Topic: SD Press - Mark II (Modifications Breakdown)
This is a list of mods I've done to my SD Press. Just for historical sake. This is a Gen. 1 Pre-order version of the Press.
(Just in case things have changed in future revisions of the Press.) This not a list of what to do, more of a list of what I've done.
This is more of a documentation then a guideline. Do your research. Even if you find a definitive guideline of what to do, Still do your homework. The more you learn about it- The better equipped you will be to handle issues if/when they arise.
TLDR;
Replaced the hotend with a E3D v6 (http://www.filastruder.com/products/all-metal-e3d-v6-hotend & http://www.soliforum.com/topic/9709/e3d … e-press/). Replaced the Heated Bed with a MK2B (http://www.amazon.com/Easy-RepRap-Solde … 00M9TQ18A/) and used a sheet of 468P (http://www.amazon.com/gp/product/B007Y7D5NQ/) to stick the PEI (http://www.amazon.com/dp/B0013HQ7B0/) directly to the MK2B.
Out of box experience.
When My press arrived. It the box was in good shape and the support packaging was very good. Leaps and bounds better than how my SD 2 arrived. Pull everything out of the box and plugged everything in. I glanced over the docs for the printer that came in the box and it said to use with 'SoliPrint'. My initial thought was I am never going to bound this to my desktop and I am going to use the same workflow I use for my SD2. RPi (B+) /w OctoPrint (devel branch). So I temp move my SD2's RPi to the press, Fire it up and detected and connected to it with no issues. And within seconds I was able to jog things around. One of the first things I've noticed was that X/Y needed to be homed before Z. Different, But fine. (Later one after diving into the FW seems like this was a normal behaviour with printers with a probes ment for detecting Z-Offsets). As for the first print. That was an adventure. If you have been following my other posts (Under the Kimono - http://www.soliforum.com/topic/9214/sd- … e-kimono/) pretty much everything that could be looss was loose. Every grub screw, chassis screws everything. Then the KOD (Knock of death). After I was able to correct all the QA issues. I was not able to shake the KOD.It would work for about 3 ~ 8 layers (or 25 mins - what ever came first) then simply stop extruding. It was repeatable. I've contacted support about it and the replies I got where like they didn't read what I was asking. At the end after a 2 week no response I got a automated email asking how would I rate their service. Then the upgrade started.
Hotend.
After many hours of screwing around with the stock hotend. I've determined this. After a couple layers would go down. The filament would deform and 'bow' inward and above the hotend inlet. Which now you have a square peg trying to get into a round hole and impedes the extruder stepper from doing its job. In my case, The KOD was due to the stepper missing steps attempting to push the filament into the hotend. After a while it then grinds the filament into a fine dust. Using a Fluke 62 Max (Yes it's a IR based and there are more accurate ways to do this.) I measured the heater block at 200C (idle - Yes I know this is not advised) then measured the 'Extruder heat sink' as soon as the thermistor detected it has reached '200C' and it (the extruder heat sink) was at 91.7C. After 5 mins it got to 110.8C. 10 mins 119.4C. At that point is where I sent my email to SD support.
This is where elmoret comes in. I found a post that was talking about just replacing the stock hotend to an E3D. Most of the folks who offered to do the CAD work did not have a Press. So I offered and v6 showed up at my doorstep in about 2 days.
The E3D v6 is a nice ass hotend. I've never used one before. I've been meaning to get one for my SD2 but didn't really have the finances or the time to pull that off. The E3D website has a very nice PDF with the exact dimensions of the hotend. And with that I was able to recreate it in Fusion 360 (ULT). The next step was to model up the "L" bracket for that mates the stepper to the gantry and the hotend. Now equipped with these models I've measured the exact length of the stock hotend (mostly where the tip is compared to the bottom of the gantry) and move the E3D there. Which places it front of the extruder stepper. The initial thought was to create a 'booster seat' for the stepper mount & the hotend. But then it collides into the lid. So the bowden config was born. Move the business end to line up exactly where the stock hotend is and designed around that.
After 17 modifications and 9 prototypes. It was ironed out and printing perfectly. The next step was to lower the hotend so it lined up with the extruder stepper for the non-bowden config. That worked out pretty well. I printed one out to see what it would look like and also confirm the new offset. Using those measurements I was able to make a booster seat for the probe. The issue was the tiny pegs that hold the microswitch in place (It's 100% a friction fit) So I needed to make something that fit into the existing mount (as if it was the microswitch) create the extension then create a receiving end to remate to the microswitch. My biggest hurdle was printing the tiny pegs. In which I scrapped that and used small nails cut to to length. Now with the increase length- The probe was not deploying straight. It varied from 1.2~2.6 degrees in either direction. This is mostly how the probe was designed to begin with. It's similar to how dog collar clips work. And the channel is rides has a bit of slop. So lately I've been working now to to replace the probe altogether. This is the main reason why the STL's for the mount is out but not the probe.
As for the prints with the v6- They were awesome. Super clean. Super sharp. They look alot better then my SD2. When I have a chance I am definitely going to get one for it too.
Heat Bed.
After installing the hotend and noticing how quickly it gets to temp. The bed up-to-temp speed was more noticeable. The hotend would get from ambient (26.7C) to 250C in just under 4 mins (3:41 to be exact) and the stock heatbed took about 11 ~ 19 mins to get to temp. So I began looking around for options for a replacement. I found the MK2B PCB Heat bed on amazon for 19.99 (free shipping if you are a prime member). For me that was cheap enough to get and experiment with.
It arrive on a Saturday @ 1pm and I was printing on it @ 1:45pm. The version I got came with the thermistor and the MK2B has a hole in the center of the bed that touches the sheet of PEI I had on top of the stock head bed. I used some 468P cut to size to stick the sheet of PEI together. Now the get-to-temp times about just as fast as the hotend (4:27).
**Edited - > The orientation of the mk2b was with the LED facing downward so that the traces will make direct contact with the sheet of PEI.
In my case, I reused the existing connectors by decrimping and recrimping then heatshrinking it. If you are not savvy enough to do that. Get new connectors.
PEI.
Get it. It's awesome. If you don't know what it is- Google it. Where to get it? Amazon or Google it. I used a sheet that was 0.06" thick. Then used 1.0mm spacers to bring it up with where it needed to be.
The Press that should have been...
After all the work listed above I feel that this is the Press that should have been released. Granted understand that the parts in the stock unit were chosen was to keep the price down. And corners were cut. And no fucks where given in the QA Dept. (Every screw was loose and not one panel was straight on my machine.) But thats the woes of an early adopter. On the flip side of all that, The price of entry (I was able to get in on the $349 pre-order) made all of this 'upgrade' feasible. It was enclosed, the z-offset function and the print area was 200mm cubed. So I am a little torn. Absolutely the 'Plug-n-play' banner should be removed from the Press (and all their printers for that matter) I think thats sending a confusing message to the newcomers to the 3D Printing world. As they don't know what to expect and how to troubleshoot something that actually considered relatively normal.
Moar Powah...
The next set of items I have planned for the press is changing out the thermistor for a thermocouple, creating an in-line filament diameter sensor, further work on the ring of addressable LEDs and a power supply upgrade. (The stock PSU gets pretty damn hot and it started making a high pitched noise yesterday.) I should mention getting the PWM port to behave also.
SD2 / E3D v6 / Direct Drive / Mk8 Gear / RAMBo v1.3 / PEI Bed / Anti Z backlash mod / Ikea Expedite enclosure.
Both Driven by Octoprint (devel) via a RPi B+ / Neopixel status alerts / GPIO Controlled SSR / (SD2) 450 Watt PSU for remote power-up/shutdown.