Missing Steps?

[OlfCNC]

I am going to change the driver settings to see if it helps. I am not sure what settings in UCCNC that i could have gotten wrong at this point. Honestly, I can hear when the machine is missing steps, which is making me think it is more a mechanical issue.[/QUOTE]

Put an oscilloscope on the step line and capture the signal when you hear the lost steps. Then you will know if it is the controller or the drives and motors or machine.
Undersized motors can be a problem. Undersized power supply voltage for large motors with high inductance can also be a problem.

[wmgeorge]

FYI I do not know anyone who owns a scope anymore. I did 30 years ago. The chances of a DIY CNC builder having one is pretty remote. I am sure its in the setup but I have searched for him and can not find anyone using that configuration. Ger 21 has seen a lot of builds, I am hoping he can help him out.

[OlfCNC]

FYI I do not know anyone who owns a scope anymore. I did 30 years ago. The chances of a DIY CNC builder having one is pretty remote. I am sure its in the setup but I have searched for him and can not find anyone using that configuration. Ger 21 has seen a lot of builds, I am hoping he can help him out.

Oscilloscopes are not expensive nowadays. Some USB scopes cost less than $100.

[drxlcarfreak]

Lets define machine. You purchased a pre-built or was it a kit Fine Line Automation CNC, purchased your driver boards and steppers direct from China, furnished your own power supply and controller and did all the wiring and setup yourself. Now its not running correctly. Here I what I would do. Go find someone or perhaps the people you purchased the UCCNC software from and try to use the setup the same as one that's running correctly. Good luck I have been in your shoes.

I am trying to find the machine you have, is it a 2x3 with ball screws?

Fair enough, I bought the stuff from China, but most of the parts that the reputable guys sell are from China as well. I didn't go for the ebay specials. I wanted reliable parts, I just couldn't afford at the time to buy a premade kit. I only bought items that I was recommended to work well for the setup. I do have a thread going there as well, there is not as much traffic over there unfortunately.

This is the kit I have, except I got it widened to 3', so it has the bigger 3060 gantry as well: https://www.finelineautomation.com/p...x-4-cnc-router

I am going to change the driver settings to see if it helps. I am not sure what settings in UCCNC that i could have gotten wrong at this point. Honestly, I can hear when the machine is missing steps, which is making me think it is more a mechanical issue.

Put an oscilloscope on the step line and capture the signal when you hear the lost steps. Then you will know if it is the controller or the drives and motors or machine.
Undersized motors can be a problem. Undersized power supply voltage for large motors with high inductance can also be a problem.[/QUOTE]

FYI I do not know anyone who owns a scope anymore. I did 30 years ago. The chances of a DIY CNC builder having one is pretty remote. I am sure its in the setup but I have searched for him and can not find anyone using that configuration. Ger 21 has seen a lot of builds, I am hoping he can help him out.

Oscilloscopes are not expensive nowadays. Some USB scopes cost less than $100.

Yeah, I definitely do not have or know how to read an O-scope. My friend just mentioned that too.

I hope he can as well... Actually as I was writing this I was going to say that I think the issue is getting resolved as I did a few cuts and was only off by less than 1/16" at the end of the cut instead of the 1/4" it was, but just as I did, I heard the machine bind and start cutting 1" away from where it was supposed to in the Y direction. So tonight I adjusted all of the belt tensions as a friend said he had a similar issue setting up a 3d printer and needed to loosen them up. I also dropped the kernel frequency and lowered the amperage of the drivers from 5.14a to 4.45a, and changed the microsteps to 8. Each time I ran a cut and it seemed to be getting better. I upped the acceleration and velocity to 30 and 650 and played with it for about 5 minutes to try and get it to stall, it seemed to run perfect, Then I noticed on the right gantry the front left middle bearing and back right middle bearing were not touching the rail after I homed it, almost as if it was racked, so I readjusted them so that they were in constant contact, hoping that would close the tiny bit of drift up. Seemed to run fine, I didn't bother upping the velocity to play with it a ton, and ran one more piece while drafting up this response. About 3 minutes in, I heard it bind and ran over to see how far off it started cutting. I think for my piece of mind, I am going to replace all of the bearings. I am not sure if it will help any, but maybe/hopefully taking it fully apart and reassembling it will remove a gremlin in the machine.

I made a video of it stalling when jogging just after the cut failed. It is right at the end of the video. To me, it seems like a mechanical issue, because it sounds like the motors are trying to move the gantry, they just can't.

https://www.amazon.com/photos/share/...yis3yF52AeFzD1

[ger21]

If you run the machine with the spindle off, does it still have issues?

Does it lose steps on all axis, or just one? I would create some test code that only moves one axis, to try to narrow down where the issue is.
Or even setup one axis at a time, and o some thorough testing to see what the capabilities of each axis are.

I would say that it's highly unlikely to be an issue with UCCNC or any settings.
But without someone having run an identical machine with identical components, it's hard to say what kind of performance you should expect to see.

Are you sure the motors are wired correctly? Are they wired bipolar parallel?

[wmgeorge]

Stupid question of the day, how or what is your machine mounted on?

[ger21]

I upped the acceleration and velocity to 30 and 650 and played with it for about 5 minutes to try and get it to stall, it seemed to run perfect,

Then I'd run it at 20 and 400. If you try to run it as fast as possible, it's always going to lose steps at the worst possible time.

[drxlcarfreak]

If you run the machine with the spindle off, does it still have issues?

Does it lose steps on all axis, or just one? I would create some test code that only moves one axis, to try to narrow down where the issue is.
Or even setup one axis at a time, and o some thorough testing to see what the capabilities of each axis are.

I would say that it's highly unlikely to be an issue with UCCNC or any settings.
But without someone having run an identical machine with identical components, it's hard to say what kind of performance you should expect to see.

Are you sure the motors are wired correctly? Are they wired bipolar parallel?

In the video I posted of it stalling, the spindle was off at that point. It had just stalled prior to that. I will say, when I moved the machine after assembly when I would turn on the spindle it would trigger one of the inductance sensors nearby. It did that last night until I unplugged the sensor and plugged it back in. So, there may be noise in the wiring, or I have the sensor too close to the spindle. I can set some fake code up to see if I can narrow it down a bit more.

After talking through it with a buddy last night, I THINK it is currently only stalling in the Y direction. It used to stall in the X when I had the acceleration and speed turned way up, but I don't think I have seen it do that recently. After thinking about it, I think that any loss of accuracy in the X may be from lack of backlash compensation that I have not been able to calibrate yet.

I believe that they are wired correctly. A+ is Black/Brown, A- is White/Tan, B+ is Red/Blue, B- Purple Yellow. These are my motors: https://www.omc-stepperonline.com/hy...s33-4008d.html

Dumb question since you had me turn the amperage of the drivers down, I thought that I had them on a conservative setting before. Would going up a step from where I had it help? I had it at 5.14A peak, 4.28RMS, and dropped it to 4.45A peak. The motors are 5.66A per phase. I am not sure if that is a peak or RMS measurement so I stuck with the conservative one. You had also mentioned, not sure if it is this thread or another about a 60V power supply. I am running a 48V since I thought that was the max these motors wanted due to their inductance. Will these run on a 60V supply?

Stupid question of the day, how or what is your machine mounted on?

No stupid questions at this point. I built this table. It is pretty solid, about 300lbs of 2x4 and glue. The machine is just sitting on it, but I added pieces at the corners to keep the machine wedged in place so it would not vibrate around. I checked that it is level in all directions. I will double check it again to make sure the bed isn't torqued.

https://www.amazon.com/photos/share/...XXH8kRn4cN0n46

Then I'd run it at 20 and 400. If you try to run it as fast as possible, it's always going to lose steps at the worst possible time.

Oh yeah, when I ran some code I dropped it to 15 and 375. I just wanted to see if the changes made an improvement. I am not trying to push the machine too hard right now, I just want reliability first!

[wmgeorge]

Looks like your making progress. I was wondering if the frame has a twist in it or not exactly on the same plane?

I think this might help you CNC Software Setup Guide | CNCRouterParts He uses a ESS controller but I am hoping the stepper set up numbers could help. How does this compare with your motors > 380 oz-in NEMA 23 Stepper Motor, 1/4" Shaft | CNCRouterParts


Are you using 2 motors for the Y or just one? If two, you have a separate driver board for each, correct? Your build looks nice.

I Have learned a lot here also > https://www.geckodrive.com/support.html
You might consider his drives, I have used in the past and they are very good.

[wmgeorge]

Oscilloscopes are not expensive nowadays. Some USB scopes cost less than $100.

Yes, I did not know they made those, maybe I need to get one.... Thanks.

[ger21]

Dumb question since you had me turn the amperage of the drivers down, I thought that I had them on a conservative setting before. Would going up a step from where I had it help? I had it at 5.14A peak, 4.28RMS, and dropped it to 4.45A peak. The motors are 5.66A per phase. I am not sure if that is a peak or RMS measurement so I stuck with the conservative one. You had also mentioned, not sure if it is this thread or another about a 60V power supply. I am running a 48V since I thought that was the max these motors wanted due to their inductance. Will these run on a 60V supply?

You don't want to exceed the motor's current rating, or you could damage the motors. Running them at a lower current **might** make them a little smoother running, but it might not.

You can run the motors at a higher voltage, if they don't get too hot. This depends on how the motors are mounted (if they can dissipate the heat) and how cool the drives keep the motors. 60V might get you 10%-20% more speed.

[ger21]

The machine is just sitting on it, but I added pieces at the corners to keep the machine wedged in place so it would not vibrate around.

You should have it bolted to the table. It may help by reducing vibrations.

[wmgeorge]

You should have it bolted to the table. It may help by reducing vibrations.

Since its a home built wood table subject to temperature and humidity I would not bolt it down firmly or it will distort the machine frame.

[drxlcarfreak]

You don't want to exceed the motor's current rating, or you could damage the motors. Running them at a lower current **might** make them a little smoother running, but it might not.

You can run the motors at a higher voltage, if they don't get too hot. This depends on how the motors are mounted (if they can dissipate the heat) and how cool the drives keep the motors. 60V might get you 10%-20% more speed.

Gotcha. So the 5.66 amp rating on each coil of my motors, is that peak or is that RMS?

They are all mounted so that they stick out of the R&P plates, but they do get warm running them for awhile, so I am guessing sticking with 48V is a better option. A friend mentioned that he added fans to his motors for his 3d printer to keep them cool. Is that something that could help here as well?

Since its a home built wood table subject to temperature and humidity I would not bolt it down firmly or it will distort the machine frame.

You should have it bolted to the table. It may help by reducing vibrations.

I had originally planned to bolt it down, but saw somewhere that others kind of just locked them in place like I did, and didn't seem to have issues.Plus, the machine is about 300lbs. It took all I had to lift each end to get dollies under the wood to slide it around the garage.Yeah I agree that the temp and humidity would change the overall size, but the way it is built, wouldn't it be all in the same direction?

[wmgeorge]

Your wood table will change as I noted, the issue is the aluminum will not at the same rate. Your only talking a few thousands of an inch to make a roller bind. My steppers on my little CNC and on my 3D printer all run pretty hot. Please read that link I posted from Gecko Drives, he goes into depth about power supplies. I wish we could find someone with the same setup as yours to compare. Mine is ball screw on all axis. Did you contact Nate at FLA, to see what others are running? Your setup I do not think is sold by him anymore as he has gone to a all steel welded frame.

BTW the motors are steppers and they run on DC, no RMS there. Its pulsed DC steps. So it seems your Driver boards are rated for VAC input or DC input. I have never seen a driver board rated for AC input?

[drxlcarfreak]

Your wood table will change as I noted, the issue is the aluminum will not at the same rate. Your only talking a few thousands of an inch to make a roller bind. My steppers on my little CNC and on my 3D printer all run pretty hot. Please read that link I posted from Gecko Drives, he goes into depth about power supplies. I wish we could find someone with the same setup as yours to compare. Mine is ball screw on all axis. Did you contact Nate at FLA, to see what others are running? Your setup I do not think is sold by him anymore as he has gone to a all steel welded frame.

BTW the motors are steppers and they run on DC, no RMS there. Its pulsed DC steps.

That is true. How is your setup? I am guessing you used an aluminum table? Yes, sorry I do need to do that, and compare the motors, from memory, I do think my torque curve was higher across the board, but I will confirm. I have been running around the field all day. I prefer to read things like that on a larger screen, it is on my todo list for tonight! I haven't contacted him directly, I haven't had much luck in the past getting a hold of him. I was hoping to hear from him posting in his sub-forum. He has two series that he sells. The original 8020 extrusion setup and the fully welded steel one. I preferred this one for a couple reasons. The main being cost, I was already realllllly stretching the budget going to this machine, hence the DIY wiring and electronics, and the second is the ability to modify the table in the future. I can always unbolt the extrusions and get different lengths to change the width or length of the table. The steel one is much, much beefier, but probably would be near double the cost of what I have into mine at this point. I will reach out to him separately as well to see if he has any ideas.

Fair enough, the table on the driver control panel confuses me and I didn't want to burn out a motor, so I stuck with the more conservative of the two. . Page 9 of this manual has a peak rating and RMS rating, I was using the peak ratings to be conservative, but am not sure which column I should be matching up to the motors' 5.66a. At first I had them at on, on, off (5.14 peak, 4.28rms), but then per Ger's advice I dropped it to off, off, on (4.45 peak, 3.71rms). If they aren't getting enough torque from the drivers because the amperage is too low, wouldn't that cause lost steps/binding as well? On that train of thought, if I jump up to the 5.83/4.86 that may help resolve issues, right? Or if it is RMS, I could even jump up to the 6.52/5.43 and still be under the motor's current rating.

https://www.omc-stepperonline.com/download/DM860T.pdf

[ger21]

I've seen a LOT of similar machine bolted to wooden bases, with no issues. Along it's length, wood may be more stable than aluminum.

If they aren't getting enough torque from the drivers because the amperage is too low, wouldn't that cause lost steps/binding as well?

If you don't have enough torque to move the machine as fast as you are trying to move it, then you'll lose steps, regardless of the current. There's no chart that tells you how much torque you need to move at a certain speed, so trial and error is the normal technique.

I had you lower the current to see if the motors ran smoother. If they seem to run the same, then increase the current back to where it was. You can probably try the 5.83 amps and see how it works, but watch for hot motors.
Note that steppers can safely run well over 100°F. If you can hold your hand on them for 5 seconds, then they are not too hot.

One thing that's important when testing is to not change a lot of different things at one time.
Change one setting, and do a lot of testing to see the results.
Repeat.

One thing you might want to try is a different motor, like this one. https://www.automationtechnologiesin...lat-381-oz-in/
You can get one and see if there's much difference between the ones you have?

[wmgeorge]

Ok, here is basically what you have kind of, PRO4824 4' x 2' CNC Router Kit | CNCRouterParts Your motors are only rated for less than 4 volts or so, you can go over that voltage say to maybe 24 the Gecko site has the details on that. Frankly I would send the motors you have now and the drivers also back and get whatever cncrouterparts dot com suggest. Read the spec's on the machine above that is what you could expect.

On a wood table I would let float and use locating bolts to hold from moving around and not bolt solid. Shim as needed to keep the machine true. The new welding frame models are a lot more ridged and perhaps could be leveled and bolted down without issues on a wood table.

[drxlcarfreak]

I've seen a LOT of similar machine bolted to wooden bases, with no issues. Along it's length, wood may be more stable than aluminum.

If you don't have enough torque to move the machine as fast as you are trying to move it, then you'll lose steps, regardless of the current. There's no chart that tells you how much torque you need to move at a certain speed, so trial and error is the normal technique.

I had you lower the current to see if the motors ran smoother. If they seem to run the same, then increase the current back to where it was. You can probably try the 5.83 amps and see how it works, but watch for hot motors.
Note that steppers can safely run well over 100°F. If you can hold your hand on them for 5 seconds, then they are not too hot.

One thing that's important when testing is to not change a lot of different things at one time.
Change one setting, and do a lot of testing to see the results.
Repeat.

One thing you might want to try is a different motor, like this one. https://www.automationtechnologiesin...lat-381-oz-in/
You can get one and see if there's much difference between the ones you have?

I do like the idea of bolting it down providing the table is 100% true or at least shimemd.

Fair enough, they were hotter than I was comfortable with, but I could definitely hold onto them. Hmm, I will give that setting a try tomorrow.

Yeah, that is a weakness of mine, and regret it every time.

Hmm, if I go with another motor, I was thinking of going with a NEMA 34 for the X&Y axes. It looks like I can convert my setup for $80 per rack plus the motor. Not cheap, but at least I can justify it to myself by getting an upgrade at the same time.I believe in another thread you recommneded this motor to me, would you still recommend this for an affordable NEMA34 motor? https://www.automationtechnologiesin...l34h295-43-8a/

Ok, here is basically what you have kind of, PRO4824 4' x 2' CNC Router Kit | CNCRouterParts Your motors are only rated for less than 4 volts or so, you can go over that voltage say to maybe 24 the Gecko site has the details on that. Frankly I would send the motors you have now and the drivers also back and get whatever cncrouterparts dot com suggest. Read the spec's on the machine above that is what you could expect.

On a wood table I would let float and use locating bolts to hold from moving around and not bolt solid. Shim as needed to keep the machine true. The new welding frame models are a lot more ridged and perhaps could be leveled and bolted down without issues on a wood table.

Looking at the software setup guide there are a few things I can do to my PC to help make it a more stable output for the CNC. I do have an antivirus on there, That couldn't be causing those kinds of issues could it?

Are you referring to the coil voltage? I had used the inductance calculation which was like 42V, which is why I went with the 48V power supply. Unfortunately, I bought these over 6 months ago. I don't think they are going back at this point. That is slightly beefier than my machine, but I used their site a ton when figuring out what machine I wanted, and also while building it.

I checked out the CNCRP motor to mine, well the more powerful one, up to 400RPM which is as far as mine shows, I actually have the same torque, their's starts out higher and drops much more quickly.

[ger21]

If you were going to switch to Nema 34, I'd recommend going with the CNC Router Parts motors. While they cost more, you know that they'll work, as they've been used on hundreds of machines. I don't think you want to buy something cheaper, and hope that they'll work.