[dodger889]

yes! or at least the back side.

[jsantos]

Rail cutting experience (inexperience?)...

Since I went from a 5X10 to a 5X5 table, I have to cut a couple of my linear rails from 3000mm to 1700mm. Originally, I had this brilliant idea of cutting the rail with a carbide end mill on my machine until I found out that the rail is harder than carbide and won't work. Okay, on to plan B - use a abrasive chop saw.

Went to my local rental place and rented an abrasive chop saw. When to HD and got a cutter. I wanted the thinnest cutter I can find but they only had a 1/8" cutter so I got that.
Got everything set up and started cutting. I've never used an abrasive chop saw before and certainly have never cut into hardened steel before, so I was very surprised at how much sparks were flying! Okay. Move the thing to a safer area where the sparks won't set anything on fire. Start cutting again. And again. And again. I knew that hardened steel is hard to cut that this was ridiculous. After 5 hours (yeah, 5 hours) of cutting, I had gone about 1/4" deep! This is not going to work!

Did a search on the net and found a video on YouTube of a guy cutting hardened steel using an angle grinder. The grinder went thru the hardened steel like butter! I didn't want to use an angle grinder to begin with as I know I will never be able to cut the rail straight. But at this point, I'll try anything.

Went to Northern Tool to pick up an angle grinder. Since I don't know a single thing about angle grinder, I got a sales person to help me. The guy was very nice and told me more than I would ever want to know about angle grinders and pointed me towards a Northern Tool brand called Klutch. I thought that it was a very strange name for a tool but the sales person convinced me that it was a very good tool for $60. The Milwaukee one that they had was about $100. Got home, opened the box and no handle! Since it was late at night, I had to wait till the next morning to go back to Northern Tool. I really wanted to finish cutting the rails!

The next morning, went back to Northern Tool to get a replacement. They didn't have any. The sales person took the handle off the display and gave that to me. Since I was anxious to get going, I took it and went home. Now put the cutter on the tool. What? I need a special wrench and I have no special wrench in the box either! Went back to Northern Tool and returned the thing. Told them that this is the first time I've bought a Klutch tool and will never buy another one again.

Went to HD to look for an angle grinder. They had Milwaukee, Makita, DeWalt and Ryobi. The Milwaukee was the most expensive one and the Ryobi was the cheapest. The Ryobi looked exactly the same as the Klutch that I just returned. But this one was for $40 and came with a case and different cutters. The HD associate said that the Ryobi was not as good as the Milwaukee, Makita or Dewalt but I like that I can put the handle on top instead of the side since I was cutting vertically. And you can rotate the switch too. Those features sold me on Ryobi. Anyway, I don't expect to be doing a lot of grinder and since it was the cheapest, I got the Ryobi.

Now for the real test - start cutting. The thing is going thru the rail like there's no tomorrow! Yay! Got thru the cutting in no time. As predicted the cut was not straight, but I'll take it.

[jsantos]

Just finished cutting, squaring and drilling anchor holes in the extrusions. Acquiring an old lagun cnc has been god-sent to this project. See the first 2 attachments.

Having the extrusions ready, I decided to experiment and find how exactly how much sag I will get from a 10-ft long extrusion. I mounted the extrusion about 6" from the floor, put a dial gauge underneath and hopped on the extrusion. I'm about 150lb and I figure that the gantry won't be much more than 300lbs.

The next 2 attachments show the results. The first picture is the reading before I hopped on the extrusion and the second reading was taken when I was on the extrusion. The sag is about 0.09" in the middle. The 80/20 deflection calculator gave me 0.181" on a 10-ft span, double the reading that I got! Looks like 80/20 is very conservative on their calculation.


Then, I added the rail to see if the rail would help stiffen the extrusion. The last 2 attachments show the result. The sag is now about 0.07". The rail stiffened the extrusion but not enough. I might try to add the CRS later on and see how much that helps.

Based on the 80/20 deflection calculator, adding another leg in the center will reduce the sag to 0.021" sag. Adding a 4th leg will reduce it to a negligible 0.006".

Since driving the x on both sides will give me more flexibility in controlling the sag, I'm going with this option.

[jsantos]

Spent most of the day redesigning the CNC. The attached is the partially finished design.

I boxed in the gantry and will replace the uprights with 1" plates.

But driving the x on the sides will present another problem. It will leave the rails and ballscrews exposed to sawdust. Given what I've seen with my first cnc, those sawdust really like to go over the side. Anybody have any thoughts on how the rails and ballscrew can be protected from them sawdusts?

If I can't find a solution, I think I might go 5X5 instead of a 5X10. It's quite a change but looking back now, I've only cut 3 full sheets of 4X8 in the four years I had my first CNC.

Can you cut a linear rail with a cutoff saw? Will the heat generated affect the rail?

[dodger889]

If you cut them it would only effect about an inch at most. So you just cut let them cool then cut again repeat until you are all the way through. Then you will not see any damage at all.
For protecting the ball screws just add a shield to keep most of the dust from falling on them. And a dust brush to keep them clear could help too.

[jsantos]

Happy New Year, All!


Based on the feedback from some people, I did a redesign. I decided to go 5X5 instead of the original 5X10. I convinced myself that I really did not need a 5X10 as I have not been doing a lot of sheet goods work anyway.

Based on people's suggestions, here are the changes I made:

1. Shorten the X from 10' to 5'.

2. I will go with 3060 instead of 3030 to stiffen the X some more. Based on 80/20's deflection calculator, this will give me a sag in the middle of 0.001" (supported two ends) or 0.0002" (fixed two ends). Don't really quite understand the difference between the two. If the worst case is 0.001", I can live with that. Besides, based on my experiments with deflection, I think that the deflection calculator is conservative.

3. As Gerry and Dodger suggested, I boxed in the gantry to stiffen it up. It will be interesting to see if I can accurately drill and tap the holes in these plates.

4. As Dodger suggested, I changed the uprights from 3030 extrusions to 1" plates. The shape of the uprights was dictated by where I calculated the center of gravity was going to be. I didn't think I was going to be over 300 lbs on the gantry but based on weighing and calculations, it's going to be a 400lb gantry.

The gantry still looks too tall. I cannot bring the lower rail lower but I can lower the upper rail. Should I do that? How close can the rails be? Given that my Z-axis is very tall, does that impact how far apart my y-axis rails can be?

Although Gerry suggested that I go with 1" plate on the spindle plate, I will go with the 5/8" since I already have that. I can easily replace that later on if I need to.


Thanks for the suggestions on how I can improve this machine. Any other suggestions before I go gung-ho on building this?

[tahustvedt]

The table is looking great, but I think you will need two ball screws to avoid racking since the X bearings are so close together.

Are you planning to support the bottom of the X axis stepper/ball screw plate? It will flex if it's not supported.

[widgitmaster]

The table is looking great, but I think you will need two ball screws to avoid racking since the X bearings are so close together.

Are you planning to support the bottom of the X axis stepper/ball screw plate? It will flex if it's not supported.

I agree with tahustvedt, when the screw pushes against the gantry, there is nothing to counter the force! You need to add triangle gusset's on each end plate!
Widgit

[jsantos]

Finally completed the y-axis plate and the z-axis back plate. I had to drill holes on the sides of the y-axis plate and couldn't do it on my CNC so I had to enlist the help of a friend who's got a very nice drilling jig from Bridge City Tools. The first picture shows the y-axis plate with the jig attached ready to drill the side hole.
The second picture is a close up of the jig. My friend was giving me a hard time that his lowly jig can do what my CNC cannot do. :-)

The third picture shows the y-axis plate, the z-axis back plate and the z-axis attached to the gantry. The fourth picture shows a close-up.

I'm a little concerned with my y-axis. While there is no binding and it moves smoothly, it seems that I have to put a little more effort to move it than I do when I move my x-axis. The effort, however, is the same effort when I tried to move all 4 linear bearings with a piece of wood without the y-axis plate attached to them. Hopefully, my linear bearings are not bad.

[jsantos]

Hadn't planned on putting supports at the bottom of the end plates but now I will. I have to remachine the plate with the motor to make my idler wheel adjustment hole bigger so I will add holes to make provisions for adding support at the bottom. The timing of your comments are perfect. :-)

I had not originally planned on using two ballscrews to drive the x- but it's been in the back of my mind. I've been researching how to drive two screws with a single servo in case I'm forced to do that.

[ger21]

Wrap a wet cloth around the rail near the cut to keep it from getting too hot.

[jsantos]

Wrap a wet cloth around the rail near the cut to keep it from getting too hot.

Did that. Wrapped clothes on both ends of the cut soaked in coolant. And stopped once in a while and put coolant on the cut. Kept the rail warm but not hot. Didn't do much to cut though.

Interestingly enough, there was not that much heat using the angle grinder.

[jsantos]

Finally finished the controller. Powered it on and none of the magic smoke escaped. :-)

Here are pics and commentaries:

1. Making the connectors for the encoders. Got some serial cables from a local surplus store and chopped off one end. The owner said that he hasn't had anybody ask him for serial cables in a long time. I wish that parts suppliers for cnc hobbyists would get away from using db-9s. They are so flimsy! The db-9 connector with the breakout board came from China thru eBay.

2. The completed controller.

The box is 3' long by 2' high and 10" deep made from aluminum extrusions and aluminum sheets.

The white stuff on the right hand side are the circuit breakers. I'm using only single phase power but I got these 3 phase circuit breakers from eBay for cheap so I used those.

The blue stuff at the bottom right corner are the power distribution blocks, from eBay.

Beside the power distribution blocks along the bottom are the power relays and the emi filters for the DMM drivers.

I found the interface modules (the green things on the DIN rails) in the local surplus stores so I decided to use those. Made life so much easier for connecting things. They are used for COM and power connections.

On the left hand are 2 power supplies. Because of the mismash of components I'm using, I have to supply 5V, 12V, 24V, and 48V power.

At the center are the 4 DMM DYN3 servo controllers. I intend to add a 4th axis in the near future. If I need to, I can add another controller to drive another motor.

On top of the servo controllers are the Ethernet smoothstepper and the DMM BOB. It will interesting setting these up as I know nothing about the smoothstepper at this time. :-)

3. The front panel

Nothing special here. Just the on/off switch and the fan filter. At one point, I considered adding meters but decided that they did not serve any useful purpose for me.

I got the idea for the fan filter from Gerry (Thanks, Gerry!). I remembered him talking about using an automotive filter in one of the threads and decided to use something similar. A walk down the aisles

of Fleet Farm yielded this nice filter. I had to make a holder for it though.

4. The filter holder and intake fan.

5. The back panel.

The black connector at the top left is a 5-pin neutrik that will drive my VFD.

The 5 terminal blocks at the left hand side are the connectors for the servo motor power. At this time, the topmost terminal block will provide power to the solenoid that I'm using to build my 'brake' for my z-axis.

The 5 connectors in the center are the db-9 connectors for the motor encoders.

The silver-colored connector at the top right is the Ethernet connector for the smoothstepper. The rest are neutrik connectors for estop, limit switches and probe.
You might notice that the bottom connector is crooked. I run out of my neutrik connectors and had to get another one. I didn't realize that the one that I got has a different mounting profile than the rest of my connectors. Oh, well.


Now, on to testing and see if this thing works!

[jsantos]

I'm using shielded wire for my limit switches so I was doubtful that it was noise causing the problem. But when I looked at the diagnostic screen, my limit switches (all of them) were flickering when the axes came close to the limit. What??? Taking a close look at the sensors, I can also see the LEDs on them flickering.

I then disabled the limit switches and moved the axes so that the magnets were on top of the sensors. The flickering stopped. Next, I moved the magnet off the sensor and stepped jog back until I see the sensor light up on the diagnostic. The light on the diagnostic screen is flickering. It looks like the limit switch is triggered as soon as the magnet is just on the cusp of being detected by the sensor but not close enough that the sensor is turning on and off.

Doing a search on this problem, it looks like this really happens with hall effect sensors. I'm surprised that nobody has run into this when people were working with these sensors with RomanLini sometime back in the "Electronic Sensors Made Easy" thread. They now have sensors with Schmitt triggers in them to prevent this from happening. I know the sensor I'm using (Hamlin) does not have this. Looks like I'll have to get the ones with the Schmitt triggers and make new limit switches.

Is there a way to turn off the limit switches in Mach3 during the homing process and then turn them back on after the process is completed?

[widgitmaster]

There are different grades of Chop-Saw discs, some are designed for cutting thin re-bar & thin-wall tubing (most common).
It is difficult to locate a disc especially for thick solid hardened steel! I know, as I bought three different discs just to cut my rails!
The wrong disc will just heat up and melt in the steel, then they stall the saw!
Widgit

[jsantos]

There are different grades of Chop-Saw discs, some are designed for cutting thin re-bar & thin-wall tubing (most common).
It is difficult to locate a disc especially for thick solid hardened steel! I know, as I bought three different discs just to cut my rails!
The wrong disc will just heat up and melt in the steel, then they stall the saw!
Widgit

I would have ended up buying different types of disc if I could find them. :-)

Yup, I had the wrong disc. It just produced a lot of sparks and a lot of heat and didn't do much else. Part of the learning I guess. That's why this hobby is such a wonderful way to while away your time. :-)

[jsantos]

Hey Jordan,

My CNC mill is primarily used for metal working and the one I'm building will be used for woodworking. Just a hobby. :-)

Yeah, I still remember Quiapo and Laguna. In fact, my brother who retired last year decided to return to the Philippines and is now living in Laguna.

I'm finishing up the electronics and I'm hoping to have the machine running in a week or 2.

[pippin88]

Are the bearing models the same? Most are available in different amounts of preload

[jsantos]

Are the bearing models the same? Most are available in different amounts of preload

They're both THKs but different models. The x-axis is SHS35 and the y-axis is HSR35.

[jsantos]

I removed all traces of my old Mach3 just to make sure and installed the latest version. I also installed the latest plug-in of smoothstepper. Powered it on and run Mach3. Mach3 sees the ESS driver and I configured the Ethernet port. Mach3 cannot communicate with ESS. Checked all the connections, make sure that Mach3 and the smoothstepper configurator can go thru the Windows Firewall, and checked all the Ethernet settings. Nothing!

I got the ESS from Soigeneris so I emailed Jeff and told him the problems I'm having getting Mach3 to communicate with ESS.

In the meantime, I connected the DMM BOB directly to the printer port to test it. Also nothing. I have been able to make the motors work before using the DMM supplied software for testing and tuning the motors. But I cannot get the BOB to turn the motors. I called DMM and told them the problems I'm having. They told me that they will send me a new BOB.

Jeff got back to me in a couple of hours and told me to install an older version of Mach3 and also sent me the latest driver for ESS. He said that there are problems with the latest version of Mach3. Installed v.062 of Mach3 and the driver he sent me. Mach3 still couldn't communicate with the ESS. Emailed Jeff again and sent him screen prints of what I'm seeing. Got an email back saying to send back the ESS and he'll send me a new one. Because I'm in crunch mode now, I ordered another one instead and told him to credit my card when he receives the old one.

Just for the heck of it, I tried running the DMM BOB again using the older version of Mach3. To my surprise, it worked! I called DMM to tell them now to ship a new BOB because I have my BOB working. I explained to them that using an older version of Mach3 did the trick.

A couple of days later (today), I got the new ESS (Thanks Jeff for the quick response!). I removed the old one and installed the new one. Wonder of wonders! The thing works! Mach3 sees the ESS and can communicate with it.

Next, I hooked up the ESS to the DMM BOB. By now, I've installed 3 motors on my CNC so I only have one motor left that I can use for testing. Loaded Mach3 and tried running the motor. It runs! And then all of a sudden, it stopped and a light on my servo driver turned on, indicating a fault. Turned everything off and on again. The same thing. Moved the motor to another driver - same thing.

I called DMM and they said the light indicates a motor failure. They asked me to return the motor and they will replace/repair it. It's a brand new motor. They should replace it. We'll see what they do.

I still have 3 other motors left so I should still be able to continue putting together my CNC. That is, assuming that the 3 motors continue to work.

And to think that everything has gone very smoothly until now. Sigh...