[TangentAudio]

A few more pics after painting last night. Not going to win any awards for my rattle can skills, but it already looks better than it did coming out of the crate.

Attachment 212028 Attachment 212030 Attachment 212032

[TangentAudio]

Don't mind the username change. I just decided it was time to get my CNCZone name in line with the name I use everywhere else online these days.

[TangentAudio]

Interesting. I noticed the spindle offset difference between the two heads was not as bad as I thought. The Tormach drawing must be wrong. Looks like the Tormach head will be about 8mm further from the column, not 17mm or so that I had originally calculated.
Attachment 212134

So, let's get this thing mounted and reassembled. Re-mounted the spindle, put on the belt drive sub-plate, and wired up the motor to the VFD.
Attachment 212136 Attachment 212138 Attachment 212140 Attachment 212142 Attachment 212144

So another serious WTF. I tried fitting an R8 collet into the spindle and it didn't fit! Turns out the locating pin for the collet was too long. What the heck kind of quality control is this? Thankfully I got the pin down to the right length with my rotary tool and a small grinder. Here's a mirror shot looking up into the spindle, before I ground the pin down.
Attachment 212146
Even with the locating pin ground down to the right length, the Tormach TTS collet binds up once it's about 3/4 of the way inserted. I eventually got it to fit, but it took pulling it in with the drawbar. Thankfully it didn't get stuck, but it's too tight. I tried one other regular R8 collet that I had around and it binds in the same place, but I couldn't get it to pull in. I guess I will need to get in touch with Tormach. I can deal with cosmetic problems in stride, but this impacts actual performance and it's unacceptable for something this expensive. This is the actual precision part for crying out loud.

But given that I could get the R8 TTS collet to fit... You know what that means. Time to make CHIPS!

Keep in mind this is before any tramming, it's just a random test. Using a 0.375" 4 flute carbide endmill, 0.100" depth of cut, 0.375" width of cut, 5300 RPM, 52 inches/minute. Material is a scrap of 2024-T351.



VERY promising results, the thing just spews chips out like crazy. And they look like proper chips! I've cut lots of aluminum on this machine over the last couple years, but the chips never looked like this, and I was never able to take cuts this aggressively. The cut just sounds a lot healthier too - not having that horrible gearbox in the stock X3 is really nice.
Attachment 212148 Attachment 212150

The finish was not bad given this was fairly aggressive and it's not trammed. The scrap was partially unsupported in the vise, so the finish varies depending on the position of the cut.
Attachment 212152

[TangentAudio]

Another quick test, using small cutters at high RPM. This is a 0.1875" 3 flute carbide endmill, 8000RPM. Around 0.500" DOC, different widths of cut via manual jogging. 24 inches/minute feed. Material is 2024-T351 aluminum.



That's definitely another thing I could never do on the stock X3. Damn!

Finally, A couple of parting shots for the evening.
Attachment 212154

[Fenza CNC]

Again looking brilliant mate, bet you're really pleased :cheers:

Are you going to get the Tormach Superfly Cutter? I've only ever heard great things about that little tool, the finish should be brilliant if you really put the effort in to get the spindle trammed.

Fenza.

[TangentAudio]

Fenza, I would like to get a face milling tool and the SuperFly is on my list to consider. I've also heard of people using the Glacern insert facemill, with only a single insert in it.

[Topdecking]

Any reason not to use just a normal 3 insert facemill?

[TangentAudio]

I've read that Tormach did a bunch of testing and found for low horsepower machines that a single cutter produced the best results. I don't have any personal experience with them, so I don't know if this is true or not.

[Topdecking]

Very interesting, I haven't tried out a facecutter and on my X2 I've been using a 3 insert facemill... The finish is nice, I can't imagine it getting better.... I suppose I'll have to try it or wait for better resources to chime in.

[TangentAudio]

This is the article I read. Could be marketing drivel, but it does make some intuitive sense.

TTS SuperFly Cutters: A Modern Update to an Old Shop Standard « Milling Around

[TangentAudio]

Tormach does excel in customer support. I contacted them about the collet fit issue and an engineer suggested that I double-check the pin length (after apologizing that I had to grind it down in the first place). He said it was a problem with some early spindles, and hadn't seen that issue in some time.

In digging through my box of R8 collets, I found one that actually seemed to fit fine, I had one that bound up a lot, and the TTS collet bound up a little but still worked. Sure enough, there was a correlation to the depth of the key slot. The best had about 0.010" more clearance than the worst. So, after several iterations of grinding a little more off of the pin and test fitting, and now they all fit like they should.

[TangentAudio]

In my talking with Tormach, I asked about a spindle break-in procedure. They told me it should have already been done, but it wouldn't harm anything to do it again. They suggested running it in 1000RPM increments, and measuring the spindle temperature with a non-contact IR thermometer. Once the temperature stabilized at a particular speed, bump it up to the next one.

I figured I'd one-up it a bit and get some IR imagery while I was at it, as well as get thermocouple data off of the stationary part of the spindle nose. Here is my test setup. It's just my beater multimeter with a temperature probe, and a Flir i7 handheld IR camera I borrowed from the lab at work.
Attachment 213168

I collected IR images and data at each speed level and did some charting and graphing during the break-in process. I've attached the PDFs with all of the data and pics if anyone is curious to see. Otherwise, here are some random IR images.

Spindle at room temp (left), 5000 RPM (middle), 10000 RPM (right). It did actually get warmer than 181 (into the low 200's) but curiously dropped down with time.
Attachment 213170 Attachment 213172 Attachment 213174


Here's the whole spindle head on the left, and the VFD on the right.
Attachment 213178 Attachment 213180

[TangentAudio]

Not much to look at, but I made some good progress this weekend writing the Modbus driver for the VFD. Took a lot of extra work because I jumped up to the latest and greatest bleeding edge development version of LinuxCNC and ran into some snags along the way. With a little persistence and some help from the folks on the IRC channel I got through it.

Spindle control is roughly working now within LinuxCNC, though I still have some fine tuning to do with stuff like multipliers for belt drive ratios. Got some nice feedback of stuff like motor load so I can show that in the GUI on a meter.

[TangentAudio]

Tonight I did some very preliminary measuring for tramming the head and column. Mostly just practice to start wrapping my head around it.

I started with the DTI in an ER20 collet holder in the spindle. First up is checking the Y axis, fore and aft. The back measurement is down .005" over ~7.25", which I think means the head is tilted back.
Attachment 214266 Attachment 214268

Next, the X axis. I actually trammed this in by loosening the head mount bolts and tapping the head with a soft hammer, so it rotated around the boss on the Z slide. Things moved a bit as I tightened the bolts, but by doing a cross-torquing pattern and going a little at a time, I got it to be within 0.0005" over ~7.25". Could do better, but it was just a quick 2 minute job. No sense in going nuts since I'll have to take it apart to shim the fore and aft to fix the error above.
Attachment 214270 Attachment 214272


Next up, an attempt to see how the column tilt is looking. I used my granite square, mounted the indicator on a mag base on the head, and ran the head up and down with the Z stepper. The Y direction is actually pretty good as-is, it's not even out 0.001" over ~9"
Attachment 214274 Attachment 214276 Attachment 214278

X axis isn't so good though. Seems like the whole column is tilted to the left, I'm measuring 0.015" over about 8". Guess I'll need to shim up the left side.
Attachment 214280 Attachment 214282


I plan to do the measurements again, then take some other measurements so I can do some trig to get some approximate values for shimming. I picked up a shim assortment from McMaster a few weeks ago, so I should be good to try once I come up with some numbers.

[TangentAudio]

I had a thought in the middle of the night last night... Perhaps that 0.015" I'm seeing in the left-right direction is not column tilt, but some combination of gib fit and wear in the ways. Maybe I should also expect to also see fore-aft movement if this is the case?

I'll have to think a bit more to see if there's a way I can determine the source of it, though there's really little harm in shimming up the column anyway. Easy to undo if I determine it's something else.

I do have that new, unfinished gib sitting on the shelf. I should really get that thing fitted before I go crazy with tramming anyway.

[TangentAudio]

So I've had this thing sitting on a shelf for a couple of years now. It's a long table that Sieg was putting on their "SX3L" machines for a while. I managed to snag one from Little Machine Shop. At 27", it's about 5" longer than the stock X3 table. I expect to be able to get around 20" of travel out of it, versus the ~15" I get now.
Attachment 214326

Obviously changing to a longer table means I need a longer ball screw for the X axis. I didn't have a lathe a couple of years ago, so that's why this project went onto the shelf. I decided it was time to finally get this part of the project moving, so I tracked down the Thomson 0.631" 5TPI lead screws that are used in my ProMiCA conversion and ordered two: one to screw up practicing, and one to hopefully be the good one.

I am not sure yet if I'll try to anneal the ends before machining, rig up some kind of tool post grinder, or just plan to burn up some inserts. From what I've read the two big potential downsides to annealing are potentially warping the screw from heat, and softening the metal on the actual screw part making it wear quickly when in use. I think the former is the bigger worry for me, as the very ends of the screw don't really engage in the ball nut.

---

In other news, I managed to get a good ebay deal ($100 + reasonable shipping) on a Hoffman A241608LP steel enclosure (24" x 16" x 8"). My electronics are already nicely buttoned up in a rack mount enclosure, but it's chock full. Now that I have the VFD, I'd like that to be enclosed. I also have grown less fond of the rack mount enclosure over time. I have plans to build a stand and an enclosure for the mill this winter, especially now that the new Tormach spindle spits out chips at such an alarming rate. I prefer the idea of having this new enclosure hanging off of the side, high up and out of the way of coolant and chips, like most other machines.
Attachment 214328

Here's a video I did almost a year ago detailing my electronics as they stand now. I am not sure what (if anything) I'll change as I move them over to the new enclosure.

[uawgmsmco]

holy progress! nice work....for all the mods you've done I'd say your tram issues are minor. I'm anxious to see the new big ol'-big ol' table. Now if only the X3 had just a liiiiitle more Y travel eh?

[TangentAudio]

holy progress! nice work....for all the mods you've done I'd say your tram issues are minor. I'm anxious to see the new big ol'-big ol' table. Now if only the X3 had just a liiiiitle more Y travel eh?

Thanks! As far as Y travel goes, the ProMiCA conversion kit that I originally used for the conversion did increase travels a bit, from 5.75" stock to 7.5". Sure, a little more would be better, but this isn't so bad.

My big concerns with the long table is that it won't be flat, or it'll sag a lot when it's hanging out of the saddle. I think I heard a rumor that the flatness issue was one of the reasons Sieg stopped offering the longer tables, but I don't know if it's true or not.

[uawgmsmco]

I didn't realize the Promica kit gave 7.5" on the Y....I think I measured about 6.40" on my old CNC Fusion kit.


I could see some saggage issues with the bigger table and increased sticktion. Buuuuut I guess it's all about what you're usin' it for too!

Cheers!


Andrew

[TangentAudio]

Received another item today: a 10 foot length of Belden 29501 VFD cable. There's no messing around with this stuff at ~$5.00/foot (purchased on ebay). It's a bit more stiff than I'd hoped, but I am sure I can still make it work and feed from somewhere on the column into the box on top of the head. I am not sure if it would work in an Igus cable carrier.

As you can see it has three conductors for the motor phase wiring, a separate insulated ground, an uninsulated drain wire, foil shielding, and an outer braided shield. Should help eliminate any potential EMI issues coming from the VFD wiring.
Attachment 214464 Attachment 214466 Attachment 214468

I would rather have found Belden 29500, which is 16AWG (more than sufficient for a 1HP motor) instead of 14AWG, but I couldn't find it without having to buy a full spool at ridiculous cost.