Return of the POS

[rbertalotto]

Damn! For a Grumpy Old Guy you sure know this stuff! I wish I new 'puters" like you do. I'd love to dable in CNC.........Go for what you did for your kids. Hope the appreciate what a great Dad they got!

[Glug]

Looking good! Do you think you'll have any issues from RF interference due to the close proximity of the VFD to the stepper drivers and cables, in the same enclosure? I've read that some folks have. Operating an AM radio as a detector around my VFD suggests it does generate a lot of noise. It try and keep it away from the CNC bits, but haven't enclosed it yet.

I'd love to dable in CNC.........

You might take a look at some of the spindex projects. With a power supply, stepper driver, and a stepper motor you can start to experiment with driving the motor. That should get the wheels turning and whet the appetite. Playing with an Arduino might also be fun for you. You can also drive common R/C servos with that, without a driver. So an arduino and $10 servo is also a way to get your PC moving stuff, reading temperatures, etc.

But don't get the dirt cheap china drivers - unless you're just playing around. It is better to spend a bit more there.

http://www.liming.org/millindex/

[MrFluffy]

Looking good! Do you think you'll have any issues from RF interference due to the close proximity of the VFD to the stepper drivers and cables, in the same enclosure? I've read that some folks have. Operating an AM radio as a detector around my VFD suggests it does generate a lot of noise. It try and keep it away from the CNC bits, but haven't enclosed it yet.




You might take a look at some of the spindex projects. With a power supply, stepper driver, and a stepper motor you can start to experiment with driving the motor. That should get the wheels turning and whet the appetite. Playing with an Arduino might also be fun for you. You can also drive common R/C servos with that, without a driver. So an arduino and $10 servo is also a way to get your PC moving stuff, reading temperatures, etc.

But don't get the dirt cheap china drivers - unless you're just playing around. It is better to spend a bit more there.

http://www.liming.org/millindex/

Good point, I wired the connections between the exposed break out board and the stepper drivers at the top with cat6 FTP cable (foil twisted pair) and grounded out all the foil connections to the main earthing points along with the BOB system ground too, if you peek at the stepper wiring, you can make out the silver foil braid because its trimmed back slightly to avoid risking a short to gnd via the shield. If it does prove a problem, I'm planning on making a rf shield cage to cover the board itself up out of some blank copper clad pcb board soldered into a cover with a ground connection to it that can be screwed on after if it needs it. After that I think it'll be ok, everything else is pretty well shielded and its being fed +5v by the pc which is shielded and smoothed. I've used this vfd previously on the shaper and it sat next to a tablet that did music duty for the shed on wifi without issues with a seperate amplifier so it might even not be that noisy. Keyboard/mousepad is bluetooth so that might stop responding when the spindle is on though. The e-stop and joypad etc on the machine will be hardwired. I'm going to connect up the original bridgeport joystick and panels into the linuxcnc. Don't want wireless issues when its all going wrong in an emergency.

Interesting spindex project too, I've been looking at mine wondering if I can add seals + a water tight motor to it and paint it to avoid corrosion, so I can put it on as a spin+burn axis on the edm or even a indexer. I have a arduino unused on the shelf, and a stepper, and a spare 5v 20amp psu left over from the edm repair that I can mount up out of waters harm. Still better finish this first before starting anything else!

[legendboy]

I recently upgraded my Boss 8 to AC servos. This is what my motor adapters look like

[DICKEYBIRD]

Dang you Brits are quick! 1st the awesome EDM resurrection in barely a few weeks and now the Bridgey? Must be the pints & pies. Good ones are hard to find over here.

[MrFluffy]

Ive been today moving the mill yet again, and I decided I wasn't happy with how the electrical cab sat in the new location so I moved the cabinet back to the right side of the machine and had to swap the monitor to the opposite face of the housing as a result. But, at last I'm happy with my workshop layout and all the controls fall nicely to hand.

Also looking at Pico systems universal stepper controller, with one of their DAC boards and two solid state relays, because it will give me lots of IO to run the various contactors, 4 axis for now and stackable for more in future, solve the spindle vfd speed control issues with the dac module, and eliminate the chinese break out board, which is probably the part I'm most unsure about being suitable. It also allows you to hook up encoders to the steppers for closed loop so you can see if it misses a step for whatever reason and flag an alert and my steppers are double ended so can have a encoder mounted on the opposite end, in that configuration they look like closed loop servo's to linuxcnc. Another factor is that Pico systems have all the linuxcnc configs to suit available as downloads on their site. It seems to make a lot of sense unless someone can talk me out of it.

Legendboy, that's very neat and what I would have done if I couldn't redrill the existing castings to take the new motor mounts without adaptor plates. Although drilling and tapping the housings seems more work, it meant less overhang on the stepper's spindle, though I suspect with newer servo motor's and their improved construction that would be much less of a concern. I inherited the steppers with this mill, so servo's were out the picture, and I'm not going to be trying to set any production speed records with it, just get a workable cnc mill.


Dickeybird, this isn't quick, its sat in the barn for a year, and I've had it sat in the workshop since may last year. I've just been occupied with other things so its about time I pulled my finger out.

[MrFluffy]

Grrr another frustrating delay. The shuttle XPC I replaced the pc with has no parallel port, which a good stable hardware one of is pretty important to this machine. But they come with the hardware for one, just not the physical db25 port. So I ordered a parallel port lead tail and it arrived friday, went to fit it and the motherboard is not a standard pin pitch.
Shuttle used 0.2mm pitch pins on their port connection (for just this one connector) and thus the standard 0.1" idc headers in use by everyone else doesn't work. They sold a special cable which is now nla, and only available on amazon for $300+, for a cable...
Some looking into it has shown that msi also did this on a select number of their industrial boards, and the msi cable is available more widely. :-

http://www.ebay.com/itm/291342585148

You can see that the "idc" plug end doesn't match the pitch of the cable and they had to do some folding to jam it in. Why they do things like this annoys me no end but now I have to wait for the header to arrive from estonia before hooking up the controls.
I'm going to get on and wire the steppers + limit switch back to the cabinet in the meantime.

[MrFluffy]

I now have 3 working parports after being impatient & making a homemade cable from a couple of spare usb connectors and a 0.1" pitch lead, and afterwards managing to get a startech pci dual parallel port card and fit it. A bit of tinkering round with the linuxCNC config which turned out to be the onboard audio causing a module issue for the parport driver, which produced a unknown symbol error in the dmesg log. I don't see the connection between missing symbols and a hardware conflict, but I disabled the onboard audio in the bios and everything booted and was happy. Something I can look into when I have more time.

I also got the cabinet powered up on 3 phase, checked the vfd, earthing, powered the vfd & pc up and also the homemade dc psu. It measures as putting out 50.1v, so I need to check my stepper drives can cope with this, if so better, if not I can move the tapping I used on the transformer to a higher voltage one to bring the output voltage back down.
Now I need to get the motor powered up from the vfd where it sits to check it works as its new to me never ran and I don't want to have to take it back off again. Starting to feel like progress though.


Next spin the spindle, then configure the break out board and check stepper details and see if it moves.

[MrFluffy]

No new photos but some progress. Figured out the BOB wiring (none included with it unhelpfully) and which parport matched with which cable off the card. Then lots of gathering of info regarding step angles and other parameters for my stepper motors.
I had read since buying them that the ratings on my first HY-DIV268N-5A stepper motor drivers are a complete barefaced lie, they are supposed to support 5amp at 50v, but there is a design fault on them and they cant supply more than 30% torque because it doesn't switch its ic into the correct mode and the 5amp rating is over the chip manufacturers rating and is a ABSOLUTE peak rating for short time periods. I wanted to be able to drive at 5amp all day long but I wanted to get it moving.

I priced up some nice gecko drives with current limiting and it came to significantly more than my budget, so decided to go with the budget but quite widely respected leadshine DM860 drivers off ebay. I'm sure these are what ARC eurotrade sells for 100 pounds apiece as companion parts to my stepper motors, but I paid 100 pound for 3 delivered. The gecko's were 200 pound per driver + shipping.

Still chinese but a lot more overhead and better spoken of & I would be using them well inside a safety margin rather than on the limit. Found a seller in Germany to avoid import taxes and am now waiting for them to arrive.
So I decided to try the DIV268N-5A's and see if I couldn't do some learning from them while waiting as maybe the manufacturers sorted the design faults out by now. And WOW, what a eye opener!

The first stepper driver exploded spectaculary when a voltage of 51vdc was applied, 1v above spec causes them to explode. A autopsy revealed no actual mechanical mounting for the pcb into the case, and no heatsink compound between the driver ic and the heatsink. Utter garbage.
So I dropped the dc output from my homemade psu to 41vdc to get it well under the absolute limit of 50, and screwed in the next one. This made the right noises, but couldn't move the axis. After some head scratching I wondered if they were just so rubbish they couldn't even move the machine so rewired the stepper motors into parallel winding configuration instead of serial hoping this would compensate for the lower drive voltage.

But nothing, then persisted, and discovered eventually that the printed on the cover instructions for the dip switch settings were completely wrong and I had them set to current limit at 0.2amp even though they were in the 5amp position.
Flicked to the opposite of the instructions for power levels, the axis began to move nicely but in one direction only then I inverted a pin in the configuration controlling direction in the stepconf configuration, and this one had its innards exit the case in a spectacular fashion also.
Finally I figured out that I had missed wiring the step dir pin up to earth on the wiring to the stepper motors. Fixed that and screwed in driver 3, and everything began to move nicely.
It became clear at this point the stepconf dip values didn't match the printed cover either, so I fiddled around to get it to 16 microsteps and was jogging the axis to check against a clock gauge set up on the X axis that 5mm jog = 5mm of travel when I hit the machine stops (no limit switch wired in yet) and once again a bang and smell of ozone was emitted from the machine cabinet after a very short delay and a 3rd stepper driver popped off to meet its maker. I suspect because it was wired in parallel and the dip set to 5amp which is outside the chip spec.

I am *VERY* glad I did this exercise on the garbage stepper drivers that were not going to be the final ones, but now I am super nervous the DM860's are going to be every bit as fragile, even though I suspect it is because the div268's were so close to their limit and overrated by the seller to improve sales. So I will have to go over my stepper wiring with a fine toothcomb to triple verify it. It was worth the $40 total for the 3 dead stepper drivers to learn this lesson and debug things a bit.
I hope the DM860's are more reliable...

[MrFluffy]

If your grappling with how to neatly terminate and join the stepper wires with a nema34 style stepper motor, this thing coupled with some connectors of choice may come in very useful if you have access somehow to a 3d printer (or use the 3dprint hubs to find one).

http://www.thingiverse.com/thing:444742

My printer is running off one to test it, and I'll be redoing the model the poster has also uploaded to remix it to mount the encoders also when I get to remaking for them.

[MrFluffy]

DM860 stepper drivers turned up today, remade cabinet mounts as they are much larger but have the heatsink of each mounted to the cabinet case so it acts a a heat dissipater for them too now which is a improvement in itself.
Set current limited to 4.5amp for now and back up to +54vdc on the stepper voltage rail and it managed 60ipm on the X axis moves. Faster and it started to miss steps, but I am happy enough with that for a first trial.
Now have limit switches wired in on X & Y axis, reused the original switches on the machine but they were wired normally open when I checked for one end of travel, so I removed, cleaned and checked them and changed them to normally closed as it seems like a more robust wiring schema as if the wire breaks, the machine will stop dead.
3d printed nema34 end caps came out great, have printed 3 already. Time to understand how the oiler is wired and get it hooked up.

[John Stevenson]

Any chance of a photo of the inside of the blown driver. Reason I ask is there are two types out there both mascarading as the 'same' spec driver but one is based on the lightweight Toshiba chip.

I'll get a picture of the ones that have worked for me but no way would I have used one on that size machine.
You have made the right choice with the drives you have ordered. My Beaver which is a tad larger than the POS was converted to originally use Gecko 210's
After 12 blew up, yup 12, and had a fault all it's life on never going back to zero I swapped to the Leadshine's which were then readily available and it's still on the original four 6 years later. Magically yhe fault with the zero also disappeared.

[MrFluffy]

I'll take one later on and post it but they are the toshiba TB6600HG's inside.
They were never going to be capable of 5amp/50v like the seller and cover print suggests, and I was relying on them delivering up to that for this, so lesson learned.
Happy to hear I've stumbled into using the leadshines and theyre better suited.

On another note, I have a ex lathe headstock set up to be a horizontal spindle for the 4th axis, I've done the physical side of it, but for electronics considering this motor setup with a 2:1 drive ratio to the headstock itself via a htd belt. Any comments if its a good idea? best rule of thumb from experience and all that of course.

http://www.ebay.co.uk/itm/1Axis-Nema34-Stepper-Motor-1100OZ-In-Driver-DM860A-CNC-Router-High-Torque-/201217825668?pt=UK_BOI_Industrial_Automation_Control_ET&hash=item2ed9845b84

[MrFluffy]

Here you go, the socketed chips in the corner are opto couplers. :-


Stepper chip itself ,note this has not been cleaned of any thermal compound, this is how they came.


Size comparison if anyone is curious.


I've just bought 3 x HG6600's to replace the blown units, so the drivers can be reused on another project after a couple of changes to make them better. Unit price is 6e apiece from mouser. I'm going to use them on a larger volume replacement for my 3d printer.

[MrFluffy]

Impatience and oppertunity kicked in & I bought the 860 + motor combo listed for my 4th axis so we will see if its up to the job.
Z quill stepper motor tested (as the X!) , heads getting hoisted into home location this weekend. This is 30% jog speed and later on 0.01mm (0.0004") microstepping. I can make it goes finer as there are finer microstep settings, but I don't see the point.

https://www.youtube.com/watch?v=k1gKDYHi4jQ