there are a couple of versions floating around.

the stl attached is basically the stock printed part - it does use a brass threaded insert, but you could just as easily use nuts...

and there is this version on thingiverse, designed by one of the members here: http://www.thingiverse.com/thing:550713

My knob looks the same, but my bracket is a thin metal thing that is kind of wobbly.  I guess I'll design something better in Sketchup.

I find 123D from autodesk easy to use and it is also fee.

http://www.123dapp.com/design

knowack - his Z knob looks exactly like what came on both of my SD4s stock... printed bracket and knob...

My knob looks the same, but my bracket is a thin metal thing that is kind of wobbly.  I guess I'll design something better in Sketchup.

Maybe I am missing something but would not a single motor be more accurate? I would think that with two motors even a slight difference in timing, belt position, and other variables could cause all kinds of issues. With a single motor linked to geared rods the only issue is the initial setup and timing.

carl - there really are no "timing" issues running 2 motors from the same stepper driver... Prusa variants have done it on the Z axis for a long time. The FolgerTech FT-5 large format has dual motors on both Y and Z with no issues.

The only "issue" to be aware of when doing dual Y motors is the actual wiring - one motor needs to turn clockwise, the other counterclockwise. This is accomplished by flipping the wiring of one motor when connecting them together. Pretty easy actually.

rsantos - nice job with the conversion. looks nice and clean.

knowack - his Z knob looks exactly like what came on both of my SD4s stock... printed bracket and knob...

I noticed your Z-knob modification...is that an original design, or did you find it somewhere?  I'm running the stock metal bracket for that knob, but I want to get something a bit more accurate.

Maybe I am missing something but would not a single motor be more accurate? I would think that with two motors even a slight difference in timing, belt position, and other variables could cause all kinds of issues. With a single motor linked to geared rods the only issue is the initial setup and timing.

Geared through a belt that introduces elasticity?  If it was actually geared through solid gears I would tend to agree but the Y belt system is exactly what have introduced half of the issues with this printer design compounded the by the SD's rigidity issues - plus ...  bad choice of belt MXL vx. GT2, loop belt elasticity, frame elasticity, rod elasticity (if not perfect aligned) and what we got was a rubber band transmission.
First it would be helpful to understand how stepper motors work. Each step should be physically exact on both motors by design. Yes, there could be problems in high torque situations that of the motors can skip a step bit but this is extremely unlikely for these running such a relative low mass (I have the same spec motors running a mass on my MPCNC on a much higher load).
Anyway, after many small and large prints now can report this printer now is actually printing extremely reliably and accurately with this mod, to the point that my prints on the SD4 are actually better than my Prusa I3 Grabber variant (and I was not expecting that).

brucesdad13 wrote:

Very cool STL of CAD files for all of these parts? Especially the unibody lawsy X carriage??

the link is at the beginning of his post

That's correct. For the unibody I was testing my printer in high speed yesterday and noted one of the bearings shifted outwards a bit. This never happened in my I3 but just in case I have just updated the file with added holes to place an M3 screw with a washer to constrain the linear bearings. On mine I have just drilled the holes . Anyway the link on the top has the updated STL.

Very cool STL of CAD files for all of these parts? Especially the unibody lawsy X carriage??

the link is at the beginning of his post

STL files here:
http://www.thingiverse.com/thing:1914991

1) New -- Eliminated Y rods/belt transmission completely - removed all previous parts and replaced with a new Nema motor - to to this I created a drilling jig and a matching NEMA motor support to help position the mounting holes.  The original motor was reverted 180 degrees.

Also

2) Replaced all MXL belts, pulley and idlers with GT2 (16 tooth pulley count for added precision - change steps per mm from 88 to 100 using Repetier direct Eprom configuration) 
3) Replaced X and Y rods with chrome hardened steel -- Way smoother and stronger than stock.
4) Replaced X/Y carriage with Lawsy's carriage derived design but modified to:
      - Be unibody (to avoid potential rod misalignment during assembly (and reduce the need for screws).
      - Be compatible with Solidoodle 4 stock extruder mount (3 holes on the aluminum bracket).
      - Add micro switch support
      - instead of snap fit, I made the linear bearings mount from the sides into a half cylinder to avoid potential loosening that could produce yaw movement. I had this happen in my Prusa I3 Graber after many prints and found this design to be a lot more resilient.

5) Mounted bed Y front stabilizer - [add link to design]

BTW... I liked the direct Y motor drive idea but that would not allow me to keep the nice Soliddodle 4 covers on (since the single drive motor has to sit outside of the case) so I decided to create this solution. The 2 motor design is based on many Rep-Rap's  Z drives and and the X and Y drives used in the MPCNC design

Printing a perfect cylinder:

Drilling jig sitting behind the original Rod (that will be later removed). This picture is rotated -- YOU ARE LOOKING AT THE RIGHT SIDE -- (I will fix it soon) . Also the bracket drawn on the metal with a sharpie is to approximate the original pulley position. 

Wiring - Note the reversal of one of the coils wiring pairs to accommodate for the second motor. A good explanation with great pictures and a diagram on this technique can be found here:  http://www.vicious1.com/assembly/wiring-the-steppers/

Left side motor.

Right side detail now showing the SD4 "filter" fan re-positioned. To do this you will need a special drill. This filter is probably useless anyway so I believe that is totally optional. You may leave that fan out.

Left side detail. Note that I had to put a very small bumper (in black) attached to the side of the Y carriage to activate the Y micro switch before it hits the NEMA mount. It is only about 3mm thick.

Finally, here is the detail of Lawsy's carriage redesigned as unibody. Printing is a bit ugly as I did it on .320 mil layer height and 90% infill. Supports are essential for the top overhangs of the cylinder since I printed it horizontally to reduce potential flex in the Y direction.