So I switched from the GRBLcontroller to the UniversalGcodeSender-v1.0.7 to communicate with the controller (CNC xPro). It adds some nice easy control (pause, cancel, ...). Also adds some info on file, estimates time left, ... The 3D viewer is nice to have as well and can be closed, freeing up some CPU cycles if a worry with slower computers. I like it. I felt better with the feedback and immediate controls like 'pause'. Although I suspect pausing at the wrong spot will introduce a bit of a error (??)

The next release of UGS should allow a smartphone or tablet to be used as jog controller and probably to monitor progress... did not try the nightly release (still in testing mode). I still have to figure out was is going on with my Y steppers.

I used meshCAM to convert a black and white line drawing of a "Fleur de lys". It was an easy process and it showed that all the generated code was ok. However, feeding it to ChiliPeppr-GRBL gave some odd results in its simulation. Nonetheless, I fed it to UGS and its 3D simulator handled everything just perfect with no a single (apparent) issue. So I ran it on myOX. I was a bit too agressive with the depth of passes. I could see myOX struggling to move along the Y axis. X and Z seemed just fine. Y steppers are starting to get on my nerve ...

I must of messed something up with the agressive routing of the Fleur de lys job. I might have to tune current up to the NEMA 23 of Y axis (and X). The 3 of them actually run quite cool. While the NEMA 17 of Z is warm towards being hot but not burning hot (I have sensitve skin).

So tomorrow (tonight now) I'll go through the entire process, step by step, just to get a handle on the situation. I played way too much with settings and such of meshCAM that know I messed up somewhere along the way.

I'll have to tune a few GRBL parameters, like the delay count and such.

Getting quite close to 'completed' status ... woohoo !

 

My first problem was a lose wire at the controller. Oddly enough, it was for the X axis which was working just fine all this time. The wheels for both Y plates, but particularly for the right one, where a bit tight. So I adjusted the eccentrics in consequence. It seem to help quite a bit. The right Y stepper became nearly as smooth as the far (left) one.

Using a 48" length of Incra T-track + (it has measuring tape built in with 1/32nd inch markings) I started measuring the longest movement possible for both X and Y axis.

X axis is darn near perfect (as much as my eyes and 1/32nd markings allow me to tell). Went back and forth for full 48" with no apparent loss. The UniversalGcodeSender allowed streaming of commands for no pause/delays between movements. Did a few shorter movements which seemed just as good. I'll have to get the calibrating down to as precise I can measure on width of myOX (about 52" in X axis).

Now for the 'mystery'. On the Y axis, the most I could go was about 19". Here's the problem : whether I did 6", 12" or 18", I seem to fall short about 1/4" ! The error was not affected by the length of the movement... Moving back the same distance brought me to the starting point. So I was not randomly missing steps. Most intriguing, if I did moves of 1", myOX was darn near dead on !!! What is that all about ?!? Now, if I sent a stream of commands, no pause in between, I would get random lengths. Why would this happen only in this axis ?

If my steps/mm (or inches) parameter is off, it would also be off for the X axis which has the exact same setup : same NEMA 23 stepper, same 20 teeth pinion, same GT3 belt. Yet, the X was fine even on 48" runs.

The only difference is the Y has two NEMA 23 to be driven simultaneously while the X only has one... The X also has less weight to move around. But these NEMA 23 should be plenty strong, right ?

Doing short moves, less than 1", in the Y axis, seems to hide the issue. Thus making the routing of small test patterns look fine ...

I check the GRBL parameters, especially the two for delay: time between commands and time to keep steppers energized. I did change them as they were quite low. I could not see any difference, so that should be good. The default feed and seek rates were also quite low (750 and 1500 respectively)... So I set them to 1500 and 3000 respectively. They get overridden by G-code anyways, right ? The faster defaults just made the 48" test travel that much faster. Tried F7000 rates along the X axis with G01 commands ... Vibration like stuttering started to be noticeable. But there didn't seem to be any loss steps.

So what could be wrong with my setup ? As some hinted in other posts in regards to problems with there machines or to make suggestions for newbies like me, I am starting to suspect my power supply is really lacking humph ! Why did I get a 12V !?! Sure, current wise (Amps) it has plenty to go around with all 4 steppers going. However, especially after checking the steppers specs, it became clear I was looking for trouble from the get go. This was back when I was a total stranger to all of this building. Turns out the specs do suggest "24-48 Volts" !!! I now know better, from first hand experience and, apparently, learned to read such little details after nearly ... 5 months.

Z, with it's NEMA 17 is doing fine. It might be running a bit hot. So while I already maxed out the CNC xPro current setting pot for the X and Ys, I might have Z's a bit too high (?) Once on a higher voltage power supply, they'll all have to be tweaked, right?

Back to my two Y steppers, the far one uses heavier gage wires which runs through the X V-rail. There is well over 1500 mm of each of the four wires. Making it longer, substantially longer in fact, than the other Y stepper which has it's original wires - cut short in act, direct to the CNC xPro controller (the mirrored/cloned driver).

Now I need to find myself a 24 or even a 36-48 Volt power supply ! If I order one from China, it will take at least 2 weeks ... If I order one from Canada or the US, it should be in my hands a lot faster, but probably at a much higher cost as well. I have little choice. The 12 V will still be good to drive the fans, lights and such.

Can those in the know confirm this is most likely my problem. Driving the 2 Y steppers with my 12V power supply, especially if all four steppers are going at same time, is probably a big factor in my issues at this point. Right ?

So close to being complete ... going through about every possible newby errors along the way.

 

I'm known to be a very patient person ... most of the time. This project is no exception as I am learning along the way. I would of preferred to learn from the mistakes of others, but, as you say, most are quiet about their learning curve posting mostly their success or they already built many other machines and putting an OX together is no big deal for them.

Rather than throw things around when they get on my nerves, I tend to take them apart to try and figure out what is wrong, or, just to see how they work (or should be working). I did that as a kid starting with my Lionel train set. I should of learned then, the locomotive never quite worked since ! (Mechanical/electronic things that is, the living things I will eventually get frustrated with and move on to others.)

Having said sad, even before your great suggestion, I was and I am still seriously thinking of doing a thread about the "do's and don'ts", at least from my experience as I am no expert (yet). Not quite ready for that as I will have a few more don'ts to discover once myOX is finally ready to truly engrave/carve something useful ... or at least intended to be useful and larger than a thumb tack machined at the far left edge of a 52" wide OX.

Hey, you noticed ? I do tend to run on at least with the typing. Maybe I should of been a writer ? Otherwise, I'm, surprisingly, a very quiet person ! Would that make me a dumb writer ? (I mean as in mute, of course.)

So off I go to hunt for a heftier power supply and see if that does the trick. My current 12 V seems to drive a single stepper fine ... and the NEMA 23s I got do suggest a 24-48V power supply. There must be a good reason for the suggestion, eh ? From it's specs I also noticed a little something about "5%" precision. That's 0.09 degree or ... 10 steps per revolution and at 8 microsteps it then becomes 80 microsteps !! Is that 'real bad', 'bad', 'ok', ... or the 'norm" ? Every one talks about squaring their OX, getting the GT belt tensioned just right, blabla. I even saw a post somewhere about making sure both Y belts are lined up just right respective to each other so microsteps move both ends perfectly in sync... But no one mentioned, in my search so far, about the stepper's exact specs like this "precision" bit. 80 microsteps off from time to time, randomly at that, could make a big difference in the resulting machining, no ? Sure, in a work of art, being art, it probably won't be noticed to the naked eye or just become "character" ...

There I go again... Off to my hunt for a 24+V power source with enough Amps to get myOX zooming at incredible speeds and the precision of reclused monk turned engineer. I will be able to not only route a toothpick out of a solid oak 16' piece of 4"x4" but it will have the full North and South engraved, the unabridged version, both English and French versions at that ! (Note : I never read it, just heard it was quite long ...)

 

Well, my hunt for a more appropriate power supply sent me back to China.

I also almost forgot the CNC xPro only takes up to 24 Volts as I was going for more volts since my steppers are best for 24-48V.

A 24V 10 A should be on my door steps in about 2 weeks for about 20$cnd shipping and all included. It was about 26$, but has luck would have it I had a coupon to use. Alternatives either were nearly triple the cost, plus shipping and, in some cases, came from China anyways, at times with lead times of several weeks, and nearly for the same item... thus no better. Best alternative, with most choices, would of been StepperOnLine, out in Asia, but the shipping is way too high for just one item even if shipping is only an amazingly fast 2 days. OBPS was out of stock, so was a few other 'regulars'. So I'll be patient and find something else to keep me busy for the next few weeks ...

Looky here, the 12V power supply I currently have seems to have voltage adjustment and, in my hunt, I noticed a lot of these xV power supplies looked suspiciously similar if not identical. Do you think the similarity is more than skin deep ? Can my 12V go towards 24V somehow ? Just wishing as I contemplate pacing in the garage for about two weeks ...

... Off to SketchUP I go for a couple of weeks, prepare some 'parts', simulate with ChiliPeppr, etc. Maybe time to do a recap in a 'dos and donts' format, eh ?

 

Darn ... we are getting close to Christmas and, well, this year at least, there could of been a miracle on my street instead of 34th.

As I more recently read in some post somewhere on the 'Net, maybe I should get the batteries out of the cars and give myOX a quick boost just to see if getting up to 24V or so would solve my problem before the delivery guy shows up in two weeks ... now let's see ... is it parallel to add up voltage, series for amps ... or is it the opposite ? Anyone for fried OX ?

 

​

A bit of an early Christmas miracle with a scratch and dent found on a far back shelf as OBPS comes to the rescue again. I should be receiving a 24V 20A power supply dusted off in the next few days. The China 10A will find an other project soon enough. Call it a spare for now ... think 4th axis sooner ?

I found some 1/4" pre-cut acrylic/plexiglass at 90% off - so I got a whole bunch of various sizes. It's going to be perfect for ... something. If I can get myOX going before they run out of stock, I'll get a bunch more

MeshCAM sent me an extra 15 days of trial (a regular offer, but I'll take it as part of an early Christmas)

I'm getting the hang of SketchUP / SketchUcam (I now only have a few days to design a few 'projects') ...

It feels like Christmas is already here ... I can't wait to open a few of those early 'presents'. I'll pay forward in some fashion soon enough.

​

 

Well, not that I looked hard, but there doesn't seem to be much of that nature around here ... The few stores like"The Source", a remake of Radio Shacks, hardly carry parts. Computer stores mostly have ATX type power supplies. All the DIY/Maker things are lacking big time here ! Mostly basement/garage type individuals each on their own ... just like with R/C trucks/cars/planes, couple of hobby shops dating back from ever with no competition and no real stock of anything ... Heck, the nearest track is 45 minutes away, each way !

Anyways, I'll soon have two 24V power supplies. A China 10A in about 2+ weeks cry and a 20A in a few days thumbsup.

 

Still waiting for either (or both) 24V power supplies to come in ... Got sick, not of waiting (yet) ... just sick, so I didn't get to do much else but "wait". Let's hope tomorrow brings a power supply to my front door.

 

Got the 24V 20A power supply late yesterday - in mailbox when I got home late from work. It had signs of being banged around, as indicated by seller. It looks like the mounting tabs are bent, all of the rest seems ok... Fuse and fuse link check out as well. No (visible) cracks or any other 'damage' beyond the frame dents. Switch at "115V" and time to power it up !

Well ... that was a flop ... It's DOA ! I get a better voltage reading off my body than this power supply. Not even a puff of smoke, much less any sparks. Just no output power ... Opened it up again and there is no sign of anything wrong ...

Hooked up the 12V PS to double check things, like my multimeter and its settings ... works fine with a bit more than 12V. So this 24V PS is ... useless.

No Christmas miracle on this street after all

Anyone with power supply "debugging" skills to maybe help me get this "24V 20A" working ? Fan and LED are working. So the input (110V AC) and all that goes to the fan and LED (lower DC voltage) is in working order. When power is off (unplugged), it takes a few seconds to shut fan and LED (capacitors discharging ?) As far as looks go, it should be working. I guess one should not judge by looks alone, eh ? Oh, if it helps, it's a OBPS unit dusted off due to the physical case 'damage'.

Now I have to wait for the 24V 10A ordered direct from China to arrive. It should only be a week or so away at this point, I hope!

Let's also hope the 10A will be good enough to get myOX working hard and precise ... Christmas is just around the corner.

 

Tell me about it. When the fan and LED turned on, I was all excited and looking forward to getting myOX going to click that "project complete" check box, even without all the bling and such in place. As long as myOX could do a good job (for me).

Another 24V 20A was apparently found in a dark corner, it was tested and is already on its way.

These guys are amazing and go the extra mile to help out. I will definitely have to pay it forward somehow at some point, the sooner the better.

In the meantime, I'll be Zen and chalk it up to "learning" ... I should really of gotten a 24V power supply from day one, or the steppers for a 12V environment. I kind of wish the CNC xPro could go beyond 24V as well, but such controllers should really have a modular approach (replaceable drivers). Maybe there will be a CNC xPro+ down the line ...

I'm also searching the web to see if there is anything SAFE I can do to get the 'doa' unit back to life. After all, the fan and LED are working. There was no smoke or sparks. Fuses are ok. How bad can it be ? Maybe it's just a bad solder joint from getting hit (the bent corners) ? I just don't want to be found as a pile of ashes in the process. So I'll only go so far in the process.

 

Actually, better still would of been buying a 24v power supply from the get go. No mods, no waiting, ... and likely cheaper than getting two ATx power supplies and mod'ing into one. Part of my learning curve, eh ?

Even the FutureShop / BestBuy don't have the lower cost ones with enough current in local (and not so local) stores. It's "order and get delivery to the store". Of course, the few in stock are closer to the 199$ a piece units. So to get two of them would get ... hmmm ... kind of expensive, eh ? They did have this one at half price. Bu at 50$ (plus the 13% sales taxes) each (about 125$ for two) plus parts for mods ... voiding possibility of warranties and returns ... makes it worth the wait for the real thing or ordering one from the beginning.

Actually, I might of been able to grab a few power supplies from 'dead' machines at work... But one post I read warned identical power supplies should be used to put them in series ('safer' ?) Those dead machines are also way old, you guessed it, with PS at fault or weaker than my good arm.

Not to forget, from what I can read, putting two (or more) power supplies in series involves being careful with the ground as the output of the first becomes the ground of the next, ... The one I would have the hardest time with (short term memory thing) IS not allowing the chassis to come in contact with each other Otherwise, ya, it is like putting batteries in series... just that there is 110 volts AC involved at each one and some mods to void warranties and possibly get jolted.

Is there a way to rig them together without doing any 'internal' mods or cutting harnesses ? Maybe doing a board to plug in the harness of each with output 'ports' for the various voltages/amps needed in a typical (and not so typical) OX build ... From the reading I've been doing, it sounds possible. It becomes an extra cost. But it would make it much easier for everyone. Maybe it could even have jumpers/switch matrix to create any permutation of output voltage(s) ?

Some of the newer (and more expensive) ATx power supplies seem to come with detachable harnesses now. Maybe just having 'special interconnect' harness might then be possible (extra cost again) for specific output voltages. I do like idea of having some form of an interconnect 'board' though ... just plug and go.

I also read how one could rig potentiometers to turn into a variable voltage/amp "bench power source" - use for multiple needs, even with multiple simultaneous outputs of various voltages/amps.

Anyways, I'm patient, at least for these situations ... but I am eager to get things going at this point. Just a few more days ... not the end of the world at this point. It's almost Christmas so I figure it will just become a real nice Christmas gift from me to me !

 

The cheapest I could find in town was about 30$, unless I go to the local electronics dump site and manage to get an OK to strip a few 'donated' computers.

Interesting, you mention they can be had for no more than about 20$ a piece, but the one you direct to off eBay sells for 24$US + 30$US for shipping. That's a total 55$ + about 15% to exchange to $CND on my calculator. I might as well get one out of the corner store or... a 24v supply at that price, no ? That's over 100$(CND) for a pair which still need to be modified ...

It certainly sounds easier then I thought, now that you repeat some of the postings read all over the 'Net. Idiots like me prefer not to play with 110+v AC to that level (although some of us do electric work around the house anyways ). So a simpler plug and pay approach feels more confortable then snipping wires and soldering inside a power supply which has high voltage caps, charged up to zap you without notice if care is not taken before diving in. The DOA I have has a 600v capacitor ... a couple of, what were they now, 350v as well. Even unplugged, power off, those caps can hold a charge for a bit longer than I would think.

Don't forget, you are talking to the idiot who, as a kid, picked up a solder gun by the wrong end ... thumb and index fingers are still marked, as the kid, 40+ years later ! Just talking about it I can feel the lesson sinking in still. I also used to stick paper clips into power outlets (ya, the 110v ones) ... cool how they would evaporate leaving just bits here and there, eh ?

The twinned ATx could definitely make for a very practical bench multi output adjustable bench power supply. Just combine 2 or more ATx outputss to make 5v, 8.3v, 10v, 13.3v, etc. The more ATX in series, the more permutations become available.

Good point about the amps which no one should forget, being in series, the weakest is the maximum amps out. So if you do go with two different ATx, make certain the weakest 12v rail as enough amps for your OX since 24v with too few amps is not much better than too few volts.

Question for you (or anyone in the know out there) : if an ATx has true multiple rails, like some seen have 2 true 12v rails, could both of those be wired in series to get 24v out of a single power supply ??? In theory at least, it should be possible right ??? That would make the better 50$ or so ATx an attractive solution ... the fiddling would be just inside a single power supply. One just has to make certain those 12v rails are distinct as opposed to "virtual". But I'm still learning and could be way off my rocker here.

 

Having left the DOA power supply unplugged for a good while, opening it in search of some easy problem to fix and reading lots of posts all over the 'Net about how these bad boys work, or should be working for the case of the one before me, I might just go scavenging for a few descent ATx units in the local "dumpster". Prauk might of twisted my arm just far enough to convince me.

Would a 'dunk' in plasti-dip, or similar isolation layer, be good enough to make certain there is virtually no chance for the multi ATx setup to 'touch' (grounding when they should not) ? You never know when something conductive might fall unto them and make the breach ... I know myself too well not to take that extra bit of precaution.

Better load up a few screw drivers and wire cutters in the car before heading out for the day ... You never know when you might cross path with a bunch of abandoned donor PCs around this time of year. Wait ... I might have a couple of donors in the attic !

Wish me luck ...

 

Oh we can get it. We just have to pay shipping through the nose. The power supply you suggest from eBay turns out with a 50$ shipping fee for Canada. The unit itself is just 33$ ! Who's making off like a bandit in these situations ? Looks like the shipper. I'm getting a 20A direct from China for about 25$ shipping INCLUDED (!) straight to my door. .. I just have to be very Zen and patient about it. Lately, it takes about 2 to 3 weeks to get stuff direct from China with no shipping fees. If they are small orders there is no duty or taxes either. Who is loosing out on those deals ? Not me !

 

Go figure, eh ?

 

Problem is most of the stuff comes out of China, or nearby, anyways. The odd part of all this is that if one orders direct from China, risking the 'one is good, all is good' the shipping is free to front door, give a few weeks to receive. Why is that ? It still goes through borders, multiple handlers, half way around the globe... including our trusted Canada Post ... Maybe instead of overcharging the local flow they should start charging the inbound flow somehow ? Oh wait, they seem to do that with States even if there is NAFTA (North American FREE Trade Agreement).

It's like flights, it's cheaper to fly to Europe and many other places in the world then doing a 2 hours flight within Canada AND the departure city is the same ! The train is even worse ... not even a high speed service ... But those are a conversation for an other forum.

Getting back to my power supply situation, I should be getting the replacement 24v 20a unit in the next couple of days. The 10a China one probably by the end of next week or just before Christmas. I looked at the attic inventory and those ATx PS were the weak kind (not enough amps anyways, ...) But it got me to clean out some of the attic (not a bad thing of itself). Dumpster dive scheduled for tomorrow (at work and the local municipal 'recycler' that is). So I should have myOX back in testing mode in the next few days one way or an other.