Joel diagnosing a commercial printer with Duet.
This is a recent episode of the 3D Printing Nerd in which he diagnoses a commercial printer with a Duet board. This is a new machine he got recently from the vendor.
A couple of interesting points
The firmware is very old. The vendor qualified it once and then stayed with that version.
Joel, as a user, can edit gcode config files. This probably makes warranty, support, and upgrades very tricky for the vendor.
The Duet is less of a blackbox like a TV and more like a personal computers where end users can easily mess it up. Do companies such as Stratasys let their users reconfigured their printers or are they more tightly controlled?
Edit: video link here https://www.youtube.com/watch?v=DXY9IilBwWQ&ab_channel=3DPrintingNerd
companies such as Stratasys let their users reconfigured their printers or are they more tightly controlled?
I have an older Stratasys Dimension 1200es BST in my garage. Things may have changed slightly since... but you cannot change (officially) anything on the system. CatalystEX, the slicer/plater for the Dimension series, had very few options to even select - two layer heights, three infill densities, three or four support structure types (and no option to disable support). You cannot manual set print temperature or modify it on the fly. You cannot change print speed, or adjust machine speed/accel/jerk/etc limits. 'Insight' iand GrabCad Print are much more powerful slicer/ploters for their printers now.
Firmware updates did occur (and were done prior to me receiving the printer) but I believe they would have occurred by a VAR (Value added reseller much like Trimech) sending a tech out to perform the updates.
One can argue that Stratasys is a destructive force in the world of additive manufacturing, but their philosophy with these machines were "press go and don't worry about it." Their locked down approach ensured that prints would come out - low fuss. Sometimes they don't, but their choice of parameters typically result in higher success rates than printers where a user can willy nilly tweak to their hearts content.
you noticed too that 38k printer he f^#%$^ the glass on is on duet and using old rrf ...
I had some experience with a lot of high end machines trough the years, both as a user and as person repairing them. it mostly depends on the industry, for e.g. medical machines are as closed as you can imagine making servicing very hard and expensive. it is a big issue as often those machines are link between a pc and a custom hw linked using some proprietary shared memory bus and what happens for e.g. is DOS floppy disk dies and you need an hour to get to it for start and then you can't fix it 'cause you are missing a driver for shared memory that was on the floppy
In the 3D printing world I don't have too much experience but the ones I have seen are 100% closed too, no access to nothing, can't even set your temp, you have to use proprietary materials that come in cartridges with chips and similar BS
Now if we turn our eyes to the CNC machines that are similar devices, most professional machines give you full access to the configuration. You can always "reset to defaults" but you are allowed to kill both yourself and the machine.
Tha automation lines on the other hand are usually something in between. The most popular PLC's are the ones you can easily hack to extract source code and edit it. Any serious factory would either require full source code and unlocked PLC's or will go with system where it is known PLC can be "open"... very rarely factory will allow automation system that's "locked"
I'd blame the breaking glass on outdated firmware..
My only commercial printer is the "Teartime" as someone (that guy ) refered to them as; slowly opening up for making slight changes to slicing parametres; but cannot change anything firmware related. And it makes sense, if you are to avoid support issues then users will have to stick to what is tried and true (even if its just the one perimeter..)
@bearer said in Joel diagnosing a commercial printer with Duet.:
I'd blame the breaking glass on outdated firmware..
They supplied him with magigoo and he decided to print PETG straight on to the "unknown glass" using "unknown extruder and unknown hotend" .... he should have enough experience to know how strongly petg can bond to glass but it all makes good TV, lot of clicks...
on the TT story, if you followed from the beginning, we had to add hardware gimmicks just to be able to print at lower temps (as normally they print ABS at 270C) but after so many ppl started doing that mod and then tinyfab came out with the CPU replacement to run smoothiware (I think he's shipping with my fork for few years already) they got pressured by the community to open up a little...
but wrt "commercial printers" look at ultimaker, expensive, nice, open... raise3d, great printers, not open source but again, fairly open, craftbot, again, pretty open... they are not all necessarily open source but are all pretty open ... you can tweak any parameter on most of them ...
and looks like even the MSLA are slowly opening up
@bearer said in Joel diagnosing a commercial printer with Duet.:
f you are to avoid support issues
I think the "closing" is 100% related with pushing own consumables (nozzles, filaments...) and not to support/quality issues
@bearer, thanks for the links! That is an awesome printer!
Good point .... but I would be afraid that it would need a lot of tweaking and ironing before you end up with a reliable passable product. I would hope that I can plug in the $38000 printer and produce reliable good prints from print #1. Sure I'd like to tweak it after the fact but the first print should come out perfect.
If you look at the cost of the printer and say to yourself "I can do better for less money" then you either have a LOT of experience or you do not value your time at all. Either is fine but sometimes you just want a product you can plug in and use.
If I would calculate my labour at minimum wage and multiply that by the hours I spent screwing around with my printers, I would probably be rather comfortable right now
Rethinking this a bit ... it's difficult to separate the hours I have spent based on hardware issues vs all the other issues that crop up so I don't know ....
Hmmmm .... might be an interesting exercise to build as if money was no object with only the best of all components used.
Did they say 2000 lbs in the video for that printer? That is a lot of weiht unless there are concrete blocks in it to increase mass!
I would hope that I can plug in the $38000 printer and produce reliable good prints from print #1.
products can be expensive for many reasons, quality and ease of use is one of them, but just because a product is expensive, doesn't mean it's easy to use. That's a very prevalent consumer fallacy. The cost of this particular printer comes from the sheer size of it (large cabinet, heavy gantry, expensive shipping and packaging, massive borosilicate bed, custom extruders, etc etc). I would argue that in this case not a lot of time and effort has been put into making it easy to use at all. They validated a firmware version, wrote the manual and documentation, and stuck with it. They likely provide hands on service and training for the end user and maybe even sell a service contract with it. The fact it's used in larger industrial scale shops proves that it does work and can work reliably, but it's definitely not easy to use.
Case in point, I would say a very easy to use printer is the Prusa mini and it is not expensive at all. Ease of use is one of it's design goals.
TL;DR Expensive != Easy
Fair point ....
I've worked on these machines a bit. The company who builds these started out of a linear motion component manufacturer, so the mechanics are pretty bulletproof. They have a dry teflon-type coating on the lead screws with wear-compensated anti-backlash nuts, it's a good setup. I think the main idea is for the machine to run reliably for a long time with minimal maintenance.
It would be possible of course to design and build something better, but it will certainly cost more money if you're paying someone else to do it, and good luck on docs/support/service. For most companies it's probably not worth the risk compared to buying this machine, bigrep, or another from an established player with everything already in place.
I would love to tackle a build on this scale, but then I would have a project rather than a printer. Not to mention I have no space remaining in my apartment to put it
@nhof I think that's your comment in the resume.g file in Joel's video
I would love to tackle a build on this scale
I build 100x100x100cm printer some 8 years ago, used it for a year and then had to move (purchased new appt) and I had to rip it to pieces as the darn thing was not able to go trough any door/window in the old apartment ... used the parts for other projects, thrown case into the garbage, learned few very important lessons
- MDF is awesome build material but you have to glue and paint it. It is strong, dampens vibrations crazy good .. it's wooden pandan for cast iron
- when you are building something you ever plan to move make sure it can go through a bloody door
- having huge volume will attract attention (clients - both for making printers and for 3d printing service) as it is a great showcase of what one can do but in reality you almost never need more than 30x30x30 ... 80% of jobs will fit 10x10x10
- if you are making a big volume printer, make an automatic toolchanger
there should be some documentation about that printer on reprap.com, it was called "cut&paste" 'cause I had both extruder and milling head on it with idea to one day write a slicer that will trim the printed part, drill holes to size etc etc .. but while it worked awesome as printer and as cnc I never managed to write software to use it for intended "cut&paste" work
now wrt "38k", for e.g. that printer cost about 6 months of work and about 5000eur in parts, 80% of which did not end up in the printer. final "BOM" value for the printer was around 3000eur for a device that prints in 60C enclosure, heated bed that goes up to 150C, milling head that can cut plastics and soft wood, in theory it worked on aluminium too but you would need mist cooling or something and I doubt MDF would last long with that even with a strong poliuretane coating that I had there..
so the dev cost, let's take the low end engineer price here of 2k/m x 6 months is 12k + 5k for parts used, I was renting the appt where taht was build cheaply for 400e/m so another 2k4 .. ignoring 12h workdays, weekends, cost of tools, friends helping out.. etc etc ... development of that machine is roughly 20keur .. and that's "single guy in the garage", no company that need to pay taxes, salaries, sick days, vacations, parking ... money goes super fast when you are designing things, every single "lets try this" can cost thousands, and to be sure what's the best course you have to try a lot of things... and you still end up with fails (this printer from this example I count as fail as it did not manage to do the cut&paste job it was made to do and 'cause it had to be destroyed after I left that appt) ... you learn from these fails but they cost
@droftarts ha, that's the page Sebastian created for me pushing me to document the darn thing, looks like I never copied the data from my dead wiki past the initial data (that image on the page is also not of that printer btw, just a placeholder Sebastian put there) ... there were some images on the google picassa I think, from what was not destroyed in a fire in 2013 (had a fire in my new appt, fileserver burned down, lost 30 years worth of projects, documents, images... on the blood sugar meds since then ) but all that is now gone too .. I have few of them on the temp upload folder but nothing from that big one... I found some of the pages from the old, dead, wiki are available on the webarchive but I can't find the pages about that printer even there ... anyhow not many pictures were ever made as I was staying away from smartphones for a long time and prefered my EOS REBELG (35mm camera) so with limited number of images per film one do not snapshot images of unfinished products like you do with unlimited storage and digital camera
IIRC Sebastian Bailard or David Buzz gave me the idea to integrate the milling head so I can do "cut & paste" ... too bad I never got the software to produce anything useful .. I still think it is a great idea
EDIT: def Sebastian, just found the email 2010-09-25 07:59AM
n : print, clean-up outline with milling cutter,
n+1: print,. etc.
cut-and-paste reprapping is a much better name than shape deposition
manufacturing, but it may be difficult to get the community to adopt
ah, the history