In my quest for a better surface finish on FDM parts from the Stratasys (especially with a mind towards having Frankie try some more investment casting), I had decided to try the technique noted in this Stratasys application note. Namely, dipping the parts in MEK (methyl ethyl ketone) in order to fuse the individual filaments together and seal the surface. I poured some MEK into a glass jar and hung two of the Mendel parts onto a length of TIG welding rod. I dunked the parts into the MEK for perhaps 10 seconds, then hung them outside to dry.  I inspected the parts the next day, and I was immediately reminded of the climactic scene in Raiders of the Lost Ark where Toht’s face is melted off.

While the surfaces were most certainly smoother when compared to untreated parts, the side effect of  severe warping and deformation doesn’t make this treatment method a viable option for these small parts.  Spraying MEK onto the parts may work much better, as the loose internal fill pattern I’ve been using makes the parts quite porous, so a little bit of solvent goes a long way.  I’m guessing part dipping may work much better with as solid a fill as possible.

An alternative solvent may also be something to try.  The Stratasys Finishing Touch Smoothing Station uses a vapor bath of a specially formulated solvent, and methylene chloride (the primary component of Weld-On #3) appears to be preferred over MEK in the latest application notes anyhow (earlier versions of the app note recommended MEK and actually mistakenly claimed that Weld-On #3 was MEK – this mix-up had me running in circles for a while).

The treated parts do indeed feel stronger (based on my unscientific method of squeezing them between my fingers to see if they have any discernible ‘give’ compared to the untreated versions), so the treatment certainly has promise beyond just surface smoothing/sealing. So much for the fail – on to the win!

My patience was rewarded on Friday when I had a box from New Image Plastics waiting for me on the doorstep at home – my fresh ABS and HIPS had arrived!  Saturday I ran into work to give the ABS a try, as I still had a reasonable amount of support material left on the spool and wasn’t in a rush to try the HIPS.  Additionally, modifying two parameters of a working system is inviting disaster. Anyhow, if the HIPS doesn’t work out as a support material, it’s not the end of the world – I suppose I could afford to buy name brand Stratasys support material, but if generic ABS doesn’t work well as a modeling material, I may as well start looking to sell the unit given what the official material costs.

The one thing that I had failed to account for was how to take the coil of ABS filament and get it onto the empty Stratasys spool I had. I initially figured I’d just wind it on by hand, but a little bit of math would have told me that 5 pounds of ABS extruded into a diameter of .070″ yields about a half mile of filament. I did end up winding it onto the spool by hand, with a swivel stool seat helping the process a little bit, but it still took a few hours, and my fingertips had a bit of wear. I’ll certainly need to come up with a better solution in the future (perhaps winding the spool on the lathe, or even better, maybe New Image can simply deposit it right onto the spool for me).

Despite being relatively fresh, the ABS had about the same amount of ooze out of the FDM 1600 nozzle as the ‘lobster red’ Stratasys ABS I had been using. Given the high humidity as of late, I suppose this isn’t surprising – I’ll make sure to use plenty of desiccant tins in the dry box. Things were looking good with feeding the ABS through the system, so I ran a single Mendel part for a test.

New Image Plastics ABS on the left, Stratasys ABS on the right

The part turned out great, though it was a little trickier to separate from the support layer than the Stratasys ABS had been.  The NIP ABS certainly equals the Stratasys ABS in resulting part quality, and I have no more worries about running it through the FDM.  In looking closely at the parts I’ve made thus far out of Stratasys ABS, I’ve noticed a bit of variation in build quality, so it will be interesting to see if the NIP ABS provides more consistent results, or if other factors are affecting the created parts.

33 comments so far

Add Your Comment
  1. Hi,

    I’m following your blog with interest, as I’ve recently adopted a Stratasys Dimension 768. I’ve only printed one object so far, as I’m getting round to wrestling an intermittent electrical fault in the machine.

    Have you considered PLA as a support material? The reprap community shows that it can be extruded ok, and apparently it can be dissolved with sodium hydroxide as quickly as the proprietary support material. (see http://blog.reprap.org/2009/10/hello-world.html, comment by Erik de Bruijn, October 06, 2009 1:49 PM)

    What are your plans for dissolving support material? The MSDS for stratasys’ outrageously expensive support material concentrate (http://www.cleanstation-srs.com/pdf/MSDS_SolubleConcentrate_1023810C.pdf) mentions only NaOH as 25% of it’s composition. It also mentions detergent in the title, so presumably some is present, too. I’ve no idea what the rest is, so I just mixed up 250g of NaOH in 1l of water, heated it in a slow cooker to around 70 deg. C, and threw some parts in. I came back to it after about 3 hours, and the support material had turned into a viscous brown slime that could be easily scrubbed from the parts under a tap water rinse. I plan to add an aquarium pump to flow solution over the parts, which I imagine will speed things up further. The whole slow cooker + 3kg of NaOH powder + pump setup totalled about £37 on ebay. Beats the several thousand Sratasys must be asking for their cleaning setup!

    All the reprap guys are using 3mm filament and we need 2mm. Do you think NIP would extrude PLA in a 2mm filament?


    • Jeff -

      I haven’t yet gotten as far as considering soluble support materials, as the breakaway material is working fine for my current needs. Also, my FDM 1600 predates soluble support materials – as best as I can tell, the FDM 2000 was the first Stratasys unit to handle a soluble filament. I’m wondering what changes (if any) Stratasys needed to make to the basic head design in order to accommodate soluble materials – I know that the 2000 (which Zak has) actually uses two roller pairs whereas the 1600 (and presumably 1650) uses only a single pair per filament, but I don’t know if this change is purely for speed, or if it handles a wider variety of filament materials as well. What does the feed system on your 768 look like?

      As far as PLA goes, NIP will definitely extrude in that size – this is actually a NIP page: http://www.reprap.us/

  2. NaOH by itself will not wash PLA away. There has to be a better solution ;)

    I’ve been discussing smoothing techniques too:

    I’m also linking back to this page for further reference.

    • Erik -

      If NaOH by itself isn’t sufficient, what about combining it with an ultrasonic generator? I seem to recall that the Stratasys cleaning station uses ultrasound to dissolve the support material more quickly.

  3. The chart here might be useful too in identifying a good solvent:

  4. Jeff,

    Stratasys just started using a newer support material now that is supposed to dissolve cleanly in water. I have not been able to get a hold of an MSDS yet but will post it as soon as I do. I have a feeling that they may be licensing it from another company.

  5. If you have to do more winding-of-filament-onto-a-spool, see me for some things which might work for you: simple Dad-made quill spinning wheel (mount empty spool onto the quill), converted sewing machine motor bobbin winding device, also Dad-made, or converted electric mixer winding device (Dad-modified).

  6. AWEsome!

  7. Erik, I’m intrigued, did you not find that PLA and the proprietary support media dissolve at the same rate in NaOH? Stratasys heat and agitate the solution with a pump/ultrasound. I found a third party cleaning machine vendor that claim that ultrasound can delaminate parts, but it was incredibly biased, so who knows.

    NaOH, slow cooker and aquarium pump are working well for me with the stratasys support media.

    Have Blue; my extruder has only one pair of wheels per filament. Looks like it would extrude any plastic with the right diameter and melting point. I think most of the extruder problems people encounter with reprap, etc. are due to the 3mm filament requiring much higher extrusion pressures, and design sacrifices made to make the extruders printable/laserable/machine-tool-free. A high-torque geared servo motor driving a high pressure spring loaded high friction pinch wheel should extrude anything, I’d expect.

  8. Jeff-

    I’ve been wondering myself if the decision to use 3mm filament for reprap was rather short-sighted versus the .070 our Stratasys units use. I think the RepRap developers wanted to maintain sufficient ‘buildability’ with either metric or imperial hardware, and 3mm is just a hair away from 1/8″. I think you’re dead-on in your appraisal of the extruder designs suffering as a result. Unfortunately, it looks like 3mm is indeed the established standard – thank goodness Jim @ NIP is around to supply us with .070″. The one advantage I can think of to 3mm filament is that it makes infill that much easier – you can basically use a firehose nozzle to rapidly put down large layers. But given the current RepRap working envelopes (8″x8″x5.5″ for Mendel, versus 10″x9.4″x10″ for my FDM 1600), I think that feature may be of limited utility. Still, I wonder if that might be a good multi-nozzle feature for RepRap – a small .012″ nozzle for contours, and a big .030″ nozzle for infill?

    On my initial subject of MEK for smoothing, I remembered that the fellow I had bought the Stratasys from used a Preval sprayer to spray a solvent (he couldn’t recall the name, but it would have to be either MEK or Weld-On 3) onto completed parts if he found that the countours were delaminating from the infill. I’ll have to try that on some parts from an upcoming batch.

  9. WOW you have put a lot of effort into this. MY HATS OFF TO YOU! I was just told about your blog and am hoping to get some insight on a project OR should I say “can of worms” I am about to undertake. Maybe I should read more on your blog as Im just starting to go threw it. I do see some of the issues you have had with trying different ABS materials. That is my quest!

    I just purchased a used Dimension SST 1200 “limited use 650 hrs on it” and of course the 1st thing on my mind is “I am not paying $250.00 bucks every time I need material!!!!! Stratasys must be crazy ABS is dirt cheep. I’m sure I can just get some off the shelf ABS filament, install it into the Dimension cartridge figure out a way to reset the internal chip inside the cartridge etc. bada boom bada bang! parts for pennies!

    Probably not that simple right?

    Stratasys claims the ABS is very moisture sensitive and if the cartridge should be opened shelf life is reduced to a matter a few days. IS there any truth to this? Any input? I dont buy it!

    Dose anyone know EXACTLY what ABS material & where to source it ? will work on my new machine?

    Next step is to develop a way to reset the internal EEPROM chip inside the cartridge. I have an electronics engineer friend of mine lined up to develop a device to reset the internal chip in the cartridge. His input is fairly confident our observations are correct and it is doable.

    IF anyone is willing to share any input and experiences down this road I would be grateful. You can email me directly.

    • Hi Bill! Do you have any pictures of your machine’s innards? I’ve been curious to know what controllers Stratasys is using on their newer machines (I’m guessing they haven’t used Asymtek for the Dimension series in order to save on the cost).

      ABS will absorb moisture, but how much this affects prints is something I’m not sure of. Despite keeping my filament in a dry box with desiccant tins, I still get a fair amount of filament coming out for the ‘ooze test’ (run a foot of filament through the extruder, stop, wipe nozzle, wait 10 minutes and measure the length of oozed filament). In a hobbyist capacity, the humidity probably isn’t terribly important – as far as I know, none of the hobbyist level printers (BfB, RepRap, Makerbot, UP!, etc.) keep the filament in a dry box – everything is out in the open. Testing the effect of water absorption on print quality would be a great experiment.

      As for what ABS material, please read my latest post for information on this subject.

      Finally, regarding resetting the EEPROM, many people have contacted me asking if I know of a method. As I don’t have one of the newer cartridge-based machines myself, I haven’t been able to hook one up to a DSO and see what sort of data is being transferred. I understand the cartridges use a DS2433 1-wire EEPROM – it should be a pretty simple manner to check the contents of the EEPROM at various stages (brand new, halfway used, empty) and compare the data. Of course, it’s possible Stratasys obfuscates the data (perhaps using the EEPROM’s unique serial number as a hash), but I doubt they’ve gone that far. I have heard rumors of people successfully resetting the EEPROMs, but have not found any verifiable reports so far – just a lot of people wondering how to do it!

      • Hey Have Blue, Whats your Name?

        Thanks for the reply. My head is spinning. Ive been going through you blog trying to decipher all you hard work. You have really acquired a vast knowledge! After reading a lot of your input I’m starting to wonder if its worth actually trying any thing other than the OEM material in my machine. At this time the machine is not under warranty has not been re-certified by Statasys, I have no service contract, I just bought it used. I am after all using it on a professional level and was hoping to actually use a another ABS for parts I will use in my work. Man I would really like to open a line of communication via email and phone if your willing. Please let me know if your willing.

        Let me ask you this. What are the potential hazards of just trying any of these ABS materials in my machine? Could it permanently render the system clogged and unusable? one thing I am considering is in the future is upgrading this machine to a 1200ES model. Witch I’m told entails swapping out some of the hardware. This would require me to re-certify the machine as well as put it under contract.

        BTW: I just contacted 3 of the companies you mentioned regarding ABS replacement for the P400. Sibco, Bolson & Argyle.

        • If I were you, I’d stick with the OEM materials for the time being – at least until we can find suitable alternatives.

          Potential hazards? I doubt that it would be as bad as ‘clogged and unusable’ – as long as you run injection molding grade materials, things should probably be fine. Extrusion grade materials, however, would be a bad idea.

          - Michael

    • Regarding moisture; I’d asked a Sys tech about issues with support not sticking- he mentioned that if the filament was moist when it when into the head the steam cause would ne trapped in the tube and cause a foaming and spitting issue. I have no personal experience so i can’t tell you more than i was told.

      myself, I went to Smart & Final and got a 5 gallon clear food container with lid. In the bottom i piled up some extra descant bags from old spools and loaded my partially used spools on top for dry storage- other than big silver Mylar bags.

  10. “Next step is to develop a way to reset the internal EEPROM chip inside the cartridge. I have an electronics engineer friend of mine lined up to develop a device to reset the internal chip in the cartridge. His input is fairly confident our observations are correct and it is doable.”

    “Finally, regarding resetting the EEPROM, many people have contacted me asking if I know of a method. As I don’t have one of the newer cartridge-based machines myself, I haven’t been able to hook one up to a DSO and see what sort of data is being transferred. I understand the cartridges use a DS2433 1-wire EEPROM – it should be a pretty simple manner to check the contents of the EEPROM at various stages (brand new, halfway used, empty) and compare the data. Of course, it’s possible Stratasys obfuscates the data (perhaps using the EEPROM’s unique serial number as a hash), but I doubt they’ve gone that far. I have heard rumors of people successfully resetting the EEPROMs, but have not found any verifiable reports so far – just a lot of people wondering how to do it!”

    reset eeprom not so easy as you think(
    all data inside eeprom are encrypted.
    each eeprom has unique ID.
    data correspond to this ID.
    and final, printer remember how many percents each chip has.
    if anybody work in the same direction please email me!

    • I’ve theorized the same.

      email me, i’ve a gmail account.

    • Hello,

      Today I have read 2 eeprom’s form a Stratasys Dimension machine. One is Ivory P430 and other SST (P400SR). For now I can only say that may be encripted, but form the support cartridge I can read “STRATASYS” on the dump. From now I have only read Them (I don’t want to write now because have material on them). I have do it with an arduino board without problems. The have DS2433 chip inside. It not needed open them to read/write the eeproms.

      I will use them in these days and I will compare data. Someone more to make things like this?


      • You should be seeing ‘STRATASYS’ at locations 0×68-0×70 on the model cartridge EEPROM as well. Yes, the data is encrypted.

      • Well,

        If helps, the purge and manteniance operations doesn’t modify the eeprom content.

        • How much material was run through during the purge operation? I recall seeing in a Stratasys patent that the material amount is stored as linear feet – if the purge operation didn’t use a full 12 inches of material, it may not have written back to the EEPROM yet.

    • Hello again,

      And about the posibility of the machine “memory” of the chips ID. Mine has Linux (RedHat) and It is a Celeron based system. It can be “editable” system taking out the Hard Disk.

      Also, some one have “Maraca”?


      • You don’t need to remove the hard disk to edit – I’ll have information in the next blog post.

  11. Hi Micheal, I’m wondering if you have any input on the MG47 & MG94 material you are having extruded working on my Dimension printer? Do you think they are worth a shot?

    btw: Have you gotten my last couple emails?

    • I definitely think they’re worth a shot, hence why I’m taking a chance on getting them extruded. Last email I got from you was on Jan. 6 – did you send anything since then?

  12. HaveBlue, Bill and others interested in hacking Dimension machines:
    I think I know how to do it (in principle, anyway). I have a Dimension 768 SST and have been planning on working on this also. Want to collaborate?

    Also, I have a lot of experience working with the Dimension extruders so I can offer advice on that if needed.

    Did any of you ever figure out what the Dimension soluble support material actually is? Is it PLA?

    I’m also building a mendel prusa in order to be able to try out mods and variations more quickly. I was considering adding a second extruder for support material.

  13. I was looking at your main board photo, and I see J1 is labeled ‘Watlows’. On my FDM2000, I also have the same configuration, but there’s a piggyback board with a PIC chip just to the right of J1. The FDM2000 has Walow 96’s (pretty extravagant) and the PIC chip appears to be providing MOD BUS control to them. When I power up, there’s a message ‘Low Temp Set’, and you have to press enter to clear the message. I found manuals for the Watlows, but they seem to be locked out from any button presses to get into them that way.

    Next to my Watlows there are stickers for multiple different temp settings for model, support, and just a single for envelope. On the Model there is ‘Run 270′ & ‘Done 50′, the support has ‘Run 235′, ‘STBY 210′, & ‘Done 50′.

    Any thoughts on how Stratasys might have worked this, or how to get into the MOD BUS? Also, how does Quickslice, or Insight send the muliple temp settings in a SML File?

    Thanks –John

    • It’s been a while since I’ve done any Modbus work, but when I did, it was – you guessed it – temp controllers! You’ll need an RS-485 serial port (you should be able to find a USB to RS-485 converter for around $40), which you can then connect to the temp controller via the appropriate terminals. I think I used some very simple software to query and set values over Modbus – shouldn’t be hard to find something.

      I don’t think Quickslice or Insight actually send any temperature settings to the machine – on the P-class machines, the temperatures are all determined via config files within the machine itself. For the FDMs, I think the machine decides on its own (with no input from Quickslice/Insight) if it’s been idle long enough to drop to standby or all the way down to ‘done’. I’ll have to dig through your FDM manuals to see if there’s any info on that…

  14. Hi have blue.

    I’ve got a functioning FDM1650 I’ve resurrected but the support extruder won’t stop extruding and so won’t let the model extruder extrude. Wondering if you’ve had this issue?


  15. Acetone works the same as MEK. Easier to get in big cans at home Depot.

    • Acetone is probably a lot safer than MEK as well – MEK is NASTY stuff, and you don’t want to be using it in an unventilated area. I just came across a paper that noted a strength increase in 3D printed ABS parts after dipping them in a 10% water/90% acetone mix.