Yamazen CNC knee mill

[KTP]

Sorry, I really can't tell much in the first picture...the glare and the trace is just not showing up well in the photo. So are you trying to say you are getting 2V of noise on your step and direction lines when they are disconnected from the Pixie?

Do you have a part # or schematic of your breakout board? I am curious how they are isolating signals and how they route ground.

[N4NV]

Originally Posted by KTP So are you trying to say you are getting 2V of noise on your step and direction lines when they are disconnected from the Pixie? Do you have a part # or schematic of your breakout board? I am curious how they are isolating signals and how they route ground.

Close, I am getting 2V of noise on the step only line when they are disconnected from the Pixie. The direction lines look just like picture 2, nice and flat. That is why I don't think it has anything to do with the Pixie, encoder, motor etc. The only common right now between the Pixie and the breakout board is the ground and +5V.

I am using a CNC4PC C11 board. I don't have a schematic. Arturo is sending me another board to try. When I told him about the noise he said "I have been thinking a lot about the noise problem you are having. It makes absolutely no sense. The board only has buffers, optoisolators, and gates.
None of those components is capable of generating noise. "

When trying to tune the Pixies, I could not have more than one at a time connected to the step and direction outputs from the breakout boards or they caused the other axis to fault. I put ferrites on the step and direction lines and was able to get them all tuned, but I was loosing steps.

Vince

[KTP]

Well if you had the Pixie, motor and amplifier disconnected from the step/dir lines and still had the noise on the lines, I would say that eliminates them as the problem

If you had an old parallel cable (like a printer cable) you *could* snip one end off and just wire the pixie board step/dir lines directly to the computer parallel port without a breakout board. As long as you are carefull about static discharge it could be a good way to test out the system. Of course if you have error in your wiring you could end up damaging your computer, so maybe it would be better to wait for a new breakout board to try.

[N4NV]

I posted the following pictures of the noise I am getting below.

http://www.flyingcritters.com/images/Noise10.jpg
This is my 5V power input to the breakout board.

http://www.flyingcritters.com/images/Noise11.jpg
Time set to .5uS and 2v/division. X axis noise

http://www.flyingcritters.com/images/Noise12.jpg
Same settings, Y axis noise.

http://www.flyingcritters.com/images/Noise13z.jpg
Z axis, there is a little jaggedness on the left side of the trace.

http://www.flyingcritters.com/images/Noise15.jpg
Sometimes the Y axis looks like this. It depends on what keys on the keyboard have been pressed.

http://www.flyingcritters.com/images/Noise17.jpg
A zoomed in picture of the Y axis noise. When I press the X > key the noise changes to that of picture 15. When I press the X< key (even though I am monitoring the Y axis), the picture looks like #10, no noise.

Vince

[N4NV]

I believe I have the noise problem solved. I was accidentally shipped an earlier revision of my breakout board that had some minor issues with the high speed opto isolators. I just installed the latest revision of the board and my scope screen looks perfect. Tomorrow I will start back on tuning my Pixie 100 cards. Thanks for all the suggestions.

Vince

[N4NV]

I replaced my opto-isolated breakout board with a non opto-isolated one and all the problems went away. I am now able to rapid all the axis at the same time at 200 ipm smooth as glass, no lost steps. I could probably go to 300 ipm, but the mill is still up on 4x4 blocks waiting for a move and it rocks more than I am comfortable with.

Now that I have the drive system up and running, I am moving forward with some of the other controls. I hooked up the powered draw bar and after replacing some air lines, it works well. I am at a loss though with how the spindle brake was connected. It should interface with the draw bar or the draw bar will spin the spindle. I can't figure out how they did that. The brake has a solenoid that operates the brake. The brake needs to engage on both the tighten and loosen cycles of the draw bar. The only way I can see to make it work is through 2 relays, and it was not hooked up like that when I got the machine.

I also need to run the fault and reset signals to the control panel. Otherwise I have to open up the pendant to reset the drive. I tend to fault the drive by not having the drive power on before attempting an axis move.

Anyway, things are looking up. I had spent almost 2 weeks trying to track down the noise problem.

Vince

[N4NV]

I went back in to work today and scrounged up a couple of 120V ice cube relays I could use to make the spindle brake work. I wired them up so that the brake would engage when either the drawbar in or out is run. It works well. I did not think it would make that much of a difference, but the draw bar engages much quicker and smoother with the brake on.

The next step will be wire in the automatic one shot lube system, and then start on the spindle control. I ordered a card that will change a step signal into a 0 to 10V signal the VFD requires. I also need to get the manual spindle controls and coolant controls wired in. These are going to be a little trickier as I have to work in a holding circuit.

Vince

[N4NV]

Putting the noise issues behind, I have been moving forward. I added a power feed for the knee. That big knee is just too much work to crank up and down.

I have been running test to see how well the repeatability of the machine is. You can see in one of the pictures I have been using the roadrunner file from Mach 3. The lines overlap perfectly. There is a double image in the picture because I did not want to change paper and just moved the origin a little and did another couple of runs.

I have the VFD running under Mach 3 control. I did find a problem though. My VFD has to have an open circuit on both the forward and reverse direction pins when it is powered up, but with the speed control board I am using either one or the other is always closed. I figured out that I will run the signals though my EPO switch so when I start the VFD the EPO will be open and both the direction circuits will also be open. I will also run the 0 to 10V speed single through that circuit for added safety. The EPO actually controls a couple of 4 pole relays so I can have 8 circuits controlled by the EPO.

The pictures are of the power feed, the pixies, breakout board and speed control board, breaking resistor (under the VFD), and the roadrunner drawing.

Vince

[N4NV]

I have been getting more into the controls now that the machine is running well. The Pixie cards can all be wired together so that when one faults, they all fault. I decided to wire them this way for safety. For instance if the Z axis faulted when it was supposed to retract, the X or Y axis could crash the tool. The Pixies can also be wired so that when you reset one, they all reset. I did that as well. I mounted a lighted switch on the control panel so that when a drive faulted, the switch would light up. Pressing the switch would reset the drives. I needed some way of having the Pixie light up the switch. I found the solution at http://www.cncbuildingblocks.com. They have an 8 relay board that will take a TTL signal and control a relay. In addition to the Pixie being able to fault each other, any of the fault signals can also supply a TTL signal at the same time. The relay board arrived yesterday and I wired it up like the manual said and it worked the first time. My lighted switch had a 120V light so that is why I needed the relay.

After playing around with it, I found that I had another problem. If the servo drives are enabled and the Pixie’s are faulted, the motors will creep. I thought about it over night and when I woke up this morning I had the answer. If the Pixie TTL output could drive a second relay, I could use that to tie into my EPO circuit which disables the servo drives. It also has the added benefit of shutting down the VFD that is driving the spindle.

After work tonight I wired in another relay to the Pixie fault output and tied that relay through the EPO wiring. It all worked. Now if a Pixie faults, it shuts down the spindle, servo drives and lights up the switch.

Next on the agenda is to install a latch on the pendent enclosure so I can open and close it easily. Currently is has two screws that hold it closed. Then it’s on to the manual/automatic switches that are mounted to the front of the quill in the original configuration. I would like to get them working again. I am thinking that I might be able to use the CNC building blocks relay board for them as well.

Vince

[N4NV]

It turns out that I did not have a noise issue after all. The noise was generated on the break out board itself. I am using a different version of the board now and everything is going great.

I installed the latch on the pendant box. This enables me to open and close it easily to turn on the computer.

Thursday is moving day. I re-arranged my garage to make room for the mill. I am having a machine mover bring it from my work to my house. I removed the head to I can fit it through the garage door. Once I get the mill in place I will have to hoist the head up from the roof joist and set it in place. It is so tall that the head will have to fit between ceiling joist. I will take some pictures tomorrow and again on Thursday.

Vince

[N4NV]

The machinery mover I hired picked up the mill at my work and delivered it to my house. They place it in the garage, but could not get it into the final location due to the low ceiling. They could have moved it all the way, but that would have cost more. They charged me $350 for a move of 7 miles and about 2 hours total. That was $300 less than any other quote I received.

I moved it by myself about 15' to its final location using pipes, Johnson bar and a pinch bar. To get it off the 4x4 block I had to tie a rope to the blocks while lifting it with the Johnson bar. It weighs 4,000 pounds. After getting it in place, I used my engine hoist and a hoist (attached to the rafters) to lift the head in place, again by myself. I made some custom bracket to lift the head so it would be level as I raised it. The whole process to move it in place and put the head on took about 2 hours.

I have some pictures of the process. My dog watched from the bed of my truck, looking quite bored with the entire procedure.

Vince

[N4NV]

Just an update since my last post. I built a plastic housing around the bed to contain the coolant. It works well with very little coolant getting out.

I had it running as fast as 450 ipm (in its original configuration it would go 100 imp), but it was going to take a lot of tuning to get it to run like that consistently. I slowed it down to 270 ipm and have not had so much of a hiccup.

I am using a CNC4PC modbus board to give me 12 more inputs and output. I need these to control the coolant, external spindle start and stop was well as a few buttons I am adding to the control panel.

Here is a video of it moving along as 270 ipm.
http://video.google.com/videoplay?docid=-3686030068739958567&hl=en
Vince