would we be on this forum today, if it wasn't for Stratasys's patents expiring and the Open-Source philosophy as well as its community. Despite many business owners advising me to maintain proprietary intellectual property, all of what my business creates will be open-source, perhaps with a slight exception or two if necessary to stay in business.

I want one of the early prototypes/early releases.

Do you have an estimate as when we might see them? I am extremely interested.