[photomankc]

That's a neat design, but isn't that going to pull all kinds of nasty right into the lower bearings?

[ChrisAttebery]

My thoughts are that the chips are going to be pulled pas the spindle nose with enough vacuum that they shouldn't be pulled into the bearings. If anything air should be pulled down the spindle past the bearings. I could always add a felt shield on the outside of the spindle nose to protect the bearings if it becomes a problem.

There's only one way to find out, right? ;^)

[ChrisAttebery]

I stepped off into the deep end of the pool this week. I have a job to cut 3D shapes out of G10 material. I've never worked in 3D before. It took me a few nights to get comfortable drawing the parts in TurboCAD Deluxe V20. Then I downloaded the trial version of MeshCAM. It is really easy to use, but it does have some limitations. I think I can work around them by using SheetCAM to do drilling and 2.5D ops and only using MeshCAM for the sections that truly need 3D cutting.

I haven't been this excited about machining since I produced my first CNC cut parts. :banana:

[Kevinj308]

I like meshcam too. For drilling it works well at locating all the holes in a model, I then go back and add a g83 to each location. It's still faster than plotting them all out and writing he code by hand.
Just as an aside fusion360 has a free year for startups/hobbyists. I just downloaded it last month and have only tried the cam side of it. It's definitely more capable than mesh but also a little more difficult to use. The tool paths are definitely more efficient. Cut machine time down by 20% or more.
I don't know if I'll pay the fee when my free year runs out, this is just for fun for me. But it's a nice cam package while I've got it.
Kevin

[ChrisAttebery]

I've been side milling 6061 bar stock with a 1/2" aluminum specific YG HSS end mill this week. I needed to remove a lot of material so I've been taking .5"-.625" DOC and .025"-.050" WOC at 27ipm and 4500rpm (shooting for .6cu in/min). Today I cranked it up to 45ipm (1 cu in/ min) and the machine actually sounded better than at 27ipm. :^)

[ChrisAttebery]

I've been trying to use more side milling lately since it allows you to spread the cut over a longer edge on the tool. It's a royal PITA to make work in SheetCAM most of the time. I don't have the money to plunk down on expensive CAM software right now so what can I do? Someone on another forum mention that Fusion 360 is now free for individuals, students and startups making up to $100k. Hey! That's me. I've been playing with it for about a week and cut my first parts with it yesterday. The post needed some work, but Autodesk jumped on it right away and now it seems to be perfect. I'm side milling .480 DOC, .050 WOC at 45IPM and the machine seems happy as a clam.

I also upped my rapids to 240ipm from 120ipm. I had problems with rapids before I installed the SmoothStepper, but I never tried retuning them after I installed it. You can really see and hear the difference with it installed. It moves with authority now and as advertised it sounds smooth as butter.

[ChrisAttebery]

I really haven't done any maintenance on my machine since I got it up and running and I've done quite a bit of machining with it over the last 3 years. Lately I've heard a bit of bearing noise from the spindle. So I tore it down and found that the lower bearing had a rough spot. I ordered a pair of new bearings from VXB tonight.

[ChrisAttebery]

The bearings arrived on Saturday and I had them installed about 45 minutes later. The noise has been reduced but did go away completely.

Meanwhile the PC I've been using as a controller started acting up a couple weeks ago. It would blue screen every few days, but it has gotten to the point that it crashed 2-3 times over the weekend. I tried running memtest86 and chkdsk on it but I think the motherboard is going out. I'm going to pick up a newer SFF Win7 32 bit system ASAP.

- - - Updated - - -

The bearings arrived on Saturday and I had them installed about 45 minutes later. The noise has been reduced but did go away completely.

Meanwhile the PC I've been using as a controller started acting up a couple weeks ago. It would blue screen every few days, but it has gotten to the point that it crashed 2-3 times over the weekend. I tried running memtest86 and chkdsk on it but I think the motherboard is going out. I'm going to pick up a newer SFF Win7 32 bit system ASAP.

[cjsamples]

Good to hear the new bearings fixed you up. Might want to look at the micro computers, picked up a great one for less than $160 and it fits in my cnc plasma controller box.
Fully Assembled PC with M350 Enclosure, Intel Dual Core Celeron (Braswell) 2.16 GHz, picoPSU-80, 60w AC adapter

[ChrisAttebery]

I bought another system off Craigslist for $50. I got Mach3, MSM and the ESS software installed last night. I'm having some trouble getting the ESS to load with my old xml settings file. I got some help on the ESS forum that I'll try tonight.

[ChrisAttebery]

I'm back in business. It turns out I copied the wrong xml file to the new machine. I copied the xml to a new name when I added the ESS last summer, but I didn't rename the desktop shortcut. Once I figured that out it fired right up.

[ChrisAttebery]

It's been a while since I've updated this thread. After I replaced the bearings my finishes really didn't improve that much. I checked the screws for slop and realized that the gibbs were pretty loose. I tightened them up and my finishes improved.

A couple weeks later I noticed that my parts were coming out out of round by about .010" or so. I did some more checking and found that the bearings at the end of the Y screw were getting loose too. I tightened them up and the backlash was reduced to about .002" again.

While I was at it I rechecked the column and found that it was out about .002" per inch in the X direction. I got that dialed in to within .00033" per inch. I checked the Y direction and it was about .0005" per inch.

Since I moved the column I had to tram the head. I got it dialed in within .00014" per inch in the X direction. In the Y direction it was within .0002" per inch.

I'm going to cut a test part tonight and see if the accuracy improved or not.

[lcvette]

Nice!!! I bet it makes a huge difference... I hated when my little piddly spanner nuts would loosen up so I switched them out and used a locking nut (the crimped thread type.. Since then I haven't had an issue with it loosening up.

For your column tram, those are some pretty big errors still. Any reason you can't get it dialed in better then that?

If you had to make anything taller then 2" it may be a problem..

Chris

Sent from my XT1080 using Tapatalk

[ChrisAttebery]

That was supposed to be .00033" not .0033". It was less than .001" over 3" in both directions.

I've never cut any deeper than 1" on my mill. The couple parts that I make that are deeper than that get flipped over and machined on the backside.

Nice!!! I bet it makes a huge difference... I hated when my little piddly spanner nuts would loosen up so I switched them out and used a locking nut (the crimped thread type.. Since then I haven't had an issue with it loosening up.

For your column tram, those are some pretty big errors still. Any reason you can't get it dialed in better then that?

If you had to make anything taller then 2" it may be a problem..

Chris

Sent from my XT1080 using Tapatalk

[lcvette]

That's much better... Lol... Post up the results!!

Sent from my XT1080 using Tapatalk

[ChrisAttebery]

The 3.5" square was 3.497" in Y and 3.500" in X.

The 3" circle was 2.995 in Y and 2.999" in X.

The 1.84" diamond was 1.841" in one direction and 1.839 in the other.

I'm not sure why it's cutting deeper in Y than X unless it's just the head flexing. I used 5mm pitch screws with 400 step/rotation on the motors. I have the steps per inch set to 2032. It might just be the ballscrew. I don't have a long travel indicator to check it with.

[lcvette]

Did you not use the steps per function when setting up your axis? Chinese screws are rarely perfect and have a little lead error. That's what the steps per calculator on the settings page is for.

Sent from my XT1080 using Tapatalk

[ChrisAttebery]

I don't think I did. I think I calculated the steps and then cut some 1" squares to verify my settings.

It looks like it's off by ~2 steps per inch in the Y direction.

(3.5/3.497) X 2032 =2033.75

[lcvette]

If you use a dial Indicator and use the set steps per inch function on the settings page it will dial it in to 0000.0000 resolution j til the Indicator is landing perfectly on your move setting. Just make sure when you set them that you make a pre positioning move in the same direction your test move will be in to ensure you aren't measuring any backlash. I actually also turn backlash comp off when setting steps per. And then retest and reset new backlash figures after I'm done. Remember after you get the steps per for each axis restart mach3 so it writes those settings into memory.

If you didn't do this you will be much happier with your results.. I can hold under .001" usually without issue. Your circles however will be what ever your backlash allows. Also for backlash, you can adjust the shuttle speed on the general config page to have it take up the slack really quick which will help keep your circles more round. Obviously the smaller your actual backlash is the better.

Sent from my XT1080 using Tapatalk

[ChrisAttebery]

I used a 123 block on top of the vise to get a 3" offset from the jaw. I approached the block and zeroed the axis when my indicator was zeroed. Then I moved 3" back to the jaw and measured the difference. I added 1.75 steps per inch on the Y axis and that got me within .0005".

If I get a chance I'll cut another test block in the morning. Hopefully that will take care of the undersized parts.