Build Thread - 80/20 5X10 Woodworking CNC Build

[erikjgreen]

I use a big SSR (solid state relay) on my VFD. It's controlled by 24v switched by my Mesa I/O card. The VFD doesn't have power until LinuxCNC is up and switches the bit on the I/O card that enables the SSR to power up the spindle VFD. In case of E-stop, it is (actually "will be") set up to cut spindle power and apply the brake.

Erik

[bobmagnuson]

I was fortunate enough to spend a couple hours with Joe yesterday afternoon. All I can say is WOW!!! His machine is spectacular!!!!

[jsantos]

Thanks, Bob!

Yes, my machine is finally working and I'm very pleased with the result!

Over the next few posts, I will go over the issues that I have run into that I have not posted here and the things I had to do to get around them. Will also post pictures.

[jsantos]

I did post this issue earlier. Whenever I turn on my controller and Mach3 is not running yet, the spindle would start spinning.

A search in the Warp9 forum (I'm using smooth stepper) yielded a thread on this issue. The issue is basically in the BOB where it does not pull down the pin that turns the spindle on/off. DMM Tech did not reply to my emails on this and did not provide a solution.

Gerry and Erik Green suggested that the VFD should not be turned on before Mach3 is in control. My VFD is powered via a power relay. The relay is connected to my 12+V supply and is energized every time I turn on the controller, thereby turning on the VFD.

Luckily, the DMM Tech BOB has a relay that turns on only when the charge pump is running. I connected my VFD power relay to this relay and it solved my problem. The VFD is not powered on until Mach3 is running and the charge pump is on. Thanks, Erik!

This is just a workaround as far as I'm concerned. For the sake of other people buying DMM Tech's BOB, DMM Tech should fix this design flaw in their BOB.

[jsantos]

My hall effect sensor homing/limit switches was not working properly. When I home, it starts to home until it hits the sensor and then stops. The LED on the sensor would then flicker.

It turned out that I had 2 issues. The first was the sensor itself. I was using a Hamlin sensor and it apparently does not have a Schmitt trigger that will ensure that the sensor stays on as soon as it detects the magnet. After a lot of research, I ended up buying hall effect sensors that are designed for Arduinos - for $0.99 each from China thru eBay.

In the first attached picture, the sensor on the left hand side is the Hamlin sensor. Big mistake using that one!. The one in the middle is the sensor I got from eBay. It's got an LED and pull-up resistor built into it. I just removed the pins and replaced them with wires as shown on the right hand side.

The other issue I had is with the power to the sensors. The DMM Tech BOB provides power pins for supplying power to the limit sensors. I used those as it makes the wiring very convenient. Unfortunately, it turned out that the BOB does not have enough juice to power the sensors. I connected the sensors directly to my 12+V supply and it now works great!

I couldn't believe that accurate these sensors are. I remember in one of Gerry's post where he was saying that he spent an hour (?) homing his CNC again and again when he first installed these sensors. I made a smart ass comment that he must not have anything else better to do. Well, now I understand! I did the exact same thing when my sensors finally worked. I kept homing my CNC because I couldn't believe that it gives me the exact same position ever I home!

The last picture shows the z-axis limit sensors lighted up in all its glory!

[jsantos]

Erik came over a couple of nights to help me tune my servos (More on servo (non-) tuning later).

One of the things he commented on was that my pulley belts were too loose. So we proceeded to tighten all the belts. We tightened them by 'feel'.

A couple of days later, I was working on fine-tuning my steps/per calculation. I was going to let Mach3 calculate my steps/per for me. Everything was going hunky dory and all of a sudden, I heard a big thump. I looked at my CNC and saw my z-axis was almost to the table. Attached to the spindle was my Haimer Zero Master and I saw that the tip was broken. It turned out that the shaft of my z-axis ballscrew broke! The pulley belts were too tight! Ouch! $50 for the zero master tip and $100 for the ballscrew!

While the ballscrew is 15mm thick, the end of the ballscrew was only about 7mm in diameter!

So I ordered a new ballscrew and at Erik's suggestion, got a belt tensioner tool. Installed the new ballscrew and used the belt tensioner tool to retighten all my pulley belts. When I turned on the machine, I noticed that the pulley on my z-axis was wiggling! Turned the machine off and look at what's going on. The shaft of my ballscrew did not break but it bent! Arrrggghhhh! Just can't win!

This time, I ordered a 15mm ballscrew from AutomationForLess with a much thicker shaft. I had to mill my z-axis to accommodate the new ballscrew but now the thing is working just great!

At the top of the attached picture is the Haimer zero master. It's a very accurate tool for finding edges. The tip is designed to break in case of operator error and replaceable but very expensive to replace!

Beside the zero master, is a Gates belt tensioner tool. I was able to get the belt deflection and belt tension from the Gates web site.

At the center of the picture is the ballscrew with the broken shaft. At the bottom, notice how thin the shaft is. Like I mentioned, the shaft of the ballscrew at the bottom did not break but it bent!

Sigh!

[jsantos]

Did several NAS979 tests. Unfortunately, I put the pieces in a bin that I usually use for garbage and my wife threw the test pieces away thinking they were garbage, so no pictures.
But here's what happened:

1. The first test really looked and sounded weird. The cut was very rough and it smelled like it was burning. Then the spindle started tipping! After hitting the emergency switch, I looked at the spindle and it looked like it pulled the screws holding it right out! But taking a closer look, the threads in the screw holes looked good. So I put the screws back in and they tightened fine. Then it hit me. When I put the spindle on, I just hand tightened them and I completely forgot to completely tighten them! DOH!

2. Second test - The cuts are still very rough and still smells like something is burning. Take a close look again. The bits are going counterclockwise! Noooo!!!!! Okay, fix that by interchanging 2 wires to the spindle.

3. Third test - Cut the same piece again. No burning and the cut looks good.

4. Fourth test - Run at 300IPM. Works fine, except that you can see chatter marks on the piece and I can hear the ballscrews beginning to complain.

5. Fifth test - Run at 200IPM. Beautiful! No chatter marks, the diamond, circle and square look and measure like diamond, circle and square.

So happy!

[ger21]

It's amazing that you can actually cut with the bit spinning backwards, isn't it.

[jsantos]

So now I have a machine ready to use. The first thing I wanted to do was put some shelves together for one of my workbenches.

Turned the machine on and zeroed it. For some reason or another, it sounded funny. After zeroing it, I moved it to do a tool change. All of a sudden, the machine stops! I look at the controller, and the x-axis driver fault light is blinking!

Turned the machine off and turned it back on again. Tried homing I and the x-axis driver faults again!

Turn the machine off again and tried to turn the x-axis by hand! It won't. The thing has seized! Oh, no!

Disconnected the x-axis from the ballscrew and the x-axis moves very smoothly. Tried turning the ballscrew and it wouldn't turn. Noooo!!!! Started disassembling the thing. As soon as I removed the pulley, I saw what happened. The setscrew of the nut holding the thrust bearings had gotten loose! See attached picture. Just couldn't do anything right!.

So I had to remove the ballscrew, tighten the nut and loc-tite the setscrew. Just to make sure, I removed the y-axis ballscrew, and sure enough the setscrew on that one was coming off too. Fix that too and put loc-tite on it too.

Try cutting again. This time, it cut the shelves very cleanly and very accurately. So happy again. See last picture of shelf.

After cutting the shelf, I decided to surface my table. I run the surfacing at 200IPM and the machine did a very good job of surfacing the table at that speed. I could not see any tool marks on the table.

[jsantos]

Okay. So it looks like I finally have a running machine and running very nicely at that. Here's what I have so far:

1. Machine runs whisper quiet at 100IPM, lets you know it's there at 200IPM and the ballscrew lets you know it's working at 300IPM. The motors supposedly can run at 700 IPM but I guess I'll never get there. I should be able to run continuously at 200IPM. I'm okay with that.

2. Backlash is 0.04MM for both x- and y-axes, 0 backlash for z-axis. Yes, I'm actually using metric. :-)

3. I've put setscrews on the spindle mount so I can adjust it as necessary but haven't had to. Count the small blessings!. :-)

4. Sensors are working perfectly and incredibly accurate. I've got an x- and y- fence attached to the machine and the position never changes for those. :-)

5. Auto-zeroing with Gerry's 2010 Screenset is beautiful! Thanks, Gerry and BigTex(?).

6. No racking on the x-axis even with just a single ballscrew. Very relieved with that.

7. Machine doesn't move when running. Added braces all around and bolted it to the wall.

Here's my remaining issue:

The motors are not perfectly tuned. I know that some people have said how easy it is to tune the DMM Tech motors. Unfortunately, I did not find it to be the case. Tuning these motors is entering some numbers in a very clunky software and seeing and hearing how the motor reacts. It would be good to hear from people who were able to tune their motors properly. Please tell me how you did it!

In my case, my motors are always hunting and there's hum coming from the motors all the time. It doesn't seem to affect the performance of the machine, but it would be nice to be able to get these motors to quiet down.

Final pictures!

[jsantos]

It's amazing that you can actually cut with the bit spinning backwards, isn't it.

Actually, I think it's more like gouging. Really stank up my whole house. :-)