slaving y axis motor for 4x8 cnc

[eloid]

question about slaving y axis motors

is it better to have a mechnical linkage between "slaved motors cuppling the motor directly"or a pinion shaft with split pinions linking just the carrages not the motors....thus allow independant motor slipage but still allowing for mechnical linkage via mechincal shaft to spur gears that would just algin the y axis carrage

[justCNCit]

I'm trying to do that. With stepper motors its a nightmare.

[jsheerin]

Why do you want the complexity of mechanical linkage? If you're willing to do all of that, why not just use dual servo motors slaved in software. You'll know if they get off by more than a certain amount, and you'll probably have more torque (depending on the specific motors) so it's less likely the gantry will get out of square to begin with.

If you do go mechanical, what are you going to do to compensate for the different lead error of your screws (or similar positional error of a rack and pinion or any other mechanism)? Are you going to measure the error differential with a laser and make a cam to allow one of the nuts to rotate to compensate? I've seen a build that does this.

Also, if you link the two sides mechanically, it seems like a bad idea to have two motors. You're just asking for something to get bound up.

[justCNCit]

all i know is about 75% of the time my machine works normally. EVerythings lined up pretty well and it abruptly stops moving and sticks at the oddest times. I'm running a dual y axis setup but wouldn't bother with mechanical linkages it sounds like a bad idea anyways.

I don't trust my machine to run reliably and lately have been pricing servos but thousands of dollars wasn't exactly in the budget. just venting a bit.

that being said when it does work it's pretty good. But my nerves are on edge and its not fun to run a machine that screws up on its own.

ps. xylotex 4 axis box w/400 oz in motors & 10 TPI acme screw 5 starts, 3 axis machine about 4x6 feet. It's all greased and pretty well lined up and the screws are pretty clean but not perfect. Maybe its cause there is a small bit of dust on the screws

[jsheerin]

Sorry, I was really responding to eloid's comments - not attacking you or anything.

I have a 4x4' machine with dual x 425oz stepper motors slaved in Mach3. They've never had a problem pushing around a 100+ lbs gantry with the same screws you have. The Z is what used to give me fits. It sounds similar to yours, at least in terms of me not trusting my machine. It ruined parts several times. Then it would go for months without doing anything. If it's sticking, you've got to trouble shoot it. Going to servos won't necessarily help. You'll probably have more torque to push through things and you'll know if you get out of position, but that only helps so much. I had two issues - one is that I was getting sawdust built up on my linear bearings and it was making the stepper miss steps. But I could never hear it miss steps when it happened, so I didn't think it was. I finally heard it one time. So cleaning my rails religiously and lowering my acceleration helped with that. However I also found out my parallel port could not source enough current to drive the G540 properly. I was amazed it was working at all. I ended up putting in a pci paralle port card, but not all of the pins on that had enough current. It was a dual port card, so I made a custom db25 cable that used some of the pins from both ports to get enough working ones to run the G540. No problems since then, but if I'd just upgraded to servos, I still wouldn't have had a good source of step and direction signals.

Having said that, the N34 size servos at Keling are what I'm looking at for the upgrade I'm working on (on my 4x4' router). They're about $130 each iirc. Then you add in ~$130 for a G320, $25 for an encoder and a power supply. Then a breakout board for the entire thing. I figure about $1500 or so to do the entire machine (4 motors). Then a bit more for timing pulleys and belts and rack and pinion parts. I've got ball screws coming for my z and thk rails for everything. Everything will be totally enclosed, so no more dust on my rails. And a 5HP 3ph spindle. I wish I'd known enough to do all that the first time around, but I guess that's how it goes.

[FandZ]

Originally Posted by justCNCit all i know is about 75% of the time my machine works normally. EVerythings lined up pretty well and it abruptly stops moving and sticks at the oddest times.

I recently ran into a similar problem. I found a bearing that had gone bad. It was just dragging back and forth. At first, I couldn't figure out what it was because it would only happen every so often. The way I finally found it was when it kept jamming on an angle when it was cutting. I ran that gcode over and over, and placed my finger over each bearing until I found the one that wasn't turning. It was a big waste of stock...lol Hopefully your problem is something simple like it.

You have a build thread anywhere? I'm all for slaving two steppers.

[justCNCit]

yes jsheerin I know you aren't attacking me its just that an hour before I posted that I was mad enough to haul my CNC to the scrap yard

considering Gecko drivers at this time.

[justCNCit]

yeah no kidding

i thought all this talk about servos the first time around was for perfectionists

now it looks like unless your machine is always 100% in perfect running order you will always wish you went with servos. It really does suck in retrospect.

I know that cleaning things off like that makes a big difference but my confidence level is still too low to do a big job on it. OR any job for that matter. Hit and miss

ps. no build thread sorry. I can provide pics though

[eloid]

Originally Posted by jsheerin Why do you want the complexity of mechanical linkage? If you're willing to do all of that, why not just use dual servo motors slaved in software( there will be salved in software). You'll know if they get off by more than a certain amount, and you'll probably have more torque (depending on the specific motors) so it's less likely the gantry will get out of square to begin with. If you do go mechanical, what are you going to do to compensate for the different lead error of your screws (or similar positional error of a rack and pinion or any other mechanism )? Are you going to measure the error differential with a laser and make a cam to allow one of the nuts to rotate to compensate? I've seen a build that does this. (not sure that there will that much compensation required fo justify all the above. perhaps just a springloaded pinioun set up) Also, if you link the two sides mechanically, it seems like a bad idea to have two motors. You're just asking for something to get bound up.

(I have seen a few set up with this concept that use liner belt track in steed of R/P.)

[eloid]

Originally Posted by justCNCit all i know is about 75% of the time my machine works normally. EVerythings lined up pretty well and it abruptly stops moving and sticks at the oddest times. I'm running a dual y axis setup but wouldn't bother with mechanical linkages it sounds like a bad idea anyways. I don't trust my machine to run reliably and lately have been pricing servos but thousands of dollars wasn't exactly in the budget. just venting a bit. that being said when it does work it's pretty good. But my nerves are on edge and its not fun to run a machine that screws up on its own. ps. xylotex 4 axis box w/400 oz in motors & 10 TPI acme screw 5 starts, 3 axis machine about 4x6 feet. It's all greased and pretty well lined up and the screws are pretty clean but not perfect. Maybe its cause there is a small bit of dust on the screws

( not planning on using any lead screw (just R/P) etc will be using way covers and will be designed to keep moving part out of site to reduce build up.)

[ger21]

Mine is slaved in Mach3, Xylotex with 250oz motors and 1/2-8 2 start. If I set it at 150ipm, it's 100% reliable. If I set it to 175ipm, it's 95% reliable. So I leave it at 150ipm, and don't have to worry about it.

If the machine is run within it's capabilities, software slaving is all you need. If you want to try to go faster than it's capable of, you'll have problems, regardless of how you do it.

[justCNCit]

I am running at 50 IPM tops and it looks like even a very small amount of dust is what does it on the screws.