Missing steps on X axis

[bigred8080]

Hello All,

I built a buildyourcnc.com "book" build 4 years ago and have been tinkering with it off and on ever since. I took about a 2 or three year break and over the past few months have got back at it. I have cut a few things here and there and they look pretty decent for what i have ha.

i set off to attempt my first "detail" work and realized i am losing steps. i first discovered this because i forgot to zero my Z and hit the estop. i clicked GO TO ZERO on mach and noticed it was roughly 1/2" off in the negative X direction. (center start).

My set up is as follows:

NEMA23 425 oz. stepper from https://buildyourcnc.com/item/electr...is-425-elcombo

i am using the BoB and drivers that are included in the kit.

currently using 13 TPI all thread.

Mach set up:

Steps per - 10400

Velocity - 20

Acc. 5

I have checked to see if my coupler is slipping and it is not.

Also, i think my steppers are running a little on the rough side. once the x gets going it sounds fine (I think) but the Z, it sonds pretty rough. I dont know if its because it is barely moving or what....just sounds bad from what i can tell. The Y doesnt sound much better but i mainly think its because it is move the gantry ha. Any clue on this as well? or can you point to a video of how theyshould sound when turning all thread/ lead screw.

Any suggestions or ideas on what is going on?

Thanks
BR

Similar Threads:

• Missing steps on x axis
• Missing/Gaining steps Y axis
• Help please with axis missing steps.
• Missing steps when jogging in more than one axis
• z axis sticking / missing steps

[awerby]

Looking at that site, it's remarkably short on specifications. There's no indication of what those motors really want in terms of power, but my suspicion is that it's more than the 36 volts supplied by the power supply provided with them. Is there a part number on the motor so you can look up the inductance? Often people size the power supplies they offer with their "kits" not to drive the motors optimally, but to avoid blowing up the cheap drives they're pushing.

You could try tuning the motors and cutting the acceleration, but the speed is already fairly pathetic at 20 ipm.

[bigred8080]

As mentoned earlier, this build was based on their book, which can be found here. The steppers they reference on page 50, which they say is the ones they offer on their site is the Kelinginc KL23H286-20-8B. Here is a spec sheet. Also, in reference to the tuning settings, that reference can be found on page 203. I have double the velocity as they mention that these values can be changed, but what worked best for them. maybe i should revert to the slower velocity and see if i get better results

Thanks for the help. Also, i was thinking of recording a video of my machine in motion so i could show the "rough" sound i am referring to.

[awerby]

Well, as I suspected, the inductance of those motors is limiting their performance. At 6.8mH, (in parallel) they want about 83vDC to run at top speed. If you want to use those drivers and that PS, change the motors for ones with less inductance (a lot less). If you want to keep the motors, swap out the drivers and power supply for some that work at about 80v.

[Biggs427]

The problem is definitely the Motor/PS match-up, but 13tpi is really big. Your motor will have to spin like crazy to achieve workable speed!

I have low inductance motor (less than 1.6mH) on my CNC and even these, I would barely be able to hit 100 ipm.

With cheap T8 rods (8mm per turn) my CNC rapid movement are at 400ipm and will be 500 once I kick my own bµtt and install the 1610 ballscrew that collects dust in my basement.

[bigred8080]

Well, as I suspected, the inductance of those motors is limiting their performance. At 6.8mH, (in parallel) they want about 83vDC to run at top speed. If you want to use those drivers and that PS, change the motors for ones with less inductance (a lot less). If you want to keep the motors, swap out the drivers and power supply for some that work at about 80v.

So to work with what I got, I should just play with the velocity and find the sweet spot where I'm not losing any steps?

The problem is definitely the Motor/PS match-up, but 13tpi is really big. Your motor will have to spin like crazy to achieve workable speed!

I have low inductance motor (less than 1.6mH) on my CNC and even these, I would barely be able to hit 100 ipm.

With cheap T8 rods (8mm per turn) my CNC rapid movement are at 400ipm and will be 500 once I kick my own bµtt and install the 1610 ballscrew that collects dust in my basement.

So I would need a 80v PS to be able to turn at full power? Like you said, I'm just using all thread so I know I'm limited there. If I was to upgrade from what I have, what would you suggest? I hope to only have this machine only a couple more months ( trying to upgrade) so keeping costs down is preferred.

Thanks
BR

[Biggs427]

So I would need a 80v PS to be able to turn at full power? ...
BR

As awerby mentionned, the kit you have probably don't have the stepper drivers that will withstand 80v. It's not only a matter of installing a bigger power supply.

... If I was to upgrade from what I have, what would you suggest? I hope to only have this machine only a couple more months ( trying to upgrade) so keeping costs down is preferred...
BR

I'm far from being a specialist but what I've learned is that it all depend on what is your expectations in terms of feed, material, precisions. If you can live with some backlash, T8 nuts are dirt cheap. You can even tweak it to reduce substantially the backlash with a ghetto anti-backlash T8 nuts, but the cutter load can't be too high because it will overcome the spring load.

T8 nuts are not nearly as good as other linear motions but with this tweak it can make an acceptable solution for a hobbyist that don't have too tight tolerances.

[bigred8080]

After talking to a guy at work, I took some advice and tried running my program with all axis unplugged except for the one I'm having issues with to see if the problem persist. After running for almost 30 minutes I stopped the program and returned to zero. It was spot on.

Does this prove that my power supply is too weak for the motors running together? If so, I noticed that, like awerby said, my drivers would need to he upgraded as well as they are limited to 36 volts.

So they way I see it, I have three options:

1. Upgrade my threaded rod to an acme screw, thus reducing the number of turns to produce the desired movement.

2. Buy another 36 v PS and power one of my motors and drivers with its own.

3. Upgrade everything ha

Am I off base here anywhere?

[Biggs427]

What is the amperage of your PS, Driver and Motor?

Acme rods wouldn't hurt.

A feed too slow will burn your cutter. When I first built my CNC, I had 18TPI rods and were not able to go beyond 45IPM. I killed so many cutters!

The acme rods are not ideal but at least I can cut at higher speed and I change cutter not because it's burned but because it dulls. ;-)

[bigred8080]

PS is 10 amp
motors are 2.8 amp
Driver is 0.9 - 3 amp. i have it set at 3

[Biggs427]

Have you tried to run your test program with all the drives plugged but with only your X axis moving?

Also , have you run a test program with no load? have you tried to swap driver (for example x and Y) and see if now the Y axis is missing steps?

[fretman_2]

Interesting thread...I too have high inductance motors and a not high enough voltage power supply. My motors need 70v and I'm giving them 45. But...I have 1/2-10 5 star screws...2 turns per inch. I really don't think I'll cut any faster than 40-50 ipm on guitar bodies. Though...I haven't cut a guitar body out yet so I don't know if I'll have missed steps.

[Biggs427]

I've put lot of research before I buy my stepper. I've found that I was better with medium sized stepper with less inductance than big stepper with higher inductance.

As an example, when I tested, I could achieve 10 000mm / min (393.7 IPM) rapid movements using cheap TB6600 Chinese drives and a 24v / 15amp PS to drive three nema 23 270oz steppers.

My stepper needs to spin the screw at 1250rpm to achieve this speed but this makes me uncomfortable... I've set my rapids to 7000mm/min (275IPM) which is more than enough for a 18 x 18 inches work area.

I'm using this setup since September and the only time a part was off was because the Y axis coupler between the screw and the stepper was loose.

[bigred8080]

Have you tried to run your test program with all the drives plugged but with only your X axis moving?

Also , have you run a test program with no load? have you tried to swap driver (for example x and Y) and see if now the Y axis is missing steps?

I've been out of town for a couple days so i haven't done much with it. The last time i ran anything was running a 3d roughing program with the z and y unplugged (same program i found with my original problem). that's where i found that after running 30 minutes and returning back to Zero, the x hadn't lost any steps. I did this with no load (cutting air).

Should i create a program of the x just moving back and forth and leave everything plugged in? I have not tried swapping any drives around.

[bigred8080]

So here is what i have done. I unplugged the y and z and ran the program. my problem went away. i plugged everything back in and just did some rapids in the x direction, just back and forth 10 times or so. everything seemed fine. i tried my 3d roughing again (cutting air) and everything seemed fine.

I decided to give it a go and hope for the best. I just finished the roughing and everything seemed fine. I'm about to start the finishing pass.

Maybe it was a connection and by me plugging and unplugging everything it worked itself out? even though i didnt unplug the axis i was having problems with.

[Biggs427]

That might be it. I installed aviation connectors on my setup to make sure the conneciotns were good from the beginning. Very happy with it.

Good luck

[bigred8080]

Here are the results from the finishing pass. Its supposed to be a Celtic weave design created using Vectric's Celtic Weave Gadget. The steps seemed to keep up, so that was a good sign. This was cut into Oak plywood (im thinking that was the wrong choice lol) using a 1/8" Ballnose bit. I know this is a bit off topic, but what can i do to help with the rip out of the tops? The bottom portion came out great, just the centers is whats killing me. I have a single speed router...27K RPM. so i really cant do much on that end.

[ger21]

I know this is a bit off topic, but what can i do to help with the rip out of the tops?

Use a different material, or take shallower cuts at a slower feedrate.
Sometimes sealing the wood with a sealer first can also help.

[fretman_2]

Oak, at least red oak, is notorious for chipping.

[bigred8080]

Thanks for all the replies guys. I think im going to take all the above advice and see what i can do. First ill try sealing what i have with some poly. if that doesnt work ill try some new material and maybe seal that too.