How to manage a farm of M200's?

Are there any commercially available options for centrally managing a suite of Zortrax machines?  I'm talking 20+ machines so more than I'd like running on SD cards!

Thanks

Paul

Maybe consider WIFI SD Cards for the machines and some cheap web cams. Haven't heard of other options management options.

Toshiba FlashAir w02 should do the trick in terms of wifi, [url=http://support.zortrax.com/m200-maintenance-guide/]regular Maintenance plan[/url], UPS and cameras. 

It would be nice if zortrax would develop a wifi/ethernet module to control basic operations. Like ultimaker.

Also, the M300 maybe an option; it's 2.25x the table area, which means about 2.25x less management needs to happen if your parts pack well on the table. M300 is 2x machine cost, but throughput may not be any lower despite half number of machines. For example a full table of parts takes 12 hours on m200, and company runs from 9-5, then on m200 you can run 1 table in 24 hours. On M300 you could run equivalent of 2x m200 tables over 24 hours on 9-5 working hours while saving employee costs. Basically "lights out" to much greater degree.

Dependent on what you printing, company scheduling, pay rates; requires careful analysis.

Also consider 2000g spools over 800g, longer life nozzles, Z-Mon for either machine.

Thanks all... still not quite the solution I'm thinking of though (although I totally get the M300 comparison suggestion though)

Lets say I have a file I want to print and I want to kick off all of my farm of 10, 20 or even more printers without the need to load a separate SD card for each one or send the print instruction to each one.  I want to kick off the job and then monitor status centrally.  Is there anyone out there even running like that at the moment?

Thanks all... still not quite the solution I'm thinking of though (although I totally get the M300 comparison suggestion though)

Lets say I have a file I want to print and I want to kick off all of my farm of 10, 20 or even more printers without the need to load a separate SD card for each one or send the print instruction to each one.  I want to kick off the job and then monitor status centrally.  Is there anyone out there even running like that at the moment?

I understand your desire to automate as much of the operation as possible. I've got a six M200 print farm at this point with plans to double it this year and I'm more than happy to share with you what I've learned and how I'm doing things. I'll share more as time permits.... -Kind Regards, David Barwin

I think the control/electronic solving of the handling is just one part.

​In my 3 Month experience i tend to print one part/job, multiparts get less quality/equality in surfaces and dimensional accuracy. If you have stringing like me, everything else is just peanuts! Fiddeling away hundreds of strings/dots with the tweezers or knive and hitting it with the heatgun… If i would print fulltime, prozessing parts would be my most hourspending employment per day.

As mechanical engineer, i would look also into hardware improvments like multiple printbeds per printer with quickchange characteristics. So you can prozess finished print parallel to the running machine and have prepared empty printbeds on stock. The next Level would be a railsystem with a stock of empty printbeds and a buffer for the finished ones, questionable where the economically border is. (production technology that gained to a stable prozess mostly goes in this direction)

David, how long/how many jobs your printers rund without you touching them inside? (apart from removing prints and scraping bed)

regards Michael

Quick background: We're a Mac based shop that entered the 3D printing industry approx 5 years ago. Initially we used five MakerBot Replicators 2s until we became aware of the Zortrax M200. We also learned that PLA was not going to work for our products because black PLA melts/softens in direct sunlight in a car. We transitioned to M200s and haven't looked back. The reliability, quality and price far surpass anything MakerBot was offering.

I would love to automate more than we currently do but I have learned processes that make an "unconnected" Farm work very well for us.

We have a library of products that we keep a small inventory on hand and assemble / ship as orders come in (eBay, Amazon and Shopify). There is a file limit of 128 per SD card (so don't waste your money on larger cards...2GB are more than enough and cost less than $5 each) and if you have a lot of files the excessive scrolling will wear out the knob and button selector (we've replaced 3 of them). Because of this we simply created an internal SD card holder for each M200 that we split our job files up on (by color...... Red, Blue, Green & White) so there's less scrolling and we can keep up to 512 ready to print jobs on the cards. To keep the cards in sync I routinely use a USB thumb-drive duplicator outfitted with adaptors for SD cards. I can't imagine having to do each card one at a time. I briefly looked into using wifi SD cards but once I learned I was still going to have to go to each M200 to select the file to start the print job I decided to keep it simple and stay with SD cards. 

I've learned that by starting the prints in a sequential fashion (instead of all at once) means they aren't all finishing at the sometime but in a staggered fashion which allows me to round-robin everything to keep the printers going as a single operator. I've also setup multiple piece jobs that take approx 12 hours  and I run those overnight when I'm not at the office and 99.7% of the time they finish successfully without me having to monitor anything. Initially I was looking for cameras and apps to monitor but I've learned that with the M200 reliability it wasn't necessary.

Our biggest ongoing concern has been the sporadic supply of Black Z-ABS. We consume approx 32 rolls (4 cases) a month and multiple times have been unable to locate any within the US for longer than we would care to remember. Our suppliers have been/continue to be 3DProShare, Octave  and ExoticDesigns3D. Apparently the supply issue comes from Corporate. Hard to make money when the printers are idle. Hopefully your products use other colors as everything other than black is usually in stock.

Here's our simple setup with six M200's in my office: MCE1.JPG

Here's our assembly and stocking area:

MCE2.JPG

Here's how we keep all 24 SD cards (six machines w/ 4 SD cards each) synchronized:

MCE3.JPG

Here's how we've split our library of .zcode files since theres a file limit of 128 per card:

MCE4.JPG

I think the control/electronic solving of the handling is just one part.

​In my 3 Month experience i tend to print one part/job, multiparts get less quality/equality in surfaces and dimensional accuracy. If you have stringing like me, everything else is just peanuts! Fiddeling away hundreds of strings/dots with the tweezers or knive and hitting it with the heatgun… If i would print fulltime, prozessing parts would be my most hourspending employment per day.

As mechanical engineer, i would look also into hardware improvments like multiple printbeds per printer with quickchange characteristics. So you can prozess finished print parallel to the running machine and have prepared empty printbeds on stock. The next Level would be a railsystem with a stock of empty printbeds and a buffer for the finished ones, questionable where the economically border is. (production technology that gained to a stable prozess mostly goes in this direction)

David, how long/how many jobs your printers rund without you touching them inside? (apart from removing prints and scraping bed)

regards Michael

Michael ~ I monitor my printers visually and sometimes can run them for over a week without touching anything other than removing print jobs and adding more material. If the initial raft layer looks too thin or too thick I will recalibrate the print bed and when I add new filament if the "loading" stream comes out at an angle instead of straight down then I'll run a .4mm wire into the nozzle to attempt to remove debris. If it continues incorrectly then I will replace nozzle (soak and clean in acetone). If I see too much stringing or odd bits I replace the nozzle and clean up through the print head (debris affects flow and retraction rates). I don't remove the print bed to remove prints .... I do it while in the printer.....careful not to knock off the bed connector). If prints start to come off too easy or lift then I'll use an acetone & filament slurry coating.

Kind Regards, -David

I don't really have a "farm", but both my M200's run production parts regularly and with very little maintenance or babysitting. I tend to run two 3-4 hour prints during the day (4 - 6 parts per print) and 12-hour prints (12 or more parts) overnight, since the risk in time and material is lower with smaller batches while I'm around to abort in case anything goes wrong. I have webcams and remote power kill on the printers but rarely even bother to check in on the prints anymore as they're so reliable. I have no problems with stringing on multi-part prints.

Marcin, I think the FlashAir has to be a w03, not w02 version.

I only have a couple of m200's but my main takeaway is to use high quality third party filament that comes in 3KG spools. I think Zortrax wont honour your warranty if they know you have done this but its cheaper and much less hassle.

ZABS has an unusually high temp of 170 I think, where most ABS is better in the 140-150 range, but one of the other filament profiles is in this range I think, I forget which. 

Also, the M300 maybe an option; it's 2.25x the table area, which means about 2.25x less management needs to happen if your parts pack well on the table. M300 is 2x machine cost, but throughput may not be any lower despite half number of machines. For example a full table of parts takes 12 hours on m200, and company runs from 9-5, then on m200 you can run 1 table in 24 hours. On M300 you could run equivalent of 2x m200 tables over 24 hours on 9-5 working hours while saving employee costs. Basically "lights out" to much greater degree.

Dependent on what you printing, company scheduling, pay rates; requires careful analysis.

Also consider 2000g spools over 800g, longer life nozzles, Z-Mon for either machine.

Well that one made me curious... can you tell us something about the "longer life nozzles"? Which one woul you recommend? Micro Swiss TwinClad? I always have a bad feeling about the dirty nozzle. Asking myself quite often if the z-detection/calibration via blank squares will make proper contact with a completely burt-material-covered nozzle... and we all know what happens if it doesn't :(

Please tell me more about these 3kg rolls of high quality filament.............

Cheers

Andy