Repstrapping with a Sherline mill

After a brief but enlightening discussion with Sebastien on IRC last night, I have decided to try my hand at repstrapping. I currently have a Sherline mill, and I will try converting it to CNC, attaching an extruder to it, and seeing what happens.

I have the following questions:
1. Are the extruder head STL files in subversion current/workable/good? If not, where can I find a good set of extruder files?

2. Does the RepRap control software have some flexibility for the types of machines it can control? For instance, provided the steppers used in the Sherline CNC conversion can be controlled by the RepRap controller boards, can the software be adjusted for the Sherline's parameters without too much trouble, or is everything hardcoded for Darwin's dimensions and parameters?

3. If the answer to #2 is no, can someone lead me in the right direction to learn how to translate the FDM concept into something the Sherline CNC controls can understand? I'm not terribly experienced in programming nor CNC, so I'll probably need a good bit of basic education or a mostly pre-built solution.

I have some time off coming up at the end of this month, so hopefully I can get all the parts I need next week and be ready to start banging on things next weekend. I'll try to remember to post some pics as soon as I get anything together that's worth photographing.

Wish me luck, and hopefully in a couple of weeks I'll know whether I want to strangle Sebastien, or owe him a beer/coffee/beverage-of-choice for putting this idea in my head.

-Eric M.

Eric,

I'm planning to do the same. I have a Taig CNC Mill which is very similar in size and function to the Sherline.

You cannot use the Electronics on the Sherline or Taig to interface with RepRap. The RepRap uses a serial port to interface and the CNC Mills use the parallel port using step/dir commands. You will need to make the electronics as per the RepRap project, however you should be able to use the Stepper motors on the Sherline as long as their current is less than 2 Amp per phase which is usually the case for that size of Mills.

Take a look at my post a few days back on the General section. I'm probably a couple weeks ahead of you. I have the electronics almost ready and I'm working on the Gcode to make the extruder out of aluminum. I'll be milling a few of the parts this weekend. I'm posting my progress on my webpage www.luis-freeman.com and I'll include the Gcode there.

to answer your first question: yes, the extruder files are up to date in subversion. however, i would suggest downloading the files from sourceforge... they are nicely available in a .zip file. it includes both the parts for the extruder and for darwin.

if you are using the RepRap boards to control the steppers, it should be no problem. follow luis' advice.

as for the RepRap software, it should be versatile enough to handle the various parameters. if it cant for some reason, its something that will need to be fixed anyway. hopefully one of the developers will be able to take a crack at it.

Luis - Unfortunately I haven't been able to connect to your web page, but I am eager to see what you have done so far.

And now another question: Please excuse my electrical ignorance, but the RepRap stepper motors appear to be 6v motors, and the ones for the Sherline CNC are 3.2v (http://www.sherlinedirect.com/merchant.cfm?pid=125&step=4). Is there anything I need to change on the stepper controller boards so that these motors will work and/or won't get fried?

This actually elicits several tangentially related questions:

Has anyone ever tried making a FDM machine by bolting an extruder head to one of these Mini CNC mills?

Would this contraption be more precise than a standard RepRap or is the impression coming from the extruding head & process?

Wouldn

I had a look at EMC2 yesterday and my undestanding is that it needs G code tool paths as input. I.e. you still need a CAM program to convert from 3D model to toolpaths. This is the main function of the RepRap Java code. I.e. it slices the 3D model and works out the extruder path. It also does the job of EMC2 in that it controls the machine, but some of that work is delagated to the PIC microcontrollers.

If you wanted to drive stepper motors from a PC parallel port then I think you would need to modify the RepRap java to output G codes and then use EMC2. You could modify the RepRap Java to drive steppers on a parallel port but IMHO you would have real time problems.

---

[www.hydraraptor.blogspot.com]

it would definitely be worth adding Gcode output to the host software. if we did that, then you could open the object, generate the paths, then use your standard control software to control your repstrap machine. we're having some trouble with the java host software, and we're looking for someone to really step up to the plate and make it work well. i dont want to beg, but is there anyone out there that wants to take the challenge?

I think the most flexible scheme is to split the RepRap Java (BTW does the program have a name?) into a slicer and dicer and a separate machine controller with an XML file in between, i.e., the scheme Forrest uses. It should then be possible to convert the XML to G code with an XSLT transform or a Python script.

I am not sure which route I will take for my machine. I can either hack the comms part of the RepRap Java to talk to my machine over Ethernet or I can hack it to output XML and use a Python program to read that and control the machine. Either way I hope to avoid G code as it looks horribly antiquated.

---

[www.hydraraptor.blogspot.com]

Eric,

To the best of my knowledge, it doesn't really matter what voltage you apply to the steppers as long as you don't go above their current limit.

I'm using a xylotex controller on my mill and it's feeding 24V to the steppers limiting the current to 2.5 Amp. The motors actually are rated at 2 Amp however so far no problem of overheating and it's been already 2 years since I got them. My steppers for your reference are Vexta PK268M-02A. Resistance of the winding is 2.25 Ohm.

I'll fire up the electronics this week and will see how it goes. By the way... my server should be up an running now... I fixed it this morning.

One last think... I have about 70% of the Gcode for the Extruder finished. I'll post pictures once I get the first parts milled.

Eric,
Looking at the stepper driver schematic it appears to be a constant voltage drive rather than constant current. The driver chip will lose about 5V at 2A leaving 7V for the motors.

This means that you will need to limit the current in software by using PWM. I am not familiar with the PIC firmware because I am not using it but I think it has a PWM capability.

---

[www.hydraraptor.blogspot.com]

Thanks for the replies - I think I'm starting to understand, but I'm still a bit confused...

Are you saying that the voltage doesn't matter as long as the current isn't too great? I'm used to circuit boards for which a too-high voltage makes things go *FZZT!* fatally. If I am understanding this correctly, the combination of 7v and 2A going to the motors is too great for sustained operation, but since the motors are happy at 3.2v and 2A, if I use PWM to send the current to the motor roughly only 50% of the time, they should be OK? Did I get that close to correct, or am I still off?

Eric,
The short answer is yes 50% PWM should be about right.

In the steady state voltage, current are resistance are related by Ohms law: I = V / R. So your 2A 3.2V motors must have a resistance of 1.6 Ohms. If you connected them to 7V DC you would get over 4A flowing though them. Not only would that overheat the motor but it would more than likely fry the drive chip.

You could fix it by putting a 1.9 Ohm resistor in series with each winding but they would dissipate 7.6W each so you would need big resistors on a heatsink.

When non steady state is considered, the coils also have inductance as well as resistance. This means that when you first apply the voltage, the current is zero and builds up at a rate of V / L. Similarly when you switch off it decays slowly. This means that if the PWM is switching fast enough you get an almost constant current proportional to the mark space ratio times the steady state value.

Be sure to use a big heatsink on the driver chip as it will be running close to its max rating.

The best way to drive motors like this it to drive them with a much higher voltage, say 24V and monitor the current until it reaches the target value of 2A and then switch off and let it drop a bit below and then switch on again and so on. This is a constant current chopper driver and achieves better torque it high speeds. The reprap board does not have a current sensor so your stuck with open loop PWM or a series resistor.

---

[www.hydraraptor.blogspot.com]