PDA

View Full Version : BUILD LOG: Sieg SX2.7 CNC conversion



Edward
03-01-2017, 11:20 PM
Hi guys, my first post after following this forum for months. I hope you've all recovered from the festivities:)

I am converting a Sieg SX2.7. So far I have converted the X and Y axes with bog standard chinese ballscrews, and Nema23 2.2Nm direct drive (i.e. no pulleys) , driven by Leadshine EM806, 68V psu.etc. So far so good.

For the controller I am using Kflop with KMotionCNC. I've tried a couple of the little cheapo chinese stand-alone boxes that were discussed in another thread. Although they moved the machine OK, the movement (particularly around arcs) just sounded a bit clunky compared to the silky smoothness of the KMotion. For many other reasons (particularly stability) I settled for KMotion, also tried the KFlop with the Mach3 plug in, which worked just as good, but to be honest, I kind of prefer the simplicity of the KMotionCNC.

Anyway, now I am at the stage of doing the Z axis. I've taken the old leadscrew out, luckily the existing ballnut support can be reused, no need to machine a new one as the standard ballnut with the six bolts fits perfectly in the hole (though only using 4 of the bolts).

BTW, the X axis was more difficult to fit, as typically there is very little room under the table for the ballnut, a little grinding of the saddle was necessary, which I did partly with a manual mill and then the Dremel to get a nice final finish, so it wasn't too dramatic. The un-ground rough part of the saddle casting wasn't level, so some material had to be removed for the length of the ballscrew to have clearance.

I am thinking of using a slightly more powerful motor for the Z, Nema23 3.1Nm. and my question is....do you think I can get away with coupling it direct (I use Oldham couplers) , with no pulley 2:1 reduction? My initial thoughts are to try it direct to start with, and if I notice that the motor struggles or looses steps when cutting, then it doesn't take too much effort to machine a new plate to house the pulley reduction and put the motor to one side. Direct coupling looks a bit more elegant and it's easier to fit, but it's no good if it doesn't work.

Since what I think I've learned from this forum is not to use motors bigger than necessary, I wonder if, from your experience, you think that direct coupling for the Z will work in this case.

The only other observation is that I have both the EM806 and the AM882 drives and with identical settings, the EM806 whistles a little on standstill whereas the AM882 is completely silent. Both are very quiet and very good when moving, with a very subdued mellow Symphony that is delightful to the ear:)

Edward

Clive S
03-01-2017, 11:59 PM
Edward Welcome to the forum its always nice to hear that some do actually read and do their research as it seems you have:cheerful:

You have stated ( with no pulley 2:1 reduction) is this not a contradiction? I personally prefer using belts with a reduction as it is better for resonance and you get twice the torque.

You have done well in your choices of drives they are both very good and the correct Voltage PSU

Edward
04-01-2017, 12:10 AM
Hi Clive, thank you for your welcome.

Let me rephrase this. I meant without using pulley reduction, i.e. no pulleys, just direct drive.

I have no problems with using pulleys either, I use them a lot for my motion control projects. And for the Z, being more distant from the swarf, it makes even more sense, plus the extra torque from the reduction, etc. Still, since I am cutting the fixing plate tomorrow, I am in two minds...

Edward

m_c
04-01-2017, 12:30 AM
Nice to see another KFlop user on here!

I'd personally go for a belt reduction, but if trying direct mount is relatively easy, it's worth trying. Whether it will work, will most likely depend on what you try doing with it.
The worst thing I would expect to highlight any lack of torque, would be during drilling, especially when you try retracting the drill, when you're fighting against the weight of the head and the drill still trying to dig in.

Regarding the EM drive being noisy, have you ran the autotune function?

Edward
04-01-2017, 12:43 AM
Hi m_c, the Kflop is great isn't it:)
I may ask you about limit switches with the Kflop, as no doubt you have done it. I would like to use IO0 to IO4 on the JP7 connector, but I've never done this. I will reserve the question for later on, one thing at a time:)

Regarding the EM drive, it's not particularly noisy, just the usual little buzz that some drives emit on standstill, which can get annoying over time. Yes, I tried the dip switch4 down and up. I did the same with the AM882 which I have on the other axis, identical setting and motor, but this one doesn't make any noise. I seem to remember I have the little yellow selector wheel with the arrow on 4 which selects a particular Leadshine motor and applies relevant parameters. I may try and set the wheel to something else and see what happens.

Edward

Edward
04-01-2017, 09:35 PM
Hi,


Today I fitted the Z axis ballscrew. I had to drill the top of the column, a hole of about 18mm. it wasn’t fun as my drill bits only go up to 12mm, and it takes forever to enlarge the hole with a Dremel. Then I remembered I had these cone shaped drills that drill in ever larger steps. Thereafter it was a piece of cake.


I haven’t machined the top plate yet, but by connecting the FK12 at the top plus a handy little lever I made to crank the ballscrew in both directions, I could get a rough idea of the effort needed by the motor.

Basically, to move the head down requires zero effort. In fact, if you are not careful it may start to roll down by itself. A question of adjusting the gib a little, I think, just to hold the head still.

Cranking it up requires more effort, for instance, more than moving the X or Y axes, but not drastically more. I think the 3.1 Nm motor will move it well with direct drive. We shall soon see….otherwise maybe a 4Nm motor...

In the future, I may attach a gas spring, if anything just to ease the effort on the motor and to hold the head stationary when the motors are disconnected.

I think for the moment, I may have to use the gib tightening lever when I am not using the mill, just in case the head decides to roll down and smash something pretty badly. I hope I remember to loosen the lever when in use!

Tomorrow, machining the thick plates to fit the bearings and motor, with the manual mill, painful, or as the French say...quelle horreur!

Edward

John S
05-01-2017, 12:46 AM
Use a 2:1 reduction and you won't need a gas strut

http://www.stevenson-engineers.co.uk/files/Z%20axis%20complete.jpg

Edward
05-01-2017, 12:18 PM
Well, in view of your opinions, I am going to cut to the chase and go for 2:1 pulley reduction for the Z axis, as I already have 40th and 20th AT5 pulleys in my stock anyway. Just ordered the right length of belt. Also, by having belt reduction, I will be able to tighten the gib a little more, making the head firmer and at the same time avoiding the rolling down by itself. On the upwards move, the double torque gained by the reduction will help too.

I will be using the Lenze type locks for the pulleys, as I prefer them to ordinary set screws. I'll have to put the motor to one side rather than to the back of the column like John shows, only because the column cover gets in the way, and also because of the depth of my table. Eventually I will also print a neat cover to protect the belt/pulleys.

Edward

Edward
10-01-2017, 12:09 AM
Just finished the Z axis as recommended, with a 2:1 reduction. It works well and not a hint of any struggle with a 3.1Nm Nema23 So I am glad I went for the pulleys.

I minimised the AT5 belt run so the motor was as close as possible to the column, to one side.

Now I am fitting a swarf guard at the front of the Y, as this particular model doesn't have any bellows there to protect the ways from dirt. So I am just fitting a U-shaped metal cover plate, 190x190mm attached to the saddle so it moves in and out with it.

I will now move to other sections of the forums to hopefully get advice or contribute. Thanks.

Edward

Clive S
10-01-2017, 10:08 AM
Edward How about keeping the build log here so that all questions related to it will be in one place as that helps others doing a similar conversion?

njhussey
10-01-2017, 10:09 AM
And pictures, we like pictures 😁

Sent from my HUAWEI VNS-L31 using Tapatalk

Edward
10-01-2017, 10:13 AM
Absolutely, I like pictures too! So as soon as I finish the guard I will post some pics of the project for other potential "converters" to see.

DamnYank
07-02-2017, 08:23 PM
Greetings from the far side of the pond. I encountered this thread via Google search, while looking for any kind of discussion related to the SX2.7. It seems to have not quite caught on as strongly as either the smaller mills, nor the larger X3. Thank you for posting your experiences with it here. I'll be following with interest.

What kind of modifications did you have to make to your ballscrews and/or the mill itself to get it adapted? Or, did you miraculously find some stock part that fit like a glass slipper?

Thanks again for sharing info!

Edward
08-02-2017, 12:50 AM
Hi there!

Well the SX2.7 fits nicely in between the two. Far better than the SX2 in every sense, but not as good, in some respects, as the SX3. I could've easily bought the SX3, but I wanted something slightly smaller and with a better Y cross travel (180mm) The SX3 has less, at 145mm.

As far as modifications, for the Z axis, the standard ballnut will fit exactly using the original housing, so no modifications needed. For the Y axis I made my own ballnut housing, from a block of aluminium, easy to do. Otherwise all is fine.

The X axis is the one where you have to work a little. I had to mill one side of the ballnut flange, take a chunk off of about 6mm and then I had to grind the saddle at the very end to make room for part of the ballnut. I also had to grind a small groove across the saddle, not deep at all, just enough to gain clearance, as the ballscrew needs to be fitted slightly lower than the original leadscrew (because of the thickness of the ballnut). I didn't have to touch the table at all.

The rest is the usual end plates for which I used 20mm aluminium.

The only misgivings I had about it is the fact that you can't rotate the head, but as it turns out, after lots of measuring, (I did it following Hossmachine video on YouTube) it is running parallel anyway, thankfully.

It is a lovely machine, I also have the SX2, which was my first machine, in my opinion, for a CNC conversion, no comparison, I mean the SX2.7 without a doubt:)



It's not unusual to modify things a little. I did it all with a Dremel and a bit of patience, not too difficult really.

Edward


https://www.youtube.com/watch?v=ObzOyfOL8vg


https://youtu.be/kXQHS8wq_As

DamnYank
08-02-2017, 01:14 AM
Thanks for the additional info. I'm looking at the Hossmachine videos now. Did you have to resize the ballscrews? I have heard this is necessary in some cases, and is a trick since they are hardened. If you have reference parts numbers for the ones that worked for you, that would be a help. Thanks again!

DamnYank
08-02-2017, 01:29 AM
Hmmm. Apparently my prior reply was sent to oblivion.

It went something like this:

Thanks for passing on the additional info. I'm reading the Hossmachine guide now. Can you share any part numbers on the screws that worked for you? Did you just match the length and diameter of the existing, and they fit without complication? I've read stories of conversions that required specific modifications to the screws itself, and it appears to be pain working with the hardened material, as well as a risk of warping the screw in the process. Any specific links you can share to point me to parts that worked for you would be welcome.

Below is a link to YouTube video by a gent in Australia. He made some slick kit for his x2.7 using his CNC router to cut fancy parts. Thanks again for your help.




https://www.youtube.com/watch?v=Kb_URlTUj8A

Edward
08-02-2017, 01:51 AM
The ballnut type is the same as the Australian.

Ballscrew, 16mm, bog standard chinese stuff.

300mm long for the Y
600mm long for the X
500mm long for the Z

But note that each length is a little longer than the above, for the bearings, thread, nut, etc.

Best to get the mill and then measure up. Bear in mind that if you get it in the US, the mill table and other specs may differ a little. I only had to cut the very end of the floating Y ballscrew, about 9mm off the end. I used an angle grinder and it cut in about a minute. You can also use the Dremel with a little cutting disk, but it will take longer and the cut won't be as clean, so you just use a file to flatten the finish. The ballscrew won't bend unless you force it. I just clamped it gently with some protection and only the very end of the ballscrew was exposed. I bought a cheap angle grinder for this, scary to use the first time, but actually it was a lot easier than I thought. I looked like an astronaut with all the protection, but better safe than sorry:)


Edward

DamnYank
08-02-2017, 02:02 AM
Again, that's very helpful & appreciated. The comments I read by people modifying their screws involved annealing the end they needed to modify -- not just cutting it short, but turning a precise shoulder on it after getting it to length. That heating process seemed to be where the warping risk was, and if this conversion required anything that sporting, I'd pass on the experience ;)

And I'm with you -- better safe than sorry, even if the big boys make fun of me.

Edward
13-03-2017, 04:56 PM
Quick release drawbar for Tormach style tools.

The drawbar is accessible and can easily be pulled out by swinging the levers, as shown in the video.

Edward



https://www.youtube.com/watch?v=w0d-ogmOIeU

komatias
13-03-2017, 06:24 PM
Hi Edward,

that is an amazingly clean conversion!

Edward
13-03-2017, 07:03 PM
Thank you! So far everything is working well.

The only problem I had with this quick tool change was that on one occasion the tool holder started to slip out as I cut with an 8mm endmill. The Bellevilles had been tighten by two full turns of the drawbar (after finger tightening) and it seems that the tool wasn't getting enough grip by the collet. So I increased the Belleville tightening to just over 3 turns of the drawbar, and now it seems to be gripping properly. These things always need a little fine tuning...

Edward