Cam GCode produces 2 identical toolpaths

I have finally finished the compact CNC Plasma build and after a load of faffing around, got the axes working perfectly.
My test profile is a Solidworks drawing of one of the stepper mount plates. Bringing in a dxf resulted in some odd inverted curves and the main hole in the centre had a gap in it (which is not on the model) so I brought in a jpg and created the toolpath in one operation.
transferring the G code to the machine results in a correct geometric cut pattern - but it does it twice. Eh?
There are no chamfered edges and the drawing and jpg are 'clean' ie no double edges/lines.
What is going on?

 

No As it is solidworks I can't save an autocad R14 version. I saved as jpeg and used that, but the jpeg only has single ines?

 

Ah. I have QCad - I will try and report back. thanks

 

Well, tat turned out the same as the dxf I exported from Solidworks ie:
I could not find a way to convert to polyline in Q Cad.
The original drawing is just fine

 

Hi Peter. Well the outside shapes and arc converted to polylines but the circles won'r - I selected a circle and went to draw/polylines./ make polyline from selection and it does not change from circle. Hmm?

 

Here you go Peter.
Thanks

 

Hi Peter
Did you manage to turn the circles into polylines?

 

Strange - I tried each individual circle rather than selecting the whole drawing. Hey Ho. thanks

 

No - not working. i imported this dxf into Qcad, dragged over everything and converted to polyline and it does everything but the circles!
What am I missing - your dxf shows all as polylines

 

Well this is what I get!

 

Ta Daah. Fixed!
Thanks Peter
As an aside as i haven't tried it on the machine (machine in workshop without internet so I have to generate Gcode here and take it down) the Z axis setup is slightly wrong. I set Z+ as down and Z- as up. and force a zero 5mm in the z- direction. The carriage raises to the safe point (near the top) then when it wants to set the plasma it drops down (to what would be the plate surface) but instead of raising 1 mm for the pierce start, it drops a further 1mm ie. wrong direction. If I put a -number in the pierce will it raise instead of lower?

 

Having sorted all the other minor bugs, fixed polylines in the dxf files and wired all the endstops I have come across a weird problem that only occurred after I finished installing the panel in the box.
The X axis was vibrating which was caused by a wire coming out of the connector block. With that fixed everything seemed normal but the offending axis only moved in the + direction. Clicking the - jog direction moves the axis, but in + direction even though the readout indicates it is going in - direction. I killed the power, pulled the plug and checked the connections and replaced and..... spot on + and_ directions working. A little after that I closed the box up and.... Only + direction again.
I have opened the box, checked the wires and made sure there is no chance of crossover (I am using 2 twisted pair cable) . The axis moves smoothly but only in + direction so I am not sure where to look. I loaded a job and it does the same.
Any ideas?

 

It is a blackbox. I did have a wire pulled free and the motor just hunted. The motor now runs free but only in one direction. Weird.
i have the door open and the wire is not under any strain. As you say, if there was a short I would get an alarm and overload.
I will play with it again this afternoon - I will check the coil resistances too.

 

Multistrand. I have a connector next to the motor. I'll check that

 

Ok. So today I checked the resistance across the coils - at the plug to the Black Box. Both the same 2 Ohm from memory.
switched the machine on and went to jog - The axis started moving in - direction and within a second switched to + and now only goes + but smoothly.
I'm puzzled. if I had damaged the output stage somehow the axis would not move at all?
Not sure where to go from here. Richard