[NI2O]

Hi folks,

Firstly, thanks for reading this post.

Secondly, I'm an absolute noob at CNC. I decided that I wanted to mill my own PCB's to try and cut development costs. I'm a radio ham not a machinist.

OK, So I have the dratted 3020. It's controller box says it's a T-D (whatever that is). I'm a Linux user so have got basic control of the machine via EMC/Axis. I've also been able to get the ESTOP button working as well as move some wires around to turn the spindle on/off. None of this was rocket science. Indeed the PCB's in the controller box have well labeled headers allowing me to make the few wiring changes.

I have been able to mill/engrave some demo items which came out OK. However, as soon as I try a PCB I get disastrous results. The Z axis seems to plunge far too deeply and no matter what I do I cannot make it plunge less deeply. I've also managed to break the 5 bits the machine came with too.

I'm using Eagle to design the PCB and have the PCB-GCODE ulp script to produce the GCODE.

Clearly I'm doing something wrong. Any ideas?

Thanks

Mark

[NI2O]

At the risk of replying to my own post. I _think_ I have it sorted out.

I think the biggest issue was that I don't know what I'm doing.

Would I be correct in thinking that there are various models of the 3020? I've noted that there are a few "howto" articles for it around the net and none of them seem to work in my case (although they are a good starting point). Indeed, even the controllers vary from unit to unit.

Upon closer examination of my controller I have discovered that it has a A axis as well as XYZ limits (or home?) too. I've not been able to get the PWM working for the spindle but can control the spindle on/off now and have also made a few "improvements" such as grounding the entire rig correctly and also grounding the power switch so that it illuminates. I can now see the condition of the controller from across the room.

If there's interest I'll write up my "modifications". I took pictures so as to be able to roll back if needed. It would be somewhat trivial to write them up.

Thanks for your indulgence!

This one is now "closed".

Mark

[hankus1942]

At the risk of replying to my own post. I _think_ I have it sorted out.

I think the biggest issue was that I don't know what I'm doing.

Would I be correct in thinking that there are various models of the 3020? I've noted that there are a few "howto" articles for it around the net and none of them seem to work in my case (although they are a good starting point). Indeed, even the controllers vary from unit to unit.

Upon closer examination of my controller I have discovered that it has a A axis as well as XYZ limits (or home?) too. I've not been able to get the PWM working for the spindle but can control the spindle on/off now and have also made a few "improvements" such as grounding the entire rig correctly and also grounding the power switch so that it illuminates. I can now see the condition of the controller from across the room.

If there's interest I'll write up my "modifications". I took pictures so as to be able to roll back if needed. It would be somewhat trivial to write them up.

Thanks for your indulgence!

This one is now "closed".

Mark

I'd be interested in more information regarding the grounding issue and spindle control. Do you have any references you could share.

[NI2O]

The grounding "issue" on my device was that there was none! Both my controller box and the rig itself were not grounded.

I first tackled the controller box. It would seem that the manufacturer had made a lame attempt at grounding by connecting the ground wire from the mains input to a standoff screw but the box was painted inside and out thus preventing the flow of electrons. Fixed by removing the standoff, rubbing some emery cloth on both sides of the case around it's hole and then replacing the standoff/screws/earth wire. Was then able to run a ground wire from the GND pin of the power switch to the case thus making its light come on when powered on. Finally, I installed a nut bolt sticking out of the back of the box to be used as a ground contact later on.

As for grounding the rig itself, that was pretty easy. I'm a ham radio operator and so have lots of coax kicking about. I pulled the braid from a piece and then using a small drill and some self tapping screws I connected the gantry to the bed on both sides and in turn ran some braid back to the nut/bolt on the controller box.

I did NOT make any effort to ground the Z axis spindle as the +ve of the probe will attach to it and so may trip the probe incorrectly. In my case the spindle did not have continuity with the gantry. YMMV.

Closer inspection of the controller PCB revealed that the 2.1mm power socket was not in fact an alternative power input as stated in the "instructions" but rather the probe input. It was labeled as such on the board's legend. Inspection also revealed negative going X, Y, Z & A limit inputs. I have yet to do anything with them other than bring them out to another CB socket on the back of the box (the hole was already there covered with a cap). My thinking is that I'll add some switches which will short to ground via the frame of the rig.

Finally, the spindle can be turned off and on from software. Pin 1 in my case. Whilst the controller had a manual speed knob the board has fittings for variable voltage (PWM??) but I've had no luck getting that going.

As stated in my original post, it would seem that there are numerous controller boards all housed in the same enclosure and so whilst my mods work for me they may not work for you if your board is different to mine.

[hankus1942]

Excellent. Thanks for the summary. My probe and XYZ limit inputs are pretty much the same as you found yours. Got my probe input functioning okay but the PWM issue is a work in progress, with a little more study required. I also thought about doing the ground strap, and think I'll do the same after reading your reponse.

Thanks for the timely updates 73's Hank - K7CHR

[hankus1942]

Just a follow up to this post. The modifications suggested by NI20 are 100% for this particular controller. I found the same grounding problem. Added the ground wire to the power switch to illuminate. The ground for mains is currently through the screw holding one of the rubber feet. I plan to follow his suggestion and add a ground stud to the back of the controllers case, and move my mains ground there as well. Now on to the spindle control PWM issue.

I didn't mean to highjack the primary purpose of this subject. I have yet to cut my first PCB, but I did find that the suggested steps per unit were off for the z axis, when I ran the axis calibration utility in Mach3.

[NI2O]

My apologies for restarting a thread that's almost a year old. I've been busy with work and family and had to put CNC to one side.

There's a axis calibration tool in MACH3? Interesting. I'm a LinuxCNC user so have not played with MACH3. I have an old PC that I could get XP going on and perhaps test stuff out.

Since my last post I've acquired a 3D printer too. It's still in the box but I suspect it'll have ll the same issues. I'm also building an SMD reflow oven from a toaster oven I got on sale at WalMart for $18. Got lots of stuff to keep me busy in the shack!!

I have acquired a new set of bits and so am ready to have another go at this project.

Mark