[TXFred]

I'm starting a new thread because my old one wasn't ever titled as a build thread, but was morphing into one.

The old thread is here.

I'm 90% done with the machining of the Hossmachine parts. I haven't installed them to the mill yet.

This week I've been working on the electronics. I am limited for space in my garage, so I didn't want a bunch of boxes with wires running all over the place. I decided to try and fit all the electronics into a single case.

$30 at the Goodwill Computer Store got me a rack mount PC case with power supply. Scrounging in my neighbor's junkpile got a servicable PC, some extra fans, and a CD-ROM drive.

It took some grinding and filing and bashing on various things with hammers, but here's the final results.

The G540 mounts in the external drive bays. It's a 3/4 height item, which means it takes up one and a half drive bays. (A modern CD-ROM drive is actually a half-height drive. Now you know!)

Behind the G540 is a hard drive cooling fan that blows onto the back of the G540, keeping it cool. Behind that is a modified adaptor that used to allow a 3.5" floppy drive to fit in a 5.25" bay. With some cutting and some taping, it now divides the space behind the G540 into an upper and lower duct. So the fans draw in cool air via the upper duct, blow it against the G540, and it then returns via the lower duct to the main area of the case. Without the duct, the fans would just be blowing the same hot air against the G540, which wouldn't cool all that well.

The DB9 below the Gecko is for the inputs, outputs and E-Stop. I'll add another connector to connect the Gecko's VFD drive to the mill. Note that the DB9 is male while the ones on the Gecko are female. This makes it impossible to accidentally plug the mill's limit switches and other inputs into a 48 volt stepper drive.

The VFD connector will either be another DB9, or something smaller.

Below the Gecko's parallel port connector, I'm going to mount another parallel connector in the bezel. I'll do the same on the back of the computer's case. Currently my 6 foot parallel cable runs from the back of the computer to the front, which looks quite stupid. With the connectors in place, I can make short cables for the front and back, and run the cable through the case, which will be much cleaner.

The silver switch on the top left is turns the power supply on and off. It's switched on the AC side, not on the DC side. I don't want to fry my G540!

The Keling power supply fits into a space that used to hold a large fan and duct assembly. This system was designed to be loaded out with expansion cards and drives, and had a serious cooling system. I don't need all that much cooling, so I could dispense with this setup.

Between the Keling and the motherboard is a steel EM shield made of a scavenged and hammered PC power supply cover. It looks cheap, but it does the job of keeping the transformer's EM field away from the motherboard.

The Keling power supply is cooled by a small fan stuffed into the front left panel of the case. I still want to make a bracket to secure it into place, but it works fine just sitting in there.

Oh yeah, and it all works. The motors spin as they're supposed to when I run the Roadrunner program. So I call it a success!

Cheers,
Fred

[Crevice Reamer]

Looking good Fred! Capital job mounting the G540 INSIDE computer case. Exellent, well-thought-out and implemented work!

CR.

[RotarySMP]

Nice clean controller. Good job!.

[SScnc]

It took some grinding and filing and bashing on various things with hammers

LOL :rainfro:

I agree with CR and Rotary, that is one of the cleanest, best done control boxes I've seen.

Good job !

[pete from TN]

Now all Mariss has to do is build an all in one drive system like the G540 only with a servo control and enough balls to drive a machine like an RF45 or maybe even a knee mill... I know I'd buy one... having all of it done for you and integrated into a nice box like that makes wiring up a cnc controller a breeze... peace

[TXFred]

I spent part of today taking the mill apart and cleaning it up. It was dirty enough that I just checked all the major pieces in the parts washer. I'll clean them up and then modify them for the CNC components.

I need help finding a thread. Another forum member is converting an X2, and he did a brilliant job of mounting the table limit switches under the table, where they'll stay clean and dry. I want to copy his work, but now I cannot find his build thread. If anyone knows who this was, let me know. It would help me search if I knew the username.

As a sidenote, while turning the X axis ballscrew today, I realized just how sloppy my lathe is getting. The toolpost was moving more than .01" when it was cutting through the ballscrew. I did a quick job of tightening it up, but as soon as the mill is done, the 7x10 is going to need some attention.

Fred

[hoss2006]

I need help finding a thread. Another forum member is converting an X2, and he did a brilliant job of mounting the table limit switches under the table, where they'll stay clean and dry. I want to copy his work, but now I cannot find his build thread. If anyone knows who this was, let me know. It would help me search if I knew the username.


Fred

I believe you're looking for bilinghm's thread, has other good mods too.
http://www.cnczone.com/forums/showpo...7&postcount=45

check the whole thread, good mods for the cncfusion kit and a column brace.
Hoss

[TXFred]

While I've got the mill apart, I'm thinking seriously about recutting the X and Y dovetails.

The problem is that the dovetails on my mill were cut with a very worn out cutter, creating dovetails that are rounded rather than flat. Also, the machine that cut the dovetails was badly out of tram.

I find it telling that the Z dovetails, which are visible, are perfect, while the hidden X and Y are done badly. That's Chinese quality for you, I guess. Make it look good and who cares if it works.

I've attached three images. The first shows what the X2's dovetails should look like in profile. The angle is 56 degrees, for some reason that I'll never understand.

The second image is an exaggerated image of what my dovetails look like. And I have already lapped these dovetails, so this is an improvement over stock.

The third image is my proposed modification, recutting the dovetails with a standard 60 degree dovetail cutter.

Recutting the dovetails will shift the table in +X and +Y by .035", which is pretty insignificant. The gap on the gib side of each dovetail will increase by .070", which is more critical. But since I'll have to make new gib strips with a 60 degree angle, it's no problem to make them wider so that they'll take up the extra slack.

Also, and I haven't confirmed this, but I suspect that the X and Y dovetails on the saddle are not perpendicular. Recutting them would allow me to fix this.

I've got access to a Bridgeport, so I should be able to get this job done over the Thanksgiving weekend.

One last bonus is that once the dovetails are at 60 degrees, I'll be able to do fun things like making my own extended X table in the future.

I don't forsee any problems from this, but I'd like to hear feedback and devils advocate comments from people here.

Fred

[Starleper1]

It's good to know my X2 was not the only one with a weird angle for dovetails. I couldn't figure out why either.

[TXFred]

I got my 60 degree dovetail cutter in the mail last night. I'm hoping to do the machining this weekend.

I turned the end on the X leadscrew this weekend, and I feel that I need to comment on something.

I am displeased with the quality of the Roton ballscrews.

When I got them, they had rust spots on them from improper storage.

And there's a worse problem. I didn't know this when I cut the Y axis, because the ballscrew was hidden inside the lathe spindle. But when I set up the X axis ballscrew (20" long since I'm leaving extra for a table extension) the problem became obvious.

The ballscrews that I got from Roton are not straight. Try as I might, I was unable to center the ballscrew. If I got both ends centered, the middle of the ballscrew would wobble off center like a jumprope.

And the fit of the nut on the ballscrew is very loose. I'm going to see if I can load oversized balls to compensate. But I think the eventual solution will be replacement ballscrews and nuts.

I've already cut up the ballscrews, so I'm committed to using them. They'll hopefully be good enough to learn on. To anyone else who considers the Roton ballscrews, remember that you get what you pay for.

Fred

[Starleper1]

I use Rockford ball screws. Automationf4less sells them

[hoss2006]

you're entitled to your opinion but i've turned dozens of sets of roton screws and never saw any
indication of rust spots, their smooth finish is what makes them attractive.
They would have replaced them for you.
They are the cheapest out there so expecting high quality is illogical,
but they are good enough for the hobbyist.
Oversize balls do lessen any backlash, I used some .1259 balls from Toolsupply.
This list can give you an idea of the costs of higher quality screws.
Hoss

[pete from TN]

I gotta agree with Hoss here. I have roton screws on my Lathemaster cnc mill and have been using the machine for a few weeks now. I machined them myself and I also machined my own anti backlash double nut setup using the bellville washer ideas. I have also recently been working on the final tuning of my steps per inch and accurizing the machine in general. This work is not anywhere near done and I am currently seeing some very promising results. I am machining a project right now that consists of three of the same part on the table machined from some 1/4 inch plate aluminum. I have run it three times now and have been fine tuning the code in preparation for a small production run. My steps per inch are slightly off but the consistence of the parts one to the next as well as the repeatability is very good, well within .001 on most parts. My X axis steps are still making my parts slightly undersize, on the order of .002-.0025 but I can tweak that in the software steps easily. I have also read a lot of people poo pooing these screws on here and as I said so far I am pleased. MY goals with this machine were to be able to repeatably machine parts to within .001" on the critical parts. At the onset and during the buildup and also with a lot of the talk on here I thought this was a pipe dream but now I think I will get at least close to that figure barring any mistakes on my part with the cad cam work and using realistic machining operations. If these screws were crap as you say I am sure I would not be able to say this. I also have to say that my screws came in a VERY STOUT thick walled cardboard tube wrapped in bubble wrap and sealed at both ends and were in undamaged, unrusted, perfect condition upon arrival. That cardboard tube is probably tough enough for me to drive my truck over it and it would not collapse!! SO in short I am sorry you are having the trouble you are obviously having but to declare in large letters Roton screws are crap is a bit much. I also agree with Hoss in that Roton have treated me well with answers and follow up orders and quick shipping as well as personal calls to me to answer technical questions over the phone at length.

Good luck fixing the problems you are having and I urge you to voice your issues to the folks at Roton, I would be VERY surprised if they were not receptive to honest and civil criticism as well as possibly replacing the damaged parts.

[TXFred]

I edited my previous post to change the inflamatory statement.

I am still not impressed by the quality. I'll see if I can post up some pictures of the rust issues and runout in the next couple of days.

My Roton ballscrews also arrived in a very strong cardboard tube. I have no complaints about the packing.

I think that the ballscrews I bought were kept on the shelf for a long time. If you remember, I ordered two 24" lengths instead of a single 48" piece. I have a feeling that when I did this, I gave somebody in the warehouse the opportunity to get rid of some old cut off ends that may have been there for years.

This would account for the rust spots. And if they were stored horizontally but not supported fully, this might have allowed them to sag, especially if any additional weight was placed on top of them.

Since both of you are quite happy with your Roton screws, this is my best guess.

I won't be seeking an exchange from Roton on these, for several reasons. I've already annealed the ends, cut the screws up on a bandsaw, and machined the ends. If I had noticed the problems before I had done this, I would have sought an exchange. At this point, I don't believe it's right for me to return the ballscrews, since I have rendered them unusable to anyone else.

Look at it this way. If you got a bad meal at a restaurant, you'd take a bite and return it. You wouldn't eat the entire meal and then send back the empty plate.

These screws are also rated at a fairly large amount of backlash, which I don't like. This is my fault for not doing my research well. I hadn't realized that there would be this much variation in ballscrews. But the next time I spend the time to turn and cut ballscrews for my mill, I am going to use better quality ballscrews than the Rotons.

So it's not to my benefit to get new Roton screws at this time or later. At some point I'll trade up to better.

Fred

[pete from TN]

I am not exactly sure what you mean by Rated at a fairly large amount of backlash but I think most of the ballscrew offerings in a reasonable price range that do not utilize preloaded anti backlash nuts will exhibit this. Take a long look around here and feel free to have a look at my build thread "finally getting started" and see what can be done to make your ballscrews zero backlash. It is very important that you can get this done either with the purchase of a quality anti backlash nut from the respective ballscrew manufacturer or use a double nut setup as many on here have done. My machine, once I got everything tightened up exhibits extremely little backlash to none at all due to a stout bellville washer and adjustment setup shown in my thread. It was relatively easy to make and can result in a satisfying performance from even these lower priced screws. A setup without a near zero backlash system will be no fun to run and circles will be a real problem, basically any time the axis reverses direction in a cut you will have issues....

Since you already have then screws in hand, already have them machined and do not intend to seek a refund or replacement you have basically nothing to loose in trying this setup. If you do find you cannot live with these screws then I would say that those nook screws look to be well regarded. Frankly I cannot imagine that an X2 with any decent ballscrew and an anti backlash setup would not produce reasonably accurate parts within it's limitations. Hoss's X2 seems to work very well with them.... Good luck man... peace

[Crevice Reamer]

You are assuming Roton won't warrant the screws after you have machined them. I think you should call them and make sure. They might just be horrified that you have received such screws from them and WANT to replace them regardless of what you have done to them. They are obviously unusable, so unsellable anyway.

Give them a call and let us know what they say.

CR.

[hoss2006]

I noticed in your other thread that you machined the ballscrews 5 weeks ago and there was no mention of rust then.
Were they kept oiled to prevent rust in the meantime?
As far as sagging goes, you did put a torch to them so I could see that as
perhaps adding to the problem.
Lead error spec is not something that I would care about with any of the rolled screws.
If you run the Mach 3 Steps Calibration, any error is compensated for and the machine will run more accurately.
If your screw is off by .003/ft (that's .002 for the X travel and .001 for the Y)
Mach gives you the actual number of steps the screw has to turn for a given test length.
Could be just a few more or less steps than a perfect screws calculated steps but it moves the axis what it actually needs to be accurate.
Art is one smart cookie.
Hoss

[RotarySMP]

I used a Roton screw on the Z axis of my mini Lathe (I also have a little one for the X axis but haven't installed it yet). It does the job.

Lately I have been buying the parts to CNC my mill, and one of the screws I got off Ebay is an NSK ground screw rated at C5, with the T backlash rating <0.005mm or 2 tenths in imperial. I got this with an NSK preloaded AC bearing fixed bearing block and an NSK floating bearing block for $100 USD.

Given the comparison, I can't see myself ever buying a cheap rolled ball screw ever again.

[TXFred]

I haven't posted for a while due to lack of free time. But I finally got this video on youtube to demonstrate the issue with the ballscrews.

Also, I've attached a picture of the X ballscrew.

Hoss, to answer your question, I didn't think much of them at first because I left the ballscrews covered in grease. I saw the spots, but assumed that they were just thick globs of grease. I didn't find out that they were rust until I'd done the machining and cleaned up the screws in the parts washer. Even then, I can live with the rust spots, since they're not deep and will clean up with a mild scotchbrite pad.

I'm going to finish the build with these ballscrews for now. I think that they'll be good enough for initial testing. But at some point I'm going to look into replacing them.

[ame="http://www.youtube.com/watch?v=Xa0ZXEFdPgE"]YouTube- Ballscrew runout.[/ame]

Cheers,
Fred

P.S. Another post is on the way as fast as I can type, with lots of pictures!

[TXFred]

So the other reason for the delay in this build is that I was waiting to get some time on a bigger machine. On Friday I went to a friend's house and took his Bridgeport out for a test drive.

Having never used one, I have to say,"WOW!" That is a fun machine, and I want one.

I used his Bridgeport to recut the dovetails on my mill with a 60 degree cutter, and make a couple of other modifications.

On the column, I slotted the holes for the Z axis stepper, which will make that mount much more adjustable. At the same time I milled away the top of the ruler, allowing me to fit Hoss's Z axis side supports. With the slotted holes and the floating mounting block, I think that these supports will be very important to the rigidity of the Z mount. Lastly, I cut the column pivot square to the column. (It wasn't before, and I had to use a lot of shims to tram the mill previously.)

The table got new dovetails, and the ends cut square. I lost .010" of table length doing this, not a lot at all.

The saddle got cutouts for limit switches and the ballscrew, and new dovetails. I took the most time on this one, since the saddle dovetails determine the perpendicularity of the X and Y axes. To do this job I did all the operations on one side, then flipped the saddle over and reclamped it in the vise, keeping the same face against the fixed jaw of the vise. The dovetails were not at 90 degrees before I did this operation. Now I know they are (unless I screwed up!)

The base got the most work done. I cut out the center hole to allow for a longer Y leadscrew and limit switches. You can see that this was done pretty quickly and roughly, but it's good enough since the hole doesn't need to be precise.

The dovetails were recut. The column mount was cut down to make its flat surface extend fully to the rear of the base. I have an idea for a block that will fit into this area that will give extended Y dovetails, raise the column up higher, move it farther to the rear of the mill, and give it more support. For now I'll keep the column in the stock location, but the mill's base is ready for these modifications.

Cutting the base was the longest process. To begin with, I mounted the base upside down in a vise, with the jaws gripping the existing dovetails. Then I faced the bottom of the base and machined the lower inside of the hole. Then I removed the vise from the mill, mounted the base right-side up to the table, and indicated it off of one of the dovetail edges. With that done, I finished all the remaining cuts.

You can see that I forgot to go back and do the second cut on the front of the base where the Y axis bearing block will mount. I was 7.5 hours into the job at that point and forgot to do that one. No worries, I can do it tomorrow on the school's Benchman.

The next step is to lap in the new dovetails using Mothers Wheel Polish. That's going well so far. The X axis is pretty good, and the Y is coming along. I'll keep doing that all week and eventually I'll be satisfied with them. Then the machine can be put back together.

Anyway, enough babble. Here's pictures.