Do you suppose this is where the words from the song came from???
"....where Buffalo roam, and the deer and the antelope play..."

 

I hope you didn't say a few discouraging words!

 

This may or may not be an issue, but if none of the above suggestions help you out, I'm leaning towards electronics. I've only ever driven my steppers at 24v or above. Possible that your steppers are being over driven, and thus loosing the torque that higher voltage @ speed would give you. Just something to think about. I found a 24v switching supply on ebay awhile back for around 30 bucks, so cost isnt too much of a factor if you end up wanting to try it out.

 

Well, when the fire started to threaten my home on the range, some discouraging words might have slipped out!

 

I know you're make mounting kits, my friend. However, my 3d printed ABS mounts are solid as a (plastic) rock and up to now I could not detect any movement in them. So that option seems to me unlikely (but not impossible!). I thought about the software or electronics and did a partial dry run (the program takes about 3 hours, so did no do a dry run completely). I will do a complete one today and see what happens. Actually the second run (where the fire started!) was only running for about 5 minutes giving the same problem as the first (which was about running for more than an hour at that moment).

 

Hi Paul.
...forgot... those are nice mounts you made!


You could greatly increase the feed rate for the dry run. Do you have a nema 23 on the Z? Some said a nema 17 is not strong enough.

I like the tool slipping in the colley idea... ive had this happen to me and it took me quite awhile to figure that out... assumed it was my bad programming.

12v is weak... everyone i head is using 24v.... some were even planning 48.

But check your usb cables... it could be a cheap one with poor shielding and picking up electrical noise from the router or other motors.

 

I checked the fastening of the bit, but it was firmly in. I have the habit to check the bit clamping twice before cutting to make sure it doesn't get out. Had something like that happening before and am warned for that. I checked if it was in firmly after this happened, it was.
What I forgot to mention was, that when this started to happen (the second time) the movement of the X axis seemed to "stutter" as if indeed it was missing steps while trying to push the bit through the wood.

The electronics are build in a metal box, with one 12V fan blowing over the controller board and stepper drivers and an other 12V fan sucking out the warm air on the top of the box. The air that comes out is cool, no sign of heat at all. The stepper motors themselves are slightly warm, I would estimate just above ambient temperature. At 12V that seems to be ok. I will have a check with one of these laser heat measure gauges of all components to check the temperatures while operating.

According to the calculations, @12V these steppers will be able to give a maximum power of 33.6 Watts, while @24V that would be double (67.2 Watts). Maximum speeds would be respectively 257.4RPM and 514.2RPM. 33.6Watts seems to me a bit on the low side for feeding power. For that I'm leaning to 24V at least for these steppers. The pull out torque graph for these steppers shows that maximum torque of Ncm 105 will be achieved at 307 PPS with a voltage of 30VDC, constant current 2.8A and half step setting.
If I see this correct (slap me if not!) that would be the same as 84Watts. Operating these steppers @ 12V results in 33.6Watts, = 40%. Missing steps could easily happening in that case?

 

I agree with a dry run on highest speeds to simulate (more or less) the torque on the system. I have nema 23 all around. I read about the issues of some with the 17 and went straight for 23. I also will triple double check the bit in the collet when I start cutting again.

48V might be a bit too much, causing issues with heat (especially in longer duration cuts) in the steppers. I do have a solution for that. On my 3d printer I mounted a 3d printed housing with 12V fan sucking cool air around the stepper to reduce heat for longer prints (some can take 12 hours or more!). I have to see if that will work with a dusty cnc router!

I thought that Robert Hummel is using 30V with the same motor drivers and steppers?

Anyway, what voltage/steppers/drivers are you using? Making such beautiful mounts and plates on the OX? Or.. are you secretly having a huge industrial mill standing there in the backyard?

 

Another thing I stumbled into last night was the use of car batteries in combination with a battery charger as power supply. I like that idea a lot, since I am completely off-grid here and still have some PV panels left that I can use. It's more efficient too, because running the cnc of 230V means conversion from DC to AC and then back to DC. I could even go cutting somewhere in the middle of a rice field!

Anybody any ideas about that one? (No, not the rice field, the battery PS).

 

It's a small mill that was junked and sitting in a friend of a friend's back yard rusting for a few years. After 1 year of tinkering and buying a lot of parts... including a new 24v power supply... it is mostly working now... in my garage and out of the rain... tho we haven't had much lately.

 

Nice! Maybe mill a borehole drill to ensure future water supply? Just saw an article about the vanishing groundwater supply in CA. Not a pretty picture. Thanks for the info.

 

An other idea just popped up! Maybe it has to do with the limit switches on the X axis. Both problems occurred when milling at the outside limits of the X axis. The cabling to the limit switches is not shielded and the EMF of the router and possibly the stepper motors might cause these switches to chatter? (open, close, open, close etc.)?

I will do the dry run with the router on, no bit in it.

 

Dry run with feed rate 4500 and router spindle on: no problems. Possibly not the limit switches, because operation near them was not effected. Steppers were hotter than before, but no alarming temperatures. Still touchable with bare hand.

So, i'll try a 24V power supply now. See if I can find one near here.

 

Darn ... this means I'll also be looking forward to getting a 24+V power supply for myOX. Right now I have a 12 V I'll play with the stepper calculator to see if I should just go to 30+ V PS. The 12 V could always be used to power the fans, lights, etc., keeping them off the power supply of the controller and steppers, which, in itself, might not be a bad idea anyways.

 

Well, quest for power supply was a dead end running into the backwards mentality over here. They still live in a suppliers world, not aware that times are already changed. I can have 100,000 battery chargers of the wrong kind for double the price they are worth and of very disputable quality. Power supplies all ranges 12V (2,4,6,8,10,15,20,30A), but nothing like 24V or even close to it. Yindee Thailand. So, google time. Grrr.

 

hey guys..
about to order my steppers, I need a recommendation on the correct torque. is 2.5 Nm ok or should I go for something higher?

 

What about two 12V supply in serie?

http://www.instructables.com/id/Two-ATX-PSU-One-juiced-24-V-DC-PSU/step1/Theory-Build/

 

Yeah, I looked at something similar on the web. The instructions were full of warnings about getting electrocuted etc. so my first reaction was RUN! But this one makes it look a bit more within my reach (not too smart with electronics) and is worth to try. I know someone having a PC shop and will pay him a visit tomorrow morning, see if he has 2 or 3 of them lying around for a bargain.

Thanks for pointing at this one dddman!

 

If I only could put my Buffalo in the battery/power room, I would have 48V available in large amounts (10.4Kw Pv system off grid). But I guess that would not be a good idea (dust etc.). Probably I first give the ATX option a change, if they deliver sufficient power to run these steppers of course. If that doesn't work out the way I want it, eBay is going to be the next option for a 24V PS. Thanks for helping anyway. Every idea can help to come to a good solution.

 

Hi Paul,

I wonder if your problem is related to the problem I am facing and investigating.

In my case, the router (a trend T4) must be generating huge electric noise. What I have discovered so far:

• After I have started the router GRBL gets in a funny state. When I start a script, the machine just does crazy things. However, when I soft-reset the GRBL before I start the script all is fine. This is very reproducible.
• At some point I was changing the router speed while the machine was in action and the Z-axes plunged and shortly after that the bit broke.
  
• I have found that when I do not run the router at full speed, the machine sometimes is unreliable, when I put it to full speed (RPM) all is fine. I suppose that the speed control generates lots of noise.

Frederic

 

While busy milling in my head for a better power solution, I took the gantry apart and inserted the bunch of v-slot connectors I printed last week. Wow! That makes a nice difference. Much stiffer and both profiles perfectly aligned to each other. Before I had a minor bend in them (from delivery on), which I solved temporarily by having them mounted in opposite directions and 1 bolt in the middle to keep them as much aligned as can be. Now they're perfect. I'm happy with it.

Some re-cabling tomorrow and trying to solve the power supply issue and than to the milling on the table again.

 

My procedure of working is;

1. Set the XY offset
2. Set the Z offset
3. Load the gcode
4. Put the router on and set to desired speed
5. Have a good look around if everything looks alright and ready to go
6. Press the play button (run the program).
7. Watch (nail biting and nervous walking up and down and around the machine with the e-stop in the hand!). Man, cnc is super stressful! LOL.

This doesn't give any problems with the program or funny behavior of the machine. So concerning that point there is no (obvious) relation to yours.

The second one I have not experienced, at least not that I'm aware off. I did change router speeds at times, but this didn't seem to have any influence on the running script or machine behavior. I use a Makita C 0701. It would be interesting to measure the possible electric noise of a router, though. It could have something to do with your problem, I guess?

I can imagine that electronic speed regulation in some routers (possibly also the Makita) is not of very high quality (after all they're not designed for CNC machines and the fine electronics in it). But this will give you the problem that practically you can only run the router at full speed, which is for the OX often to fast I think. Depending on the bit you're using, slow speed and high RPM of the router means that you will burn the material (like what happened to me!) an dull the bit in no time. I think my expensive Freud is gone to heaven (maybe I can sharpen it, have to try that).

Did you use shielded cable for the connection between controller board and stepper drivers? I think that the power cables from stepper driver to stepper motor don't need to be shielded, since they only transfer the power? The limit switches need to have shielded cable too, I believe. What stepper do you use for your Z? a 17 or 23? The 17 seems to be not strong enough. I use a nema 23 of OB for that (all axis).

 

indeed.

Indeed. I need to set the feed rate rather fast. I do not see burn marks or anything, but still the RPM is too high, so I need to get this solved. This is a temporary workaround.

IMHO, all leads could function as antenna's and lead the noise to the electronics. I use NEMA 17, seems enough for me...

Frederic

 

Just thinking, wondering about your grounding of the electronics
Did you follow the sandard star grounding method?
I find if the ground of your arduino is not bonded to mill ground grbl likes to act funny
Causes missed steps random movement and total crash at times.

Start with the simple, it always tends to be a silly over looked thing that causes tons of hair pulling and yelling at the big man upstairs lol

 

Another little trick is to replace your router cord with a shielded cable "hospital grade PC plugs"
It requires you to take your router apart but it helped a few friends that had issues in the past with noise

 

Reading around here, the nema 17 for the Z doesn't seem to be strong enough, which could explain your Z coming down? Maybe in combination with the pulling force of the bit (if you were using an upcut bit)?

 

Robert, do you mean ground as in earthing (like in AC systems)? Or connecting the DC - to the mill?

 

All of your ground wires AC and DC should be earth grounded to same point.
Having multi grounding points or separate ac dc grounds will cause a ton of noise that the arduino does not like.

This is a good little lesson
http://www.lh-electric.net/tutorials/gnd_loop.html