[dfmiller]

Bob,
That is looking really nice. I am glad you are keeping the smoke where it belongs.
Sorry for not keeping a closer eye on things. I am either missing or ignoring the update emails.
Soon the real learning will happen.
Thanks for all the pictures.
Dave

[boblon]

Thanks Dave

I'm a connector or two away from connecting it to the computer and seeing about getting some initial movement.....

....however this moment is bittersweet. On one hand I'm REAL excited about getting some movement, on the other hand, I shoved everything in the garage to the outside to have room to build this thing. I've dissassembled my old machine and am trying hard to get rid of as much stuff as I can (painful as it is to through ANY scrap of wood away) not only to make room in the shop for the new toy, but I'm finally fed up having tools I don't have room to use and things I can't find.

So....I've been cleaning the garage,err, shop. I really want to get things organized, but right now I'll be happy to decide where the big stuff goes and deal with the rest after the machine is set up.

My trash collecors are hating me right about now

Anyhow, it won't be long now

BobL.

[dfmiller]

Bob,
Just checking in to see if you are still alive?
So close but yet so far. ;-)
Dave

[WillsCNC]

Like Dave,

Hope all is ok

Will

[boblon]

Hey guys,
Still here. Been a lot of life to contend with lately but have made some progress.
The good news is I have been able to cut some air For the most part the control box checked out good. With all the opportunities for a crossed wire or something I was surprised that other than one loose connection things checked out ok.
The bad news is I blew my last panel voltmeter. I haven't done much trouble shooting there but did discover that somehow I managed to forget to ground the 50v PS chassis. I don't know if this is whats causing the meters to blow. Two different digital voltmeters read the supply fine.
The other thing is one of my motor couplers is the wrong size where it connects to the Y axis ballscrew. I had to wrap some .005" shim material around the shaft to mate the motor for checkout. I have not contacted XYZ about it yet, trying to decide if I need to remove the motor and mic out the shaft and coupler to supply this info to them.
Currently working on the home switches. I am debating whether to even put any on the Z at this point.
It looks like I might have this entire weekend in the shop, that will be great. Will post some pictures soon.

BobL.

[boblon]

I know I haven't posted in a bit but progress has been made.

There is a still alot of work to do but I once again have a functioal CNC machine. Man does it blow my old one away.

It's nowhere near done and when I get tired of playing with it I'll get back to work on it I haven't even really set it up yet... But to catch up...

I cleared out the spot where it will rest. Then I connected the motors and turned it on hoping the motors would lock. They did. I quickly set up some inputs and outputs and tried to get some movement. Gawd, the noise was horrendous. And the gantry racked. Movement was jerky.

Long story short, I had two problems. One was a loose connection at the wire nuts I used just outside the control box. I left them there just knowing I was gonna get at least one wire crossed between all the connection points and connectors. But dang if they weren't all correct. Just a loose wire nut !!

I forgot to check the straps internal to the Geko drives before I installed them. I deduced they were set to half step and have verified that. I set them to 10 micro steps but am not convinced I will leave them that way.

Anyhow, after that movement was much better. I still need some motor tuning but hey, good enough to truck on.

Next I leveled the machine with just a carpenters level. A good one at least though.

I debated on how to attach the home switch on the gantry and ended up just screwing a bracket on the side. I also made some quick and easy magnet holders to trip the switches. I figured I'd worry about adjusting them later. (I had and idea to use an eccentric to adjust). Anyway, I initially just set the home magnet postions by using the lighted switches as a guide. Slide the magnet just until the switch tripped and tightened them down.

Next I taped a piece of paper on the table and stuck a sharpie in the collet. I marked the four corners of a rectange on the paper at 22 x 34 inches. I took a steel rule and measured the diagonals. Dang if it wasn't but 1 millimeter off ! I loosened one magnet and gave it a tap. Tested again and once again it was 'close enough' to proceed.

I set my homeing routines up and used some particle board and mdf I had laying around to put down a sacrafical table and did a couple of surfacing passes.

One thing for sure a dust shoe needs to come quick. I'm trying to come up with something quick to get it on there. (What this really means is a reason to keep playing with the machine).

So anyhow I've just been having a blast the last few days.

I havent' installed any switches for the Z axis and am not sure when/if I will.

I haven't got the router working though Mach yet, have been setting the speed via the knob.

Thanks for looking.

BobL.

[WillsCNC]

Good job Bob,

Nothing better than starting to cut. Your machine looks great and your going to really like it. Keep working on a z limit switch, I would be lost without mine. Make's setup so much better. Keep us updated.

Will

[boblon]

Good job Bob,

Keep working on a z limit switch, I would be lost without mine. Make's setup so much better. Keep us updated.

Will

I'm surprised you find your Z limits so valuable.

I zero to the top of the workpiece and if you're going to plunge the bit into the work then a limit won't stop that.

As far as the upper limit I figure a soft limit will work. Even driving the Z to the top just caused the motor to stall.

Of course I don't tend to run the machine without keeping an eye on it in some fashion.

But then I'm often wrong around this place so am eager to find out if I'm missing something and correct for it.

Thanks,

BobL.

[WillsCNC]

Bob,

Using the soft limits should work just fine. The z limit is just a stop for me. Your little baby is really looking good. Keep us updated.

Will

[dfmiller]

Bob,
Sounds like you are getting there. I will try to stay tuned better than I have been recently.
Had a minor flood here so have been busy. With any luck I will have an office back in a few weeks.
Its been a good thing.
The rebuild to my CNC is on hold as it was in the middle of the action.
My router was outside in my shop so other than boxed in with a few treasures its ready to run.
I am looking forward to seeing more of the first steps and cuts.
Dave

[boblon]

Sorry to hear that Dave, hope the flood didn't cause too much damange.

On my end I am losing steps in the Z axis but I don't think it's mechanical.

Haven't had any shop time this weekend so looking forward to going out and taking some readings.

Good Luck,
BobL.

[boblon]

I had this annoying issue with my Z axis. After running a given job the Z would have missed steps. It was consistent across many runs. I was fairly confident it wasn't mechanical but it took me a while to find it. In the end it was a configuration setting. Had to be set by me but I'm going with an errant mouse click when setting things up.

Here's my version of what was happening.

On the Z axis pin settings on the computer I had the DIR pin to the motor set to trigger on the falling edge of the pulse and the STEP pin to the motor set to trigger on rising edge of the pulse. Don't ask me why it was set that way?
The drive manual says the DIR input must be true 200 nanoseconds before and after the STEP pulse edge. Sounds like the pulse was able to meet this criteria until the timing caught up with it and a step pulse was missed. The software put out the pulse (so it counted it) but the pulse timing caused the drive not to fire. This added up and due to the cyclic nature of the timing it was off the same amount for a given size job. Makes sense now, but was a bear to find.

This could have much, much worse. This is great !


BobL.

[boblon]

Didn’t realize it has so long since I’ve posted any updates. Project is still ongoing albeit with a few glitches and obstacles. It's been pretty hectic between a family crisis or two, a virus I caught and changing employers along the way.

When I last posted I thought I had found the last of the issues causing me to loose steps on the Z axis. Well I was wrong. I did get it down to where the lost steps didn’t show up until a job was run for a considerable length of time. I was rearranging the motor cables in the e-chain and noticed it affected the next run. Assuming it was noise being induced I pulled the cables and put shielding over them. I re-installed them and the problem was the same.

Oh, along the way I ordered an Ethernet Smooth Stepper and installed it. The ESS install was a breeze and appears to work great. I initially had some issues with zeroing but upgrading my Mach install and using the latest plugin for the ESS solved that issue.

So I ordered some IGUS motor cables and installed them. I ordered the PVC jacketed stuff as opposed to the TFE stuff on the recommendation of the IGUS Rep AND the fact that the price was almost half. The PVC stuff is a little larger in diameter and it just fit in the cable chain, AFTER I removed the power cable to the router. So I will have to make other arrangements for it.
Well, replacing the motor cables did NOT solve my problem either. After many, many test runs, tweaking settings and motor tuning I could not get good results. My Z usually ends up above zero after the job but at some motor tuning settings it wold occasionally be lower? I don't know if this clearly indicates a noise issue or something else?

Anyway, I decided I had wasted enough time messing around with trying to trace this down considering the control box was temporary. I decided to just go ahead and re-do the control box. Not planning on having to do this yet it kind of left me in a lurch as to what to use for a control box.

I knew I wanted to seperate the power stuff from the logic stuff. Long story short I picked up a couple of old computer cases.

So, the current plan is to join the two computer cases with the power supplies, contactor, etc mounted in one box and the logic boards and motor drives in the other box. Initially I am only going to install enough to get the 4 axis moving and 'hopefully' verify operation. No frills to get in the way (meters etc).

That's it for now, I'll post some pics when I get going.
BobL.

[dfmiller]

Bob,
So glad to hear you are still alive.
I was beginning to wonder. ;-)
Sounds like life has been keeping you busy.
Hopefully everything is back on a even keel now.
So how much Z variation are you seeing?
Any chance you have backlash issues?

I am looking forward to seeing some picture of you new arrangement.

Not much happening here. Can't get at router right now. Got a couple of things piled in front that I need to dispose of. Office is just getting back to normal after Spring Cleanse. Still have lots of boxes to unpack and finish negotiations with Insurance.


As Winston said. KBO..

Dave

[boblon]

So how much Z variation are you seeing?
Any chance you have backlash issues?

After an hour or so of running a job the Z would be about 1/4" above zero.

I hope it's not backlash, but now that you bring that up I started to think back.

I didn't feel any movent in the Z before I mounted it, seems like the backlash issues you and Will had were obvious in that manner.

I was going try to do some specific measurements to determine if it was backlash but thinking back I can't explicitly remembering getting valid info for that. By the time things settled down my ESS and cables were here and I set about getting them in, hoping it would help. Ran a few tests after that with no change.

Oh yeah, the PC PS I was using in the control box smoked and took out a couple of meters but fortunately nothing esle was damaged (as far as I can tell). Bottom line is I will take some measurements to try to see if backlash is the culprit.

I'll let you know what (if anything) I find out.

BobL.

[dfmiller]

Bob,
A quarter of an inch.
Crike.
That's 2500 steps lost.
You have a serious issue there.
Have you chatted with Steve at PMDX? I would be looking at setup of the Mach XML and hardware issues for sure.
That's not backlash for sure.

Dave

[boblon]

Well, that 1/4" was after a ton of Z movements.

My measurment tools consist of a handfull of dial indicators (0.001") and some digital calipers/depth gauge but the actual movement measured over 1 inch of travel is equal to the desired movement, at least as near as I can tell with these tools.

Still working on trying to get some coherent backlash measurements.

Right now I'm paying the price for throwing together that control box that I know was not optimum (to say the least). Hard to get too critical with all the potential issues in play. So I'll work on the control box.

Backlash is another problem entirely and I need to rule that out if possible. The thing that sticks in my craw is that with all times I altered the motor tuning settings I couldn't find any setting that was right on even when setting the velocity and acceleration very low. The ESS and good motor cables had no effect so I'm hoping it's in the box.

BobL.

[dfmiller]

Bob,
Got time for a phone discussion on this?

Is your number the same?

Dave

[boblon]

Yep, give me a call

[ger21]

It's not backlash. Backlash isn't cumulative, so even if you have 1/4" of backlash, it will still return to zero, provided you return there from the same direction that you used when you first measured.
By moving to a SmoothStepper, I would have though that any Mach3 issues would be eliminated.

What is your acceleration set for in the Z axis? Most of the time, the Z gaining position is due to the acceleration being too high.

Other than that, the only other culprit would be some electronic issues.