The never-ending saga of the Hurco that's "ready to make parts"

Re: Code issue, please help. This stopped being fun a few days ago.

Yet another chapter in the CNC mill that is "ready to make parts", or so the ad claimed.


So I start taking parts off to get the ball screw out and about halfway through I decide to check something. I have the nut unbolted from the housing but the screw is still installed and as I was moving it I noticed how smoothly it moved. I didn't hear the clunking I heard when the table moved. Put an indicator on it and nothing, no play. I'm thinking maybe the play is only noticeable when it's moving a heavy table so I grab hold of it as best I can and try to rock it back and forth... nothing. What gives? So I have a brainstorm and bolt it back up. Now I take the indicator and put it on the end of the shaft that's sticking out of the front of the machine. Guess what the measurement is............ .004". That's right, the exact amount of slop I had in the table. I had checked the end nut and thought I had it tight enough but apparently not. Gave it a few light taps until I got to the next tang on the lock washer and took another measurement. I'm getting somewhere 1/8º and 1/4º turn in the pulley (it's really hard to measure that small) but more importantly my indicator, which is .0005", got barely a wiggle.

Looks like it's not the ball screw after all but my instead it's my inability to remember to check the simple things first. Still good news though. I'll get it all back together soon and make another test cut.
 
Re: Code issue, please help. This stopped being fun a few days ago.

Good catch. You'll be up and running in no time. Pretty sure this is going to make a huge difference.
 
Interested in the 1micron comment? , please help. This stopped being fun a few days ago.

Jim, this sounded interesting. Not to hijack this, but have you posted about this magnetic scale system you setup? Sounds interesting?

"micron magnetic scales on the table rather than rely on the encoders for positioning. I am looking at the actual table position, "
 
Re: Interested in the 1micron comment? , please help. This stopped being fun a few days ago.

countryguy said:
Jim, this sounded interesting. Not to hijack this, but have you posted about this magnetic scale system you setup? Sounds interesting?

"micron magnetic scales on the table rather than rely on the encoders for positioning. I am looking at the actual table position, "
Click to expand...


The best write up I've done is here http://www.dawsoncontrols.com/millupgrade.html

I'll be happy to answer any questions you have. Maybe start another thread for that.
 
Re: Interested in the 1micron comment? , please help. This stopped being fun a few days ago.

That is a very neat write up Sir. I always pick up a ton of brain cell material reading your stuff. Thanks for always taking so much time w/ us! Really appreciated.

Now, Here's something you don't read everyday: from your link : " my WD-40 / kerosene mix (my preferred coolant for aluminum machining), " ;-) Wowza! You are a steely eyed missle man!



JimDawson said:
The best write up I've done is here http://www.dawsoncontrols.com/millupgrade.html

I'll be happy to answer any questions you have. Maybe start another thread for that.
Click to expand...
 
Re: Interested in the 1micron comment? , please help. This stopped being fun a few days ago.

countryguy said:
That is a very neat write up Sir. I always pick up a ton of brain cell material reading your stuff. Thanks for always taking so much time w/ us! Really appreciated.

Now, Here's something you don't read everyday: from your link : " my WD-40 / kerosene mix (my preferred coolant for aluminum machining), " ;-) Wowza! You are a steely eyed missle man!
Click to expand...

Thank you for the kind words. It's my pleasure. This gives me something semi-useful to do when I don't feel like actually working. Maybe I can pass on some of the experience that I have gained in my 50 or so years of doing fun stuff. Being about half crazy helps too.
 
Re: Code issue, please help. This stopped being fun a few days ago.

I've gotten everything back together and hoped to have some good news, but alas I do not. One good thing i guess is the thud is gone. We can rule that out as having any effect on matters.



Here's a couple new videos of the issue I'm having as it stands today.



Still walks a bit but now I have a jump included. See the area around 0:08 - 0:11 of the video below.

https://www.youtube.com/watch?v=nHVVEaRYM_4&feature=youtu.be

Interesting thing is it does not appear to walk while traveling "in" only "out". Not sure what to make of that. The jumping is intermittent too. Sometimes it doesn't happen but usually it does.




Here is a video of me cutting a circle. I slowed he feed down in the trouble spot to make it easier to see. When it get's to the bottom of the circle (0:30), it jumps just like the video above.

https://www.youtube.com/watch?v=_9iLqSq0FdQ&feature=youtu.be


The readout seems to keep track of this in both instances and appears to try and compensate. In a previous exercise I got the voltages mostly balanced for in and out as we suspected this may be a culprit. So if it's not the voltage and not free play in the screw, what's next to check?
 
Re: Code issue, please help. This stopped being fun a few days ago.

This is becoming more interesting. There is one mechanical problem that I can think of, and that is a floating gib. If it is sticking in one spot then the motor torque has to increase to overcome the stick, this can cause the motor to overshoot the position and then try to correct. In watching the videos, the motion seems to indicate something like this, but it could have other causes also.

At least, you know the ballscrew is pretty accurate now.

The other external possibility is that the encoder is flaky and is outputting a ''fuzzy'' signal. This is a somewhat common problem. The only way to check this is with a dual channel oscilloscope. The output from the encoder consists of two square wave signals that are out of phase by 180 degrees. These signals should be well formed and crisp. This gives the controller the direction and position data. When these signals are ''fuzzy'' it confuses the controller.

It is possible that there is a intermittently bad connection in the encoder cable, maybe a broken wire that makes contact most of the time.

You might switch the X and Y cards in the computer to see if the problem moves with the card.
 
Re: Code issue, please help. This stopped being fun a few days ago.

I'll check the gib first since it's the easiest.

When you say switch the cards, you are referring to the card with the pot on it I adjusted correct?
 
Re: Code issue, please help. This stopped being fun a few days ago.

Yes, physically switch the cards in the controller cabinet. The adjustments are probably not exactly correct but should be close enough for testing. Just stay close to the E-Stop button when you power up, and during testing. I don't think the motor will run away, but you never know.
 
You must log in or register to reply here.
Back
Top