I feel like I’ve been running a marathon and I just broke my leg a few feet from the finish line. I still cannot straighten out my motor tuning issues.
Since this is a used Ethernet smoothstepper, how do you know that it isn’t faulty in someway? Contact Warp9 and see if there is some diagnostic testing program.
Try different pin i/o for your step/direction lines. There are plenty to choose from.
If it were me I’d have been running all along while assembling. Then when I was done I’d still be running. But a lot of people don’t do that. Coincidentally a lot of people find themselves right where you’re at now too. Funny how that works, isn’t it? Almost as if there’s some kind of a correlation going on there.
So just for laughs take a motor off and run it disconnected from your machine mechanically. Just to make sure it isn’t your machine being out of whack being the cause of you not being able to get it to run. If you can get a motor running fast and strong not connected to your machine then you need to readjust a few things.
And no, you’ll be far from the first person this has ever happened to. Problem solving is a matter of eliminating variables and breaking problems down into their simplest form. Right now you’re just dealing with too many things to isolate the problem you’re having. Once you’ve figured it out it’ll be obvious. But right now it is hiding in a thicket of confusion. You need to burn the whole tangle down. Then take things one step at a time. Current, step spacing, and acceleration. But take your machine out of the equation.
Just get a motor running. That will be progress. It is what you should have done at the outset. So it is time to go back and check that box now.
@Jim_Fong That sounds like a good idea. I think I actually bought it from Warp9TD. Do you think it could be faulty? Trying another pin set sounds like a good idea, so I’ll try that first. But, it seems like I probably have something misconfigured. I’m running it with an ST-V2 BOB, which works but it’s not the best BOB. I’ll need to check everything again tomorrow.
@Paul_Frederick I’ll try that too. I did that once before and it wasn’t a mechanical issue.
@George_Allen If you tried it then trying it again won’t net you different results. You remember when I told you to get an oscilloscope right? It would give you more information than your malfunctioning drives do. Then you could literally “see” the problem. Or at least you could see the signal your drives are trying to run on. The shape of the waveform, the timing of it, amplitude etc. Everything you need to know in order to know if it is right, or not. That information would narrow things down considerably. Otherwise you’re just flying blind. If you go really slow you could just use an LED and see it blink on and off. Because the optos in your motor drives are LEDs. You just can’t see them because they are encapsulated in a package and they are infrared. In the very very beginning I blinked LEDs with my parallel port. Once I saw that happening I was confident that I could get a CNC machine running eventually. But you should be able to set it up that you can see an LED blink. You’d really have to bump up your setup time and slow down your speed. As anything over a few Hz you can’t see the LED go on and off really. LEDs ain’t oscilloscopes. But you could still use that method to prove your electronics work, to a point. If nothing you do gets the LED to blink then that’s a problem. Though getting the LED to blink wouldn’t prove your electronics can work any faster than that. So it still leaves things up in the air. But blinking would prove some kind of functionality. You don’t just want to hook an LED up. You want to use an LED and a current limiting resistor. With a 5V signal that’s 330 Ohms. Polarity matters with the little buggers too. They only light up one way. The big problem with electronics often stems from trying to draw too much current. That’s when things go poof. We want to avoid that. But it may have already happened? If it has that would be a problem. Your sundry colored LED (not white or blue) with a 330 Ohm resistor at 5V draws about 12ma if memory serves me. Which is generally safe for integrated circuit outputs to drive. Your BOB had better be able to deliver that much current. Or you’re never running a stepper drive with it. Optos draw at least that much. So what I’m suggesting should be completely safe to do. It should work. Actually on the direction line you should be able to turn an LED on and off just by changing direction. On will be one way and off the other. That doesn’t even change at speed so that would be a definitive test.
@Paul_Frederick it sounds good. I couldn’t get the right voltage on my oscilloscope when I put it together.
@George_Allen BTW are you completely sure you have your motors wired to your drives right? Like have you ever had your motors run right? If you have a pair of wires flipped around a stepper motor will run stupid. There’s a way of phasing a stepper motor “right” but there’s really a couple of “right” ways. Just one “right” way the motor will run backwards. But it’ll still run. You can literally connect the leads at random until the motor starts spinning. The worst thing that’ll happen when it is wrong is it just won’t go. But never disconnect a stepper motor lead while a drive is powered up. That can lead to blowing a drive up. It doesn’t always, just when you really don’t want it to. Even Gecko drives are susceptible to damage by disconnecting live motor wires. Or so they say. So never do it.
@Paul_Frederick One thing I meant to tell you, I won’t need to worry about setup times if I can get it running. The motion controller takes care of that.
@George_Allen I’m thinking you’re using the term motion controller for something different than I do. Mach 3 is your motion controller. There’s no way it can divine drive timings either. That logic is just not made available in any way shape or form. It is used internally in a motor drive sequencer. The dingle donger that makes the coil step signals. Well it fires the bridge. It is actually a bunch of logic gates. I’ve made sequencers out of logic gate ICs. They’re always built into motor drive ICs these days though.
@Paul_Frederick I am pretty sure the motors are wired with the correct orientation. I checked them at the beginning by testing the resistance to make sure they were paired correctly. I think I could get my oscilloscope to work if I got correct voltages. Why wouldn’t I be getting the right voltages if my 9v battery was delivering enough power?
@Paul_Frederick And yes, I’ve learned the hard way how inadvertently or accidentally disconnecting wires from the motors while they are powered can destroy electronics…I have 2 TinyGs RIP’d resulting from that…true Homer Simpson moments! I would like to say I’ve learned my lesson, but that was what I said after the first one.
@George_Allen it sounds like you learned your lesson to me. Have you made any progress sorting out your machine?
If it was me I’d bag your present controller for a GRBL setup just to get running. It is the cheapest thing I can think of that has the potential to make the most difference. In one fell swoop you’d be changing out your whole controller that so far hasn’t done you any favors.
Who knows, you may turn into the biggest GRBL fan going too? Stranger things have happened. Heck if I was there I’d give you an UNO to use. The counterfeit ones only cost a few bucks. I don’t think you need the motor shield either. That’s only if you want to use the little drives. You should be able to wire right up to the UNO board.
But never having used it myself I’m no GRBL expert. I’ve still looked at it a bit. It seems legit to me. I think one problem GRBLs suffer from is susceptibility to noise. I’ve seen people have problems with them and the cheap air cooled spindles. But they put cap filters on the motors and that sorts it out.
You might have a noise issue going on. They can be hard to track down. Noise injected into a machine controller can make it go haywire too. Noise is caused by EMI, ground loops, coupling, just bad signal shielding. The trouble is control logic is extremely sensitive. But big drive motors are giant noise emitters. So noise becomes a huge factor with CNC.
When I made my BOB I intentionally made it not very sensitive. It needs some current to trip. The chips themselves will trigger on picoamps. You could practically breathe on them and get them to switch. But I used resistors that makes it take milliamps to get them to trigger. I have one of them crappy Chinese air spindles and I never filtered it. My machine never missed a beat either.
Noise is one of the big CNC bugaboo pitfall traps though. A bad noise issue will cripple a CNC machine. Just mess it up.
@Jim_Fong I think I’m going to try to switch input and output pins for my Z driver, maybe later today, if I get my windshield replaced in time & I’m feeling a little better. I’m beginning to think you may be right about that PIN on the ESS not working. It seems as though the others are working okay.
I wanted to ask you a question about your gecko driver: do you connect the pc +5v to the common terminal from the BOB or the pc gnd? Documentation also states that you can connect +3.3v from the BOB to the common terminal. How do you decide which to use?
Also, if necessary, it should be fine to use the gearbox stepper on my Z axis shouldn’t it? That axis shouldn’t need much speed and I know that the geared stepper is functioning properly. I could use the Nema 34, which is also working, to drive the X axis.
@Paul_Frederick The sensitivity and noise issue could be the problem. You’ve given many excellent ideas and areas to sure up and I certainly appreciate your input. It’s obvious this isn’t your first Rodeo. I’m going to stick with this controller for the time being. I actually have a grbl controller here and it works, but it lacks some things I want with this machine. I may be wrong, but I feel if I can identify this motor issue, I may be satisfied with the results, if I can manage to keep the drivers and motors from overheating during a long run. I realized yesterday that I had the 2 Y axis motors set for 2 different step frequencies, which caused one to work and the other not. As soon as I returned those to the same they both worked and were quite powerful and rapid for my taste. I probably will need to slow the velocity and acceleration on that axis a little, but I am pleased with it. I’m considering moving the Nema 34 from the Z axis to the X axis. I think that would insure the necessary speed and torque for that axis. What do you think?
If I do that, wouldn’t it be sufficient to use the geared stepper on the Z axis? The reason I would like to use it, is b/c I know it’s working well, it would supply a little additional torque that another Nema 23 wouldn’t have, and I’m not entirely certain that my other Nema 23 is functioning properly (so I’d need to get another motor.
@George_Allen wait until you run a job. Z is the most used axis. Every new cut you have to feed down, then when you’re done go back up again. Plus Z has to fight gravity too. I put the strongest out of the 3 motors I used on my Z axis. You’ll see. I suppose it matters what kind of work you do but I mostly do engraving so my machine does a lot of small cuts. Up and down, up and down. One of the biggest factors when I am cutting a job is the Z safe height I set. The closer the better.
@Paul_Frederick Yes, that’s what I figured initially. Someone on Facebook (who I had never written to claimed that I didn’t need that kind of motor on the Z axis). I thought, as you commented, that it would likely need the most powerful motor.
@George_Allen It depends on the design of your machine but often lifting Z is the most challenging load. X and Y can be made pretty balanced. Work is defined by lifting a mass though. Which is what Z is doing half the time. What’s most surprising is how much Z ends up moving in a CNC job. Every cut start and stop involves cycling Z. We see the cuts but don’t imagine all of the up and down that went into making them. You’ll see when you get running. You may be like, half the time my machine is just going up and down. Which for 33% of your axis is a disproportionate amount of use.
@George_Allen I only have older geckos that are +5 common. The new geckos can be either. It doesn’t matter what you choose since it will both ways. +5 is found on USB port and it convenient.
I run my Z as fast as X&Y. When doing simultaneous 3sxis movement such as 3d carving, machine speed is limited by your slowest axis.
I actually have a servo on my Z so it can run really fast. Good if I have to drill a bunch of holes. I don’t like to wait for the Z.