[PsyKotyk]

I was getting seriously frustrated at trying to get the PICStep gerber file into Eagle for milling purposes, so I did my own rewrite. Similar connections, with a small modification on the motor output terminal block (for unipolar convience)
Also, the current limiting oscilators have a 10turn pot for adjusting.

This is sort of a mash up of Alan's awesome design and the equally excellent EAS design. For firmware, use Alan's code.

BE WARNED: I have not built or tested this design yet. It *should* be okay, but who knows. No warranty implied or expressed.

Included are the Eagle library files for the LMD18245 and PIC16F62x. They need to go into the eagle "lbr" directory.

If you make improvements or corrections, please post them back here on CNCZone.

[pminmo]

Nice job! If your wanting to be easily compatable with the open source stuff, SV1 is out 180 degrees. http://pminmo.com/interface.htm Still works you just need to remember to twist the cable once.

[pminmo]

OOP's you have a couple of gotcha's. DAC ref, Capacitor physical sizing on the electrolytics maybe an issue. 78L05Z issues, input voltage, power.

[PsyKotyk]

OOP's you have a couple of gotcha's. DAC ref, Capacitor physical sizing on the electrolytics maybe an issue. 78L05Z issues, input voltage, power.

Ahh good points. I'll drop the regulator all together. The interface board can deal with that.

[pminmo]

It's a good idea to handle the 5V, but use a to220 sot223 or to252 package.

Your 1uf package is to big and the 470 to small with at 63vdc.

[PsyKotyk]

Here's a rev with the 7805 removed. I haven't resized the caps yet, as I'm still waiting for my parts to arrive, at which point I'll take some real world measurements.

This version includes Alan's latest firmware (2.0), the GPL license and a README with the various "may asplode in your face" warnings.

[PsyKotyk]

I've gone ahead and changed the capacitor sizes to match the 390uF 63V ones I got from b.g. micro.

Also made the traces thicker to aid production methods like milling or glossy-paper + toner transfer.

Still haven't done a trial of this board yet, so builder beware.

[PsyKotyk]

Moved the input signal connector to allow more space near the board mount holes. Also allows for keyed connectors.

I've finally gotten a board made and all the parts, so we'll know real soon if it's a keeper.

[PsyKotyk]

Here's a working revision. I strongly recommend not building any previous revision. The 20K pot in parrallel with the 6.6K resistor is dangerous. I've changed them to be in series which should protect the LMD from going beyond its limits.

Had lots of issues with crystal resonance. Added a series resistor as recommended per the 16F628 datasheet. Also added a noise reducing capacitor across the +5V. This made a huge difference in how cleanly it ran, and at what speeds it would step.

[PsyKotyk]

New release with some cosmetic board changes. This ensures that all traces are separated by at least 1mm (for easier milling with a wider bit).

Added a few more fill areas to save on etchant/milling time (if you enable that).

The picture below (not to scale, so don't use it directly for making a board) was touched up in Gimp, than tiled so there were 6 on a sheet, and then printed back to postscript, ready for toner xfer.

[opcode1300]

im new to all of this. but do you have a parts list? digikey or mouser?
thanks a lot.

[PsyKotyk]

im new to all of this. but do you have a parts list? digikey or mouser?
thanks a lot.

I used to until my laptop hard drive decided to take a dirt nap yesterday. (chair)

I'll try to put one together this week. A lot of them you can find off the EAS website (see the first post in this thread). I picked up the capacitors, crystals and some other parts from B.G. Micro for quite cheap. They also have the PIC's for cheaper than Digikey, but are sometimes back ordered.

[PsyKotyk]

Just a final follow up. I tried using some different crystals I found at the local Computer Recyclers (2 for $3 yay!), and things started working perfectly. I must have been using some non-parallel cut x-tals. These ones were in the tall can instead of the short one.

The resistor in series with OSC2 can be ommitted completely, so either jumper it, or edit the schematic/layout, or just cover the gap with a sharpie if using toner x-fer.

I had a successful test at 52V (the stepper was certainly moving quickly), but when I touched the heatsink to check the temperature, I felt a definite surge of voltage. I thought the heatsink was supposed to be GND?

Anyone tried liquid cooling for their heatsinks?

[opcode1300]

when i get my HF mill up and running i was gonna run some copper waterblocks to get it setup and to help w/ cooling.
but i keep smoking my xylotex board, so i gotta get some diffrent motors to try.

[tfmacz]

I thought I would share a couple pictures of My PICStep Controller.

I am a fan of pcb milling and cut one of PsyKotyk's boards except I put the 5V reg back on the board as my system needs it there.

I am running a tripple stack NEMA 23 stepper at 2.8A so need the big heatsink.

Thanks Alan for the code to make this run.

[tfmacz]

Here is my version of Alan's board Have a working board. Works great.

I am salvaging 20mhz crystals off of old 10meg ISA network cards. so far I have not had any that did not work.

[Xephius]

Hi all. I have all the components to build 3-4 boards for my new CNC machine. Would anyone be willing to knock out the PCB's for me? I am happy to pay for them, and I know that I can buy the PicStep boards (probably the same price or cheaper) but I would rather work with Open Source works. The project looks great, I hope I can be part of it.

-John

[pminmo]

Which version? I'd be willing to as I've finally got my pcb cycle pretty well tuned. Plus I could add the code to my website.

[Xephius]

PMINMO,

Thanks so much. I would like to use PsyKotyk's latest board revision. Using a external 5V regulated powersupply is not a probelm for my setup. You can contact me directly at jc (at) xephius (dot) com. I look forward to working with you.

-John

[sdantonio]

Here is my version of Alan's board Have a working board. Works great.

I am salvaging 20mhz crystals off of old 10meg ISA network cards. so far I have not had any that did not work.

Hi,

I just got my CNC router finished as of Sunday night and I'm already looking at modifications, such as new driver boards. I was looking at your set of gcode files. Could you elaborate a little on how you would go about setting up a board for routing, registration points, which NC files to use and in which order. Size of router bits and drill sizes etc. Hopefully, with a little bit of explaination and if I look at things long enough then it will begin to make sense.