I might do that once I get the new stepper. If it is a bad bearing, I would have to find a replacement to fix the issue ... might not be as easy as it sounds.

Wouldn't a bad bearing cause the issue randomly instead of always and only for any full revolution of the shaft (360°) ? If I do the 360° revolution in any number of segments there is no binding no matter how fast or slow they are made. The stepper just doesn't like doing a 360° (about 3/4" of travel in my setup).

 

Sounds like the latest US election fiasco eh ? The "people" voted ... we just don't know who these people were.

Oh well, let's get back to our OX ... myOX works fine, as long as I don't do a Z move of more than 3/4" in one shot. So I'm back working on 'thin' material, waiting for the new stepper to come in.

 

Interesting thought ... but it worked fine for the longest of time. Why would loose fingers and toes ? That would be a controller issue. It will still be there when I change the stepper ... late next week.

But it brings an interesting thought, when I tested I was sending gcode one at a time and the controller was probably not doing much look ahead logic. It didn't combine say 3 separate but consecutive 1/2" moves (in same direction) into a single 1.5" (over 3/4') move. If I do tweak any of my jobs to split an up/down move of more than 3/4" I have to make certain the look ahead logic doesn't recombine them. Probably need to insert a slight move in opposite direction between steps ...

 

Thinking of it as I wait for the replacement stepper .... I could just rewire from the controller and do the 'exercise' done earlier to test. My other 3 steppers are nearly identical (dual shaft version) to the Z (single shaft). It would confirm if it is the stepper ... I certainly hope at this stage.

At worse, I'll end up with a spare stepper and would have to borrow the controller from the other OX build while I get a new one. The extra stepper could be used to add a 4th axis , even if controller separately (at first) ... all it needs is 360°, so the stepper which might be bad would do the trick, never having to do more than a full rotation for what I expect to use this 4th axis for.

 

Nearly everything is surface mount on the CNC xPro, even first generation, so finding a bad capacitor or such won't really help short of pointing to the controller instead of the stepper. My bet is still on the mechanical side, thus the stepper. I'll soon find if I'm off into far left field ... ready to receive a "I told you so" with a spare (and brand new) stepper as prize.

For now, I'm happy I can work "thin" projects instead of wearing off the bit of hair left on my head.

 

Received the new stepper yesterday while at work, an exact replacement to the existing stepper. Did the change earlier today, a bit nervous ...

Started testing with a conservative 1/8" move of Z, up and down. Everything was perfect - right direction, correct travel.

Did 3/4" Z move, up and down. Things are looking good ..

Then came a 1" Z up. Wow ! No more binding. Down, up, ... all is working great ! The binding problem used to be at 3/4". I was one happy camper.

So I figured I could go for broke, and do a full 2" Z move up ... and ... ... problem surfaced. A bit of testing, it now binds on a 1.5" (or so) move. WT... is going on ?!?

I went through the above a few times, painfully, and the binding moved from single 3/4" (what seems a full rotation) to 1.5" (2 full rotations) moves ... on a brand spanking new stepper.

Did I mention all of these tests, especially the part about the binding are 'rapids' (G0) or equivalent (G1 with no F specified, so it defaults to... max feed rate, right ?) This also happens to be when myOX was binding in Z during jobs : between full retract and moving to the next area to be worked on. Why didn't I think of slowing down the moves earlier ?

So I tried something different. I did the 1.5" and 2" moves with G1 specifying a feed rate of 50 ipm. There was no binding !! Up and down, with pauses at first and then as fast as I could press the up arrow and press return on my keyboard (with the look ahead logic working it's magic). Not a single missed step. All was perfect and, to be honest, without seeming much slower ...

I then increased the feed rate to 75 and all was well; 100 and all was still going great, 150 and I might be wasting my time here, 200 and ... bingo the binding showed up. I bring the feed rate down to 175 ... still binding. Dropped it back to 150 and ...no binding ! All was "perfect".

What is going on ? Why was it working for the past year or more with no issues and, suddenly, I need to set the maximum Z feed rate to 150 ?

So I now have a spare stepper (the old one) and I just need to drop the maximum feed rate for Z axis to 150 ipm. Since this is the Z, typically less than 1" being travelled in rapids, I can live with rapids limited at 150 ipm.

All of this time, weeks gone by, and I could probably of just reduced the Z maximum feed rate to 150 ipm ...

Maybe the driver has something gone a bit off ? Frequency of signal to do the stepping being affected after a certain number ? Still odd that it seems to be for a 'full rotation' and not just random or some other number of steps ...

There's something new to learn every day, eh ?

 

might be, since I had a couple of accidents when first surfacing that thick oak slab ... about the time the problem surfaced as well. As you say, that should affect a specific height rather than only rapids of certain length.

I have not tried with the other controller or testing the Z stepper with one of the other drivers on existing controller.

For now, at least until I upgrade myOX (longer Y axis, ...) I am going to live with a maximum rate of 150 ipm on my Z - even with rapids, since I usually don't cut beyond 100 ipm or so anyways and Z rapids are relatively few (to date).

Fingers crossed ...

 

Well, there is still an issue. When it rains, it pours ! Can you believe it's actually raining outside in mid January above 45 parallel, here in Canada. Freezing rain is expected as the temperature will dip back under freezing point ...

After running the job I could not run earlier I ... noticed that the Z binded again, on a down stroke much further into the job... pulled the plug and retrieved into a small space to think about this and not throw anything around.

Testing, as I have done in the last while, showed that up stroke of any length (at least 2" and some) work great since having replaced the Z stepper (identical model). However, since the situation on this last run most down stroke of, say 1/2" or more things are not looking good at all.

As I manually turned the leadscrew to retract the router from the work piece, my sharp beagle eyes noticed something I had missed earlier. The leadscrew appears to have a smallish 'chip' ! There's a shinny little nick just about the right areas for the latest issue to kick in. Unfortunately, this could mean with each pass of this bad spot through the nut block, especially at the higher speeds, it likely compounded the damage done thus far to the block ...

I figure when I was surfacing the big thick slab of oak some large wood shavings worked their way into the lead screw / block area. But is that enough to damage the leadscrew and/or block ? There was quite a bit of shavings, some pieces of bark, ... vacuum, being held by me, was not keeping up with all of the material being removed. My bad, thinking it's only wood, it cannot damage anything.

This hit me yesterday as I was gearing up to run a few jobs with the rapids limited at a bit less than 150 ipm (all axis since I haven't done the update to GRBL !) It worked for about 20 minutes, at most.

I've taken some time off, not wanting to stress over this. I will do a quick check by extracting the lead screw for an up close inspection. I hope I can just flip it over ... the top half of the 8" length never being used to date ... for a quick fix, hoping nut block's material is not damaged. I also have a spare length left over from when I was building myOX.

Otherwise, I'll have to order a new block. If I do that, I might as well go for the anti-blacklash version ? I should also order a couple of lengths of v-slot to swap the existing 750mm by 1000mm to give me a few added inches of Y... I would then need several feet of GT3 belt to compensate for the added length, at which point I might as well do double belt upgrade. Make a shipping fee worth the trouble ...

If that doesn't fix myOX, after kicking myself in the butt a few times, I will have to swap the controller ... and hope for best !!

I should also finish the installation of a vacuum holder I have already designed for myOX and clean up a few aspects left to "when I get around to it". Time to pass from being a pass time to more serious hobby mode with production for a few hobby shows coming up starting in April, only a couple of months away already ! I need to start making a bit of revenue to pay for maintenance, upgrades and ... my errors to date.

Things were going great until I started working on that chunky slab ... Yet an other lesson learned.

 

Man, that's a bad turn of events. I hope your car is back in working order and that your computer was just taking a break ... you could try connecting an external monitor to see if it wakes up. Although the 'no lights' is not foretelling good news ... unless it is just battery / charger issues. If battery is nearly at 0%, it may take a bit of time to reach a charge level allowing it to start. If left uncharged too long, depending on the type of battery, it could be trouble ...

Maybe it was my jinx which was contagious ... Let's think positive thoughts and scare the jinx away, far far away.

40c ?! You are not anywhere near my neighbourhood for sure. It's a balmy near 0C here. Above average for the season, but a bit cold for my aging bones. I'd take the average between our two temps though, year round at that.

 

I didn't think the lead screw was at fault, but did extract it with minimal disassembly to have a closer look. It looks fine. As you say, it is some form of hardened steel and would need more than wood to damage it in any way.

However, the fit with the block seems loose. There is a bit of sideways play, some wobbling. Pulling up or pushing down is, as much as I can tell, firm. But sideways, it's quite loose ... I am not certain if it is a good or bad sign. With just the flexible joint on the lea screw will work its way down through the block - spinning, not sliding.

Having the stepper off the rest of the Z axis, I decided to exercise it on it's own since I could direct it to do very long moves and really get it going ...

Wo ... it ran fine for a bit, 2", 5", ... even 50" in both directions. Now I was confused. Then it started going nuts ! There was no load, no lead screw, nothing to get in its way, short of a piece of painter's tape to have a visible marker that it is stepping even with the smallest of moves (0.005 mm !)

I kept everything powered, fans going, for hours, and it still behaves at random. Steppers are just warm to the touch, the Z a bit warmer then the other three (they are just sitting idle).

When working properly, I had it do 50", -50", 5, 5, 5, -5, 5, -5, ... with not a single bind.

Then, out of no where, it would bind. From there on, it seemed to bind no matter what ! Then it would stop binding and be ok ! Only to get back into a binding cycle.

Changing the seek rate (rapids) to 5000 mm/min or even 7500 mm/min didn't seem to affect the binding situation that much. It's getting hard to tell ...

This is about the time I wish I was using some form of quick connectors to power the steppers. I could easily swap them without bringing out the wire cutters and soldering iron.

Next best thing, connect the stepper directly to the controller ... which would also remove my wiring from the picture. Just a few screws to play with. I could bring the old stepper back in the picture to save a bit of time.

At this stage it feels like the controller is back to being suspect number one.

Maybe not enough juice is getting to the stepper ? Am I sitting just at some magical limit ? Why has this surfaced only in the last few weeks ? It can't be the steppers, one is brand new and still very shiny ?

Time to swap controller, eh ?

I'll be back.

 

But the problem started with an apparent pattern (full rotations of stepper - how the heck the controller would know that ?)

It did start, from memory, about the time I gave myOX a heavy workout surfacing that oak slab. I had it running for hours none stop, or just about, with a few situations. Even had the router nearly stall a couple of times taking too big a bite on high spots.

I'll swap the controller over the weekend. Right now I'm trying to flash the old one... just to see how to upgrade the other w/out bricking it (with my luck). Using the MacBook seems to make this harder then it needs to be ... The saga goes on, err, learning curve. Breath innnnnnnnn, exhallllllllllle ... Ummmmmmm.

 

Okey dokey. I replaced the controller, a v1 CNC xPro with GRBL 0.8 by a v2 CNC xPro with GRBL 0.9j.

I configured GRBL settings based on my old set up. First testing with X and Y, known to be ok. Nice and slow, just to be safe. and discover the miswired 2nd Y stepper. It was not moving at all ! Having fixed that, I was ready for the task at hand.

Now to the PITA Z axis. Things were not looking up. On the positive side, the 'binding', of any kind, seemed gone. I was, however, moving it very slow and just a few mm at a time. Checking for proper direction ... luckily ! Then I decided going for broke, starting to go for the longest travel (above 2.25" being safe). Up, down, up, down, ... Thinks are looking ok ! Then, with no warning and little effort, it just quite. The only way to bring my Z stepper back to life was to power all off & power all back on. At this stage, the controller is out in open air, fan blowing near it, but no heatsink (yet).

The heatsink are larger and wider then the driver chips, so I thought maybe, when I removed, cleaned and reinstalled them on the other controller one or more of them might be touching something they shouldn't be touching, etc. Others are running CNC xPro with just fans ... and NEMA23. So I should be good, at least that was the thought for now.

Looking at the old controller, I did notice the heatsinks were not sitting that well in any case. The silicon does hold things down but it is far too flexible to apply any force. It just held the heatsinks where they happen to be as the silicon dried. Not to mention a bit of silicon from the previous 'cleaning' was left at the wrong place - just on edge of chip. So very very fine dust was working it's way between the chips and heatsinks. This makes the heatsinks not only completely useless, but it also made them shields blocking the air blown by the fans from properly cooling the chips ! Thus I did not use the heatsinks on the 'new' controller... for now.

But why is the Z now just stopping ?! If it was the chip over heating, it would eventually just start moving again. Here, nothing ... without doing a full power cycle.

I took an other look at the wiring especially the new cable used for the Z (and X given I had cut the wrong wires during the last intervention). I was not that impressed, the gage was quite lite for the need... All I had on hand was ribbon cable used for LED light strings - lower voltage, lower current. In other words, not the brightest move on my part, eh ? Even if the LEDs do shine.

So I dug deep into the reserves... found some 18 awg cable left over from an unrelated job (rewired a trailer a while back). So went ahead and rereplaced the wiring for Z (and X). I had JUST enough of all four colors as well. Destiny ?

Took my time to do real nice soldering (I think), using a different approach to minimize stress on the soldered portion (saw this instructable with 'military / space agency soldering of wires' even of uneven gage). Sure looks good. Feels strong too.

While at this stage, I tested to see just how fast I could do rapids at least with the X and Y axis. I have them cranked to 25000 ! mm/min (just about 985 ipm) and myOX is not skipping a step (obviously not cutting anything, just pushing air). Mind you, given acceleration control and all the other fun settings, it would reach those speeds only on long rapids any ways, right ? Still that is fast (at least for me). Cool how it hardly makes any noise at those speeds. It might shave a few minutes on a big job using most of myOX safe cutting X is 52" and Y is 20.5" working area even before tweaking the g-code (for those repeat jobs worth the time it takes to squeeze out every possible seconds).

Back to the PITA Z axis ...

It still stops after a few moves, any travel length. No binding mind you, it just stops taking commands. Doing a full power cycle, things are back to normal. What the ... ?!?

Any suggestions ? Really, any suggestions will be welcomed. I'm running out of options ... short of finishing the other build from which I grabbed the v2 CNC xPro and see if I have better luck.

This Z issue of mine is getting ridiculous ... and embarrassing !

 

It's on the list ... right along 'sledge harmer' and a few other choice tools.

I will get it going. It can't be that difficult. It is testing my patience and is well beyond my wife's patience. Maybe I'm just trying too hard to understand the situation and I should just try as random solutions as the situation is ?

 

Good ideas.

I will try the Z stepper through one of the other drivers. The X seems a good candidate since it can fly that entire Z assembly right and left at 25000 mm/min with no issues. I can even try the old Z stepper on both drivers and compare behaviour. And should also try the X stepper with the Z driver (role reversal often does miracles to resolve issues).

I set the pots to about the same on this controller as they are on the previous one. All four steppers are near identical. The Z basically being the single shaft version of the other three NEMA 23. Of course, these probably need some tweaking. They're around 3/4 full at this time (hefty NEMA 23 being used).

The controller board is as good as brand new, but that doesn't mean much. There could be a bad solder or such on it. That's kind of luck I typically get ... If there's a cherry pit in a fruit salad bowl, I'm the one who gets it 99.999% of the time !! This has absolutely no (positive) effect on my chances of winning any lottery. ;(

It could be as simple as an overheating issue with this one driver as it doesn't give me much running time even with the small travel. Z stepper does have to 'fight' to keep the router up. This might cause the driver to heat up faster? But that was not an issue with the old controller... unless it's the cause for the binding on longer travel ... but why apparently only on full 360 turns (until the last few tests) ?

Firmware (GRBL) settings sound more likely. 0.09j does have a bunch of extra/different settings compared to 0.8. But I have both X and Y running nice (I won't say great just now, not wanting to jink myself). I'll swap the X/Y for the Z (just rewire at bus connectors).

Tomorrow is an other day ... well, actually, it's already tomorrow, just much too early to start making noise in the garage, even less to start using choice words to release the potential stress.

Oh I'll get that Z to behave, somehow. I just can't spend much consecutive time on it ... or it might get the better out of me.

 

I am totally lost with my Z situation

Using the 'new' CNC xPro V2 controller and the new Z stepper, connecting it to the Z driver, the stepper freezes within a few commands for moves of any length. All other steppers (1 X and 2 Y), same NEMA23 except Z is a single shaft version, fly... even after the Z freezes. So it can't be the controller ... except maybe for the Z driver. Also, when the freeze happens, I just need to unplug the power supply (power to the steppers, thus the drivers), keeping the USB power on through the computer, CNC xPro remains powered, and then turn the power supply (24volts) back on. Everything works for great except for Z. If I power supply is not cycled, Z never comes back to life yet X and Y are happy as can be.

So I connected the Z stepper to the X driver, and the X stepper to the Z driver. The wiring, by the way, are using same cable type, run through the same path along the gantry, inside the plastic chain and over a v-slot for support. It is the new cable ... Everything works perfectly fine keeping in mind the settings, as number of steps per revolution, were left alone: Z commands control X and vice versa. I just swap the wiring at the controller end (easiest place to do).

Wait, why would inverting the Z and X suddenly makes everything work fine ? The switch is done at the controller end. The cable length is slightly longer going to the X stepper than the Z. Both steppers are within inches of each other. None of this wiring is shielded now or before. Any interference would exists on both runs, being along same path and all. Right ?

Ok, so I swap this 'new' CNC xPro V2 for the 'old' CNC xPro V1, wiring the steppers properly : X is X, Y is Y1, A is Y2 and Z, obviously, is Z. The problem of Z binding mostly on longer moves (just about 1.25" at this time) is back ! The controller ?!? I can't see the steppers or wiring being a factor given behavior with the other controller. If the Z moves are less than 1.25" and things are just about perfect ... but it does bind the very odd time. These tests, obviously, are done manually sending the 'up' then 'down' commands by clicking the Z+ and Z- alternatively on the 'machine control' mode of UGCS.

I need to do the exchange of Z and X on this controller. But I just threw my hands up in the air and joined the wife to watch a movie ... The garage is too cold. Which reminds me, I do have a small heater running on the same electrical circuit, different outlet. But if there was interference or fluctuation of power, why just the Z stepper is being affected ???

Does this controller/stepper behaviour makes any sense to any one ??

I needed to run a job for Thursday. Being a small and relatively simple 2D job, I might just cut it using the scroll saw I got for Christmas ... I could do the cutting in the warm basement.

I'll finish building the other machine over the weekend to test both controllers with completely different steppers (NEMA17), ...

 

Good news, of sorts.

I have the C-Beam standard configuration running with the V1 CNC xPro. All three axis can travel at least 3" without any bind/unbind or freezing. Ok, I will admit I just wired and powered everything and sent only a few commands, 1st mm to check directions of travel and right away long travels to comfort myself.

So myOX has something quirky on the Z axis. What, remains to be discovered. This is the bad news. It's not the controllers, which is now more obvious. I did changed the wiring of both Z and X with no improvment on Z nor issues with X. Changed the Z stepper with no improvment either. Changing the controller is changing the behaviour of Z only, but not for the better. Instead of bind/unbind stepper freezes until power supply is cycled off/on. This leaves only the power supply. But why would it affect only Z and only when moved a certain distance (V1) or randomly (V2). I'll redo the tests now that I now it is not the (V1) board. Spare 24v power supply is ... somewhere. So I can switch that has well ...

I'll try with the V2 board in a little while. I just want to exercise the C-Beam for several minutes just to be certain I'm not prematurely jumping of joy ...

Things are looking up ... I think.

 

Having strayed and stayed away from myOX for a while, the sun shining with temperatures reached 19c, birds singing, ... this long weekend, I went at the freak Z issue of myOX one more time before completely taking it apart to eventually build myOX 2.0 ...

First, I changed the wiring to the Z stepper. This time, as I cleaned up some dark corner of a closet, I used some twisted pair shielded cable (low capacitance). This, for those who lost count, is the third rewiring - fourth type of cable !

I left the heavier AWG on the X stepper. It is on its third type of cable, with no apparent issues. Odd since it is using the same path between controller and stepper as the flaky Z axis.

The Y steppers each remain unchanged from day one (1+ year). One is wired direct to the controller while the other is actually using solid (!) copper wire which does not move beyond the entire machine's vibration : it hardly moves.

So I went on to do some quick (desperate) last testing of myOX using the V2 CNC xPro.

1) Z driver driving Z stepper

Driving the Z stepper randomly gives me a "freeze" regardless of travel distance. The only way to get the stepper back into operation is to remove power to the 24 V power supply. To close the Universal Gcode Sender connection and to disconnect the USB cable does not make any difference to the Z stepper status.

The X stepper works great.

2) Z driver driving X stepper

Driving the X stepper with the Z driver DOES randomly "freeze" the stepper. I just need to compensate travel distance for fact that the steps/mm are quite different (and I didn't adjust for the test, just swapped wiring at controller) :

View attachment 22579

The Z stepper works great (being careful with the travel distance : factor of ~ 8x !

Conclusion ?

After all the multiple cable changes, using 3 different steppers (original Z, new Z and the original X), this leads me to think something is wrong with the Z driver ON the CNC xPro. Just my luck as I was trying to determine if something was wrong with my V1 CNC xPro controller when my Z started bind and unbind, first with travels of about 3/4" or more and then, after having changed the wiring, 1.5+" !

Next steps

So I did the same tests with the old V1 to see if the Z stepper (still) binds/unbinds.

Guess what ?!

It doesn't bind/unbind anymore !!!

The only thing changed is the wring ... for the nth time. And, as a side note, the Y axis steppers not being connected ...

Did multiple 2" rapids up and down for several minutes ... to the point of getting carpal tunnel on my clicking hand. Not one bind/unbind ... as if it never happened and it was just a real bad dream (nightmare !) for the past, what, two months ... This is about the time some people would be throwing stuff around, eh? I, on the other hand, still want to know what is going on.

So it seems, with my luck, the wiring to the Z went bad on me (flexing at the soldering within the chain track ? or whatever could cause wiring to go nuts ... although the regular height change needed still is a mystery for me). I made the issue no better by replacing with worse wring (low AWG then high AWG). Which then was, apparently compounded by a bad driver on a 'new' V2 controller. It just didn't make sense ...

Note that I have not reconnected the Y steppers at this point ... just too many controller swapping to bother since it seems fine through all of this ordeal anyways.

Mind you, the far Y stepper is wired with the solid wire which goes through the V-slot which supports the chain guide which contains the Z and X wiring.

So I will reconnect the Y steppers and run through the test again with the V1.

Then, I will reconnect the V2 to the C-Beam and see if the Z behaves or not. At worse, I can jumper controller to make A driver a duplicate of Z and go from there. The C-Beam, stock version that is, doesn't use the A driver.

Wish me luck, looks like a need tons of it at this point.

If all goes well, and the OpenBuilds Store special is for the entire long weekend (today included), I might start ordering a few lengths of v-slot and C-Beam to start transforming myOX and the C-Beam. Looking forward to cutting my own plates for myOX 2.0 ...