Bridgeport Manual to CNC conversion...

[Davek0974]

Thoughts over coffee - Limitations of the system?

When i was building the large scale traction engines, everything was made with a tolerance of only 0.1mm or better - easily obtainable on a hand controlled Bridgeport with a three digit DRO. It really was not needed to go tighter on these engines and many say that if too tight they never worked properly due to expansion when running etc.

Anyway, the third digit on the DRO was pointless as you could never set that precision by hand and merely looking at the mill out of the corner of your eye made it wobble more than that!

So, now (when done) I will have ball-screws and servo's with Mach3 acting as my DRO display - this is all running on my bench very nicely now.

But - the encoder/motor is connected to the screw via a toothed belt - surely this must introduce an error possibility, especially with a 100kg sitting on the bed??

My main point here is - If i want to use the mill manually, say to drill some holes without cam-ing the job, I can put Mach into reset, this drops out the servo motors and i can spin the handles like before and Mach tracks the position on screen - I am hoping here that this is a workable situation with decent accuracy when positioned manually??

[JAZZCNC]

But - the encoder/motor is connected to the screw via a toothed belt - surely this must introduce an error possibility, especially with a 100kg sitting on the bed??

Not really if the belts are tensioned correctly and decent quality. Lots of very expensive VMC have belts on the screws.

Regards Manual operation then it will Display position just like Manual DRO's would.

[Davek0974]

Was playing around with PID tuning today, reading the manual on how to do it manually (ooh-err) It was not happy and no amount of fiddling got it better, long positioning time, not sitting at position etc. That was until i stuck my finger on the end of the shaft and gave it a tiny amount of resistance - instant improvement rapid stopping on position etc. Lesson learnt, seems unloaded motors don't like tuning ;)

I have ramped it up to 3000rpm, been deliberately messing about with rapids for several metres each way plus millimetres each way, then put mach into reset and jerked the motor around, reset and did a go-to-zero and the damn thing snapped straight back to my zero mark - I was most impressed with this feat ;)

JAZZ - how did you wire the enable line to the drive - there is a signal mis-match here as CSMIO puts out 24v but the drive needs to be pulled low - I stuck a relay in which works but....

Also, what size fuse would be good for a drive and 750W servo?? Thinking 5A slow-blow?

[JAZZCNC]

JAZZ - how did you wire the enable line to the drive - there is a signal mis-match here as CSMIO puts out 24v but the drive needs to be pulled low - I stuck a relay in which works but....

I do it with a relay tied to the HV option on IP-A that enables all the drives.

[Davek0974]

Thanks, if it really is relays then thats fine, I have that working on the bench setup, just wondering if that was the only option here :)

[Davek0974]

Ok, where did I slip up...

Previously had the motor tuning set as book, 2000 steps per (10,000 / 5mm pitch)

When I fit this on the machine it will be at 2:1 ratio so i set steps per to 4000 (10,000 * 2 / 5mm pitch) and set velocity to 7500 (3000rpm / 2 * 5mm pitch) acceleration is at 750mm/s/s

Works ok in jog mode, works ok in code mode at speeds up to 7500mm/min

If i do a "Goto zero" it runs for a while then stops with ePid fault.

It also allows me to set feed higher than it can handle and then goes into ePid fault.

Is there a reason here?

[Davek0974]

Ok, found it - it seems it does not like running right at the limit of max speed - if kept just below max it all works perfect.


If I set the max at 7500 and command a feed at 7500 it faults, if I set 7400 it works perfectly in all modes.

Won't get t hat high in real life as the bed is not really big enough i think for a 7500mm rapid ;)