McWire unexpected behaviour

First the good news.
I finally have my McWire RepStrap completely integrated and running.
All axes run all the endstops are set and behave as expected, The extruder extrudes.

Now for the odd news.
In the home position for all of the stages the extruder nozzle just clears the X,Y stage and every thing is as it should be, when I try to print anything,all of the stages home (seems right) Then they X and Y index to ~X,Y offset seems right, then the z stage lifts up about 3 1/2" (Seems odd) the extruder comes up to temp, prints a little line in the corner and proceeds to print the stl. Did you notice that the print nozzle is ~ 3 1/2 " away from the X,Y stage???
my actual working volume is about X=9", Y=6", Z=6".
What I have set in the host preferences is ~200, 140, 200 this is because any value for Z below 200 trashes the program. The gray background turns black and while the tools work I can't manipulate anything on the work space.

If I set the workspace to Z=300 (Default) instead of the z stage raising 3 1/2 inches it is still raising past Z max when I kill power.

Who has a bright idea of what might be going on?

N.T. we really need a separate set of preferences for motor holding torque VS running torque. I have to use 75% torque to get my stages to move, at this holding torque the stepper driver is getting way to hot.

All help is appreciated.

Best,
Dan

Dan,

Where is your z axis end stop connected ?? it sound like it is right. Do you have a jumper on the max end stop or a max and stop wired up? you need a jumper on the 2nd and 3rd pins if no end stop You also need sync wires connected on all the stepper boards. (this is not related but you will find if you do not have then that you will..

I agree with the torque setting mine get hot also but I have not had any problems the driver chip will take a lot.

Bruce W.

Good luck!!!!

Bruce,

Endstop is connected to K2 and works fine just like the X and Y.
Z endstop stops the stage about 1/2mm from the stage (this is adjustable).
both sync lines for X,Y,Z are all wired, even the unused one.

The device is doing everything exactly as it should except Lifting the extruder away from the X,Y stage. With no HDPE installed and with the extruder temp set to 25C I have simulated the building of the testpiece and the minimug. Both performed as expected except for the extruder being nowhere near the stage.

Best,
Dan

Edited 1 time(s). Last edit at 12/03/2007 10:55PM by Dan Putman.

Problem found,
There is a separate extruder offset on the extruder prefs page.
The default is 500mm.I set this to 0 and problem solved.

Dan

fantastic! let us know how things are going and hopefully post pix / etc when you have some news to report!

Now that the all of the systems are working, I decided to try and lay down some HDPE.
Using the best info Forrest's foam board (works great), and nophead's temperatures I was able to make a repeatable mess. So here is the next McWire RepStrap problem. McWire uses 1/4" 20 threaded rod for motion on all 3 axes. I am using 400 step steppers the conversion to millimeters is ~ 314.96 (400 steps/revolution * 20 revolutions per inch./25.4) Darwin's X and Y stages are 11.4855 steps/mm. McWire is about a factor of 27 slower. The current problem is that I can't extrude material slowly enough (motor stalls) to match the maximum speed of the McWire X,Y stage. Adrian was just working on stepper speed control so I am hopeful that there is a fix. I am currently running extruder speed =75, stepper speed =240. torque =75. The steppers stall at any higher speed, I haven't cranked the torque all the way up to see if I can get the speed up, but I don't think there is a factor of 27 lying around. I will post pix once I clean up the wiring and pwb board mounts. I have a few more experiments to try to see if there is a sweet spot.

Best,
Dan

Dan,
Sound like you are getting closer Way Cool!!!!


I have found that using Forrest's foam board and then taking a HDPE milk jug cleaned out or a gallon water jug and cutting out a small piece and gluing it to the foam board (i have used a glue stick) has helped even better.

I wonder if changing the firmware to be 400 steps would help you out. now it is only stepping at 200 steps even though you are using 400 step motors.

Bruce

yeah, hopefully you can get that speed up. from what i've experienced with my mcwire machine, with the steppers moving at 200 RPM it definitely moved the stage fast enough to keep up with the extruder. i imagine some tweaks are going to be required to the code to get that going.

I have been exploring the speed envelope of the current firmware and host software.
I have been able to get stable motion at a speed setting of 249. This moves the stage at about the right speed for printing ~3mm second. If this was the behaviour while printing I would be a happy camper, of course this is not the case. When I print the testpiece or minimug the stage movement is orders of magnitude slower. The test segments on the testpiece which run parallel to the Y axis move at the proper speed. As soon as the extruder begins printing the real article things get very slow. Since it is clear that the steppers can move much faster I am guessing the lack of buffering between the host and the stepper firmware is at least part of the problem. I have not found any prefrence item that has any impact on this. Any thoughts?

Best,
Dan

Ok, so now that I have had some time to think, I have concluded that the simplest way to fix the X,Y movement speed issue is to replace the leadscrew drives with belt drives. It is actually pretty simple, but still a days worth of work. The Advantage is that other than build volume the McWire RepStrap would then be a very close proxy of a darwin. I have attached a simple drawing of a modified McWire X stage to make it easier to see what I am thinking of doing. Of course if darwin parts are going to be available in the immediate future, I would cannibalise My RepStrap and build the real deal.
All thoughts are welcome.

Best,
Dan

yeah, it is definitely possible to have belts power the mcwire repstrap... once i get it up and running and fully documented, that would be an interesting hack to apply to it. with some of the new improvements i've gotten from Bre, it would probably be pretty easy to attach everything correctly.

if you want to give it a go and document it, i'd be grateful!