Looking for new contol boards and firmware

Not sure where best to post this, so I've settled on the "General" section.

In a nutshell, it's time to move on and away from Duet products and RepRap firmware. It's a long story and one that I don't want to go into here. But I've only ever used Duet Hardware and RepRap firmware since getting my first printer which was a RepRapPro Mendel (tricolor) and which used the very first Duet 06 boards. then the 085, then Gen 2 Wifi and Ethernet and now Gen 3. So I'm more than a little out of touch with the alternatives.

What are you guys using now? What's best? I tend to build complex machines.

My ideal choice would be to find a vendor or vendors that I could work together with. That's how it used to be with Duet. I would come up with ideas, we'd bounce it around, they'd do the firmware and together we'd debug it and get it working for the benefit of all. But if that's not possible, I'll settle for a fixed spec solution.

So over to you guys - what are you using (hardware and firmware)? What do like about it? What don't you like about it?

Cheers

Edited 1 time(s). Last edit at 03/01/2021 03:19PM by deckingman.

---

[somei3deas.wordpress.com]
[www.youtube.com]

I love SKR Mini and Marlin for low cost, but the Marlin source code is a nightmare to work in, and I think it will only get worse over time. Smoothieware code looks a lot nicer, but I haven't actually used it before. I've heard good things about Klipper too, which uses an interesting architecture where the main program runs on a Raspberry Pi, and the regular 3D printer motherboard (Arduino+RAMPS, SKR Mini, Duet, etc.) only relays commands to the stepper drivers. Might be good for your 6 extruder hotend since it can control multiple motherboards at once.

@deckingman, I did not think I had anything to add to this discussion as I have kept to 8-bit processors with additional functions such as print head swapping being controlled by slave processors. This architecture may not be to everybody's taste, but it suits me as my high-level language skills really suck although I can do useful work in assembly language and have some hardware skills.

I had not followed developments in Klipper over the last few years but was very intrigued when I followed up dekutree64's suggestion. I had incorrectly assumed that Klipper was another casualty of the great apathy which has struck down much of the RepRap enterprise.

I know now what my reading matter and tinkering will be for the next few days and I am sure I can find a spare Raspberry Pi as I have quite a few in my project boxes. The distributed approach is particularly attractive as it allows failed and obsolescent parts to be replaced and to add in additional functionality which may never have been thought of by the designer of a monolithic controller.

Mike

Hi Mike,

Yes, Klipper has been mentioned to me a few times over the years. I have to say that I hadn't really taken much notice until recently.

I particularly like the distributed approach - it's useful on a large format machine with many motors. In that respect, Duet Gen 3 would appear to be a good idea with it's CANbus expansion boards. But I don't like the fact that if one wants (say) 3 stepper drivers, you have to have something like 18 additional IO connectors of one sort or another, which are largely redundant (especially in my case as I have 3 of those boards). It means that expansion boards are much larger than they need be, which is problematic when one fits them to a moving gantry. Also, I have this gut feeling that many of the problems with Duet Gen 3 are somehow related to things getting out of sync or losing data between boards. I have nothing to substantiate that gut feeling other than poor and inconsistent print quality issues.

I was pleasantly surprised to see the quoted attainable step pulse frequency rates with Klipper. I didn't know it because nobody would tell me, but for around 18 months Duet Gen 3 expansion boards were limited to around 12kHz. Just about enough drive a 400 step per mm (using 16x micro stepping) extruder at 1800mm/min (30mm/sec) retraction speed. This has recently been resolved and it's now around 130 kHz which is on a par with Duet Gen 2, and means that I can now run my extruders at 128x micro stepping, as is required when using a mixing hot end and where one or more extruders might be feeding single digit percentage of the whole extrusion amount. The problem being that with high micro-stepping, if the maximum step pulse frequency is too low, one cannot then retract the filament at a reasonable speed. So I was somewhat amazed to see that Klipper could attain 700k with 3 steppers active (depending on hardware of course).

Lots to think about........... I guess the big question is, can a 67 year old climb the learning curve?

Ian

Edited 1 time(s). Last edit at 03/02/2021 04:13PM by deckingman.

---

[somei3deas.wordpress.com]
[www.youtube.com]

Just a note that I have been playing with Klipper and so far it is giving me few problems other than my own hamfistedness - like not putting my thermistor definition in the Printer.cfg file before I called it for example. It would be nice to try something that had an active community like the RepRap forums used to be in the old days.

Mike

Edited 1 time(s). Last edit at 03/08/2021 10:44AM by leadinglights.

There is a pretty active discord server for klipper, it isn't as useful as a forum in my opinion, but you can usually get help there including the availability of the main dev.

[discord.gg]

I have been occasionally looking at the discord for Klipper. As you say, it is not as useful as a forum - but at least there are some signs of life; you can't say the same for the RepRap forum.
Mike

Given the popularity of hobby 3d printing, you'd think this forum would be booming.

It would be nice to coordinate with some popular vendors to host official support subforums here to generate more traffic to the site and rebuild the community.

Quote
obelisk79
Given the popularity of hobby 3d printing, you'd think this forum would be booming.

Personally, I like the self-repairing, self-replicating aspect of RepRaps that use lots of 3D-printed parts and only common hardware as "vitamins". But, as far as I can tell, modern designs aren't really RepRaps anymore since their self-replication capability is diminished due to the use of many custom structural parts that cannot be 3D-printed. I suppose most people these days are interested in buying a ready-to-go 3D printer and using 3D printing as a means to an end, instead of being interested in the self-repairing, self-replicating aspect of a DIY RepRap -- which might explain why this forum isn't as active as it used to be.

... another aspect of the "diminishing" is the fact, that any interesting development, we made in the forum am posted as OS/OHW was pretty fast copied and used "comercially"

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Quote
qrp-gaijin
........................................
Personally, I like the self-repairing, self-replicating aspect of RepRaps that use lots of 3D-printed parts and only common hardware as "vitamins". But, as far as I can tell, modern designs aren't really RepRaps anymore .......................................................

That is what appeals to me as well: Originally, the RepRap movement was about finding and publishing knowledge of materials and techniques that could be used to make tools that make tools - and goods. It seems though to have become little more than a discussion about which commercially available extruder, controller, hotend etc., to buy.

It saddens me that if I research a readily available resource, and find it equal or better than commercial alternatives, that this bit of knowledge is dismissed as wrong simply because it is not commercially made. If I could find some new hobby where I got as much pleasure in the simple seeking of knowledge as I did at the beginning of the RepRap movement I would do that instead.

Mike

... one of my personal goals since my joining in 2007 was "laser-assisted multi-material fabbing" -- e.g. melting sand with a laser or focussed sun-light to solid parts ... or powders from plastic, lignin, glass, ceramics, metall , ...

This too was then and is today done comercially with metal powders, but in pretty big/pricey setups, so not ideal for DIY and hobby.

It was mostly me, finding the pumping IR-laserdiodes, salvaged out from broken fiber-lasers with powers of then available 5 and 9Watts on a spot of 0.1mm as cheap source (compared to comercial systems) ... but only sintratec made it "big enough" to get comercially ... all other starting groups and individuals didn't get enough momentum, to trigger media interests ... or lost interest ... or were "cleared"

After several related R&D-projects for companies I'm now again free to make some more progress "on my own" - so me (in Germany) and a friend (in Spain) are again busy with developing drivers and modules for laser-sintering ... actually I'm building around IR-laserdiodes with powers of 200 and 270Watts on a spot of 0.2mm -- so more than enough energy density for much more than "only melting sand"

But it's not so easy and fast to DIY such a project alone in my basement ... so can take a while (again) ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

I was a little shocked that you have outgrown your duet & all the expansion possibilities, & what must have been quite a close exchange of ideas to work on, I doubt i'd have ever outgrown it's use...but couldnt decide to spend that amount knowing that if I went down the cnc mill/lathe route something else would be better suited. Break out boards so cheap, currently waiting for the tiny SKR Dip, didnt really want to go BTT but for £14 it should be fun trying to set it up. I wonder if in the future these tiny boards could be made rackable on some backplane, maybe like a row of M.2 drives.

I don't imagine that deckingman has outgrown the Duet, simply that he has a problem which he is unable to solve and the Duet people are unhelpful. Considering the price of the Duet, along with its expansion boards as well as deckingman's contributions to the troubleshooting and popularising of the Duet controllers, it is easy to sympathize.

Mike

I agree with the sentiment that the community writ-large has shifted towards commercial off-the-shelf options. Not that the concept of self replicating/repairing has completely died, but it is certainly no-longer the prevailing mindset of the 3d printing newcomer. I also generally agree that large percentages just 3d print as a means to an end. I personally spend much more time tinkering/designing/modifying my machines than I do using them to make other useful objects. Often when I have new (to me) ideas, I find that they were already explored to at least some extent here.

Oh well, I'm happy to be here in good company still, even if things are slower and quieter than ever.

VDX

What is your source for powders for sintering?

I have found small quantities of PLA or other powders very limited

Tnx

... I have two big bottles full of white and black PA powder from a friend ... somewhere a bag with 2kg lignin-powder from Tecnaro ... and some 100g-bags of different metal materials (iron, steel, aluminium, titanium, brass, bronze, ...) for testing purposes from another German company ..

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Quote
MechaBits
I was a little shocked that you have outgrown your duet & all the expansion possibilities, & what must have been quite a close exchange of ideas to work on, I doubt i'd have ever outgrown it's use...but couldnt decide to spend that amount knowing that if I went down the cnc mill/lathe route something else would be better suited. Break out boards so cheap, currently waiting for the tiny SKR Dip, didnt really want to go BTT but for £14 it should be fun trying to set it up. I wonder if in the future these tiny boards could be made rackable on some backplane, maybe like a row of M.2 drives.

As Mike (@leadinglights) correctly surmised, it is/was not the case that I had outgrown the Duet "ecosystem" but more the fact that after "upgrading" from Duet 2 to Duet generation 3 back in July 2019, some 20 months later, I was still missing some of the basic functionality that I had with generation 2. Seemingly the ability to run PID tuning on heaters, or run >16X micro-stepping on extruders due to severely limited step pulse frequency limits, were less important issues than (say) controlling NeoPixel LEDs for fancy lighting effects. Call me old fashioned if you like, but I'd prefer to actually be able to produce decent prints than be able to view "dodgy" prints under RGB lighting.

Matters came to a head when after "upgrading" to the latest stable firmware, the extruders would, more often than not run backwards instead of forwards and I got "homing failed messages" every time I homed the printer. After 3 or 4 weeks of not being able to use the printer and having no response to the post I made about those issues, in desperation I made a YouTube video pointing out my frustrations.

It was around that time that I started this thread. However, I can now say that the video served to focus peoples' attention and those matters are now all resolved. So I can now say that my printer once again has the functionality that it had with Duet Generation 2, and the only thing I can "winge" about is that it took nearly 2 year to get to this point. So as of now, I'm still undecided whether to stick with RRF firmware and/or Duet hardware or go the "Klipper" route with Duet hardware or some other hardware. At my age, I have to consider if I can "afford" to spend the time climbing another learning curve............

---

[somei3deas.wordpress.com]
[www.youtube.com]

I'm surprised at the focus on stepper and interrupt rates. With the latest range of TMC drivers and the like, the interrupt rates aren't so important. Like with 2209 or 2208 even.

The microstepping is now all 'simulated' in the driver and therefore you get microstepping without the interrupts.

Quote
NovaHuta
I'm surprised at the focus on stepper and interrupt rates. With the latest range of TMC drivers and the like, the interrupt rates aren't so important. Like with 2209 or 2208 even.

The microstepping is now all 'simulated' in the driver and therefore you get microstepping without the interrupts.

I don't see any mention of interrupt rates anywhere in this thread. You aren't conflating on chip micro-step interpolation with the maximum step pulse frequency that the firmware is capable of sending to the driver chips are you? The fact that the driver chip does "internal" interpolation between individual micro-steps doesn't help much if the frequency of those micro step signals being sent to the driver chip is too low. Or have I missed something?

---

[somei3deas.wordpress.com]
[www.youtube.com]

Quote
mishaelbrown
Importing data from a variety of database types into our format. We gave some basic information on the process that we were using within our company and no documentation on the database itself. I have done an excellent job of familiarizing myself with our database and other database types, creating their own processes to more efficiently import the data, and working with customers to keep them updated on progress.

Assuming that this may have something to do with the thread "Re: Looking for new control boards and firmware" I guess that you have missed out a link at least, perhaps a few words, a line, or even a paragraph went into the digital black hole. Perhaps, driven by the same Twitter pressure that would shorten the whole of Hamlet to "to be or wot???" you have succumbed to Brevity19 before completing your posting.

Please enlighten.

looks like robo spam to me

and to go back to Deckingman, and end that convo,
interesting stuff, luckily i'll never run into those issues, good to hear it's all working fine, though if you keep going with the printer development it will either outgrow the duet 3 or the room. Do you plan on developing other ideas for other printers where you would need even more out of the equipment or something else with less complexity.

Edited 1 time(s). Last edit at 04/09/2021 04:58PM by MechaBits.

Quote
MechaBits
looks like robo spam to me

and to go back to Deckingman, and end that convo,
interesting stuff, luckily i'll never run into those issues, good to hear it's all working fine, though if you keep going with the printer development it will either outgrow the duet 3 or the room. Do you plan on developing other ideas for other printers where you would need even more out of the equipment or something else with less complexity.

TBH, I'm taking a break from 3D printer development. The 2 years of frustration with Duet gen 3 firmware have left me a bit disillusioned with the whole thing. I do what I do mostly for fun and the last couple of years haven't been much fun. I just need a break so I'm throwing my energies into other things at the moment - mostly home automation.

But I still plan on revisiting the mixing hot end to make one which will actually mix. Mostly because I is see it as a challenge and also because my last iteration was showing promise. But realistically, that only works for colours of the same filament type. For the future, maybe I'll look at multi material printing which will need a tool changer of some sort and probably need to be enclosed in a heated chamber. An infinite Y axis might be good as well. I don't think anyone has managed an infinite Y axis inside a heater chamber so that might be a challenge.

My printer really isn't that complicated - at least as far as the firmware is concerned. There are many people who use mixing hot ends. There are others who use the CoreXYUV kinematics that I came up with to have the extruder(s) on its/their own gantry(s) above the hot end. And there is at least one other user who has a the third AB load balancing gantry but that doesn't actually help with print quality. The only reason it's still fitted is that I can't be bothered to remove it. The CoreXYUV can be run by mapping the U motor to the X axis and Y motor to the V axis so it comes down to being a CoreXY with a mixing hot end. The Z axis is just a single motor driving 3 screws. It's mechanically flat, level and stays that way. So I don't use any form of firmware levelling or flatness compensation. The Z homing "probe" is a simple as it gets. It's just a brass plate on the hot end and another on the carriage which complete a circuit when the hot end is seated and breaks the circuit when it is unseated (when the nozzle comes into contact with the bed). It's the simplest form of switch as far as the firmware is concerned - no interface board board required, no probe to deploy. The complexity is in making a kinematic mount that allows the hot end to move in Z (against a spring) but not in X or Y. The Bondtech BMGs work really well and there us no extruder slippage so I don't need firmware based non-linear extrusion. The mass of the hot end is such that I don't get ringing so I have no use for dynamic acceleration adjustment. etc etc. The list of firmware "features" that I don't need/use is quite extensive.

---

[somei3deas.wordpress.com]
[www.youtube.com]

@deckingman
I will await your return to the forum and of course to YouTube fully invigorated with lots of new and interesting ideas and designs.

btw, you are but a spring chicken. I celebrated my 75th a week ago. I would normally keep my antiquity top secret to avoid inducing culture shock in the under 30s, but I seriously doubt that there are many who read this forum.

Mike