PDA

View Full Version : CONVERSION: Converting an AMAT25LV



Lloyd Barnes
04-01-2017, 06:18 PM
2021220213
As per the title, the plan is to convert an AMAT25LV. It seems a well trodden path so am hoping with the info available and some nudging back on track where required from you guys all should be well. I do have some questions at the bottom so would really appreciate your comments;
p
For info, I'll be machining aluminium perhaps 80% of the time, with occasional steel parts thrown in. Mostly car components for various petrolhead projects. I can always do some of the serious hogging out on the Bridgeport if it becomes a better option.
p
Things I've done/bought so far....
p
Electronics
• NEMA23 4nm motors for X & Y axis
• NEMA34 8nm motor for Z axis (I plan to add power drawbar etc and had read that a bigger motor would be a good idea for the Z axis
• Digital steppers, 5.6A for NEMA23 and 7.2A for NEMA34
• CP0-10V 4 axis breakout board. Only planning on using 3 Axis initially but wanted to leave option for 4th if it ever became desirable
• 48v power supply for drivers/motors
• 24v power supply for breakout board
• I've added an additional 4 relays and seperate 5v power supply to control box for future upgrades
p
At this stage I've pretty much finished up the machine control box. (you'll note from the pics that even if I cant afford a Tormach, I can at least have a fancy Tormach inspired logo? lol)
p
The current thing I'm working on is an enclosure. I want a really solid stand (I had the RF30 on a flimsy stand and it rattled like a good un). The plan is that whatever I make will also be bolted to wall/floor. I also want something enclosed to keep the chips down. I've pretty much made the base and am starting to fiddle with the tray and enclosure next.
The mill should arrive in a couple of weeks. I went with Amadeal as the price seemed fair and also they had an R8 version in stock. All of my Bridgeport tooling is R8 and whilst I'd want at some point to dedicate tooling to the CNC mill, seems to make sense to keep options open.
p
I do have some questions. Mostly thoughts I've got based around what I've read others doing;
1. I've read that bracing the Z column helps rigidity a lot. It sounds like common sense but has anyone tried it and able to comment or advise on what works best. I've put a box section upright on the rear of the base to give me something to brace to (or could go to wall if necessary)
2. Is it worth upgrading the spindle bearings to angular contact (see next question)
3. Is it worth increasing the spindle RPM? Playing with a feed/speed programme 4,000rpm seems a useful speed to aim for when I'm machining aluminium?
4. If so is it simply a case of swapping out pulleys or will the motor struggle? It's sold with a 1kw brushless DC motor.
5. I was planning initially to run mist coolant although I'll build the enclosure with flood in mind. I understand that chip evacuation is my primary goal with lubrication a second and cooling a third priority. I'll read through other posts but any major objections to mist?
6. I'm a little baffled by the pro's and cons of mechanical limit v proximity sensors? I have some limit switches but if its the right way to go, then happy to ditch them and switch to proximity, or am I getting this back to front and I need both??
7. Can anyone confirm how I should work out the amp setting for the drivers (am I being thick again?!)

Appreciate any thoughts :smile:

Clive S
04-01-2017, 06:51 PM
Welcome to the forum Lloyd. My comments would be up the volts on the drives if you can 68V is a good number with AM882 or EM806 drives and it definitely needs to be higher on the nema34.

I take it you are using ball screws.

I have done the conversion on the WM18 and used 23s with belt reduction all round and it runs fine.

Good luck with the build

Lloyd Barnes
04-01-2017, 07:10 PM
Thanks Clive

I'm using drivers from CNC4YOU.

5.6A, 50V CWD556 for the 23's
7.2A, 80VDC or 60VAC CWD872 for the 34.

I guess that means my best bet would be to keep the 48v supply for the 23's and add a 68V for the NEMA34 then?

Clive S
04-01-2017, 07:37 PM
Thanks Clive

I'm using drivers from CNC4YOU.

5.6A, 50V CWD556 for the 23's
7.2A, 80VDC or 60VAC CWD872 for the 34.

I guess that means my best bet would be to keep the 48v supply for the 23's and add a 68V for the NEMA34 then?

That's the problem that has been highlighted many times on the forum with buying kits.

Personally I would change the drives and built a toroidal transformer type PS.

Lloyd Barnes
04-01-2017, 11:05 PM
That's the problem that has been highlighted many times on the forum with buying kits.

Personally I would change the drives and built a toroidal transformer type PS.

No kits here Clive, I prefer to choose parts/screw things up myself :wink:

John S
05-01-2017, 12:44 AM
Go for 2:1 reduction on the head and you won't need gas struts etc.

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

Lloyd Barnes
06-01-2017, 04:19 PM
Thanks for the heads up on this. Have spoken with the motor supplier and done some further research.
P
Can I check my understanding. As I understand it (be gentle I'm, still learning this stuff!) running the driver for the NEMA34 at 48v is going to generate a much lower torque. Rather than the 8.7nm its capable of, it's probably going to be running somewhere closer to 4.5-5nm. Speed however shouldn't change? Is that right?
P
Assuming I'm getting this right, then my options are to belt drive the motor to the ballscrew 2:1 thereby doubling the torque if halving the speed, or up the voltage to the 34 to something closer to the 80v the driver is capable of handling. Rather than rip apart everything I've done, I could add another PSU running 68V and hook that into the NEMA34 driver. As a bonus it would free up some load on the 48V PSU if I ever wanted to add a 4th axis.
P
I added the 34 to cope with the added weight of a future power drawbar, but you guys are running the Z axis on 2:1 geared belt with NEMA23? Are you saying the you need closer to 8nm to happily run the Z even without additional weight on it
P
Just trying to understand the problem and my options going forward. Have to say there is a tonne of information out there, some of it conflicting. Reading through various sites I've seen people converting the G0704 with 4nm NEMA23 direct driven to the Z.

Lloyd Barnes
08-01-2017, 09:17 PM
Made some progress on the enclosure this weekend. Quite a lot yet to do but with luck I'm hoping to get it in primer and enamelled next weekend.
P
Have set it up for flood coolant for the future. There's also box section at the rear to bolt the Z column to. If nothing else should be able to hold everything pretty rigid.
P
Mill arrives next Saturday so hoping to make a start taking that apart soon. :smile:

20308
20309

Lloyd Barnes
09-01-2017, 05:25 PM
I'm using the Oldham style couplings from CNC4YOU on my build. On the NEMA23 they clamp around the shaft with no other locating method (the NEMA34 motor has a woodruff key in the shaft, so I'll cut a corresponding slot in the coupler).
p
What's everyone's experience using these? Do they slip at all or is there not a problem. There is a flat on the NEMA23 motor shafts, so I can drill and tap the couplers and add a grub screw locking onto the flat if its worthwhile?
p
Appreciate any advise :smile:

John S
09-01-2017, 05:32 PM
Yes Grubb screw as eell

John S
09-01-2017, 05:34 PM
Well, stupid phone

Lloyd Barnes
09-01-2017, 05:37 PM
Great, thanks John.

Lloyd Barnes
16-01-2017, 11:34 AM
Bit more progress over the weekend. Pretty much finished up my enclosure. A 3rd set of rails for the tops of the doors to be added, and some more polycarb for the windows now that I've measured it right! Some final wiring of lights etc and its about there. I've plastered it in sound deadening and its bolted to both the floor and the wall so very rigid. Will be nice to have a bit less cleaning up to do, the Bridgeport fills the whole shop with chips.
P
I've also put a fixing in at the rear direct through the enclosure to the wall to give me a strong point to anchor the column to to beef up its rigidity further. I'm planning to add some adjusting screws to help tramming the mill. How well do the columns line up on these little Chinese mills normally? The mounting faces look nicely ground but I've not done any measurements at this stage.
P
The mill arrived so starting stripping that down as well. First time I've bought a brand new machine then proceeded to start throwing half of it away!
P
203802038120382

Lloyd Barnes
10-02-2017, 07:24 PM
Bit of an update. Have started building machine back up.
P
Machined the base out to get a bit more Y travel, and generally clean it up. The ways look OK for a Chinese machine but the casting is the typical Chinese effort, bit wibbly wobbly! Had to do some slightly clearancing to the underside of the table where the casting was very uneven to make room for the ballnut mounting. Both axis are now running nicely under Mach3.
P
Column also needing machining to straighten up the slot for the ballnut mount which was a little cocked over. That's back in place. Next job is the Z axis motor mounts and ballnut mount and should be somewhere near ready to try some cuts.
P
Was going to upgrade the spindle to 5k rpm but I'll need to make some fresh motor mounts for that so might as well get the machine running and use it to make them as a test.

20780207812078220783

John S
11-02-2017, 12:05 AM
Aaaarrrrrggggg.

You have made a fantastic job of the enclose and the rest of the machine and then mount the steppers on 4 pillars ?????????????????????????????????????

It's a CNC machine tool not a fooking clock !!!!!!!!!!!!!!!!!!!!!!!

Ross77
11-02-2017, 12:15 AM
It's a CNC machine tool not a fooking clock !!!!!!!!!!!!!!!!!!!!!!!

Still one of your pet hates then John? :0)

Nice job on the mill and enclosure.

Lloyd Barnes
11-02-2017, 10:19 AM
I take it you don't like my spindly legs then?! Lol.
.
Have to agree they aren't the most manly looking of things, although in my defence as a newb I'm simply copying what ive seen elsewhere. They do seem solid.
.
I'll be hiding them with motor covers soon and we shall be rid of the sight of the abominations for ever! :shame:

JAZZCNC
11-02-2017, 12:43 PM
Thanks for the heads up on this. Have spoken with the motor supplier and done some further research.
P
Can I check my understanding. As I understand it (be gentle I'm, still learning this stuff!) running the driver for the NEMA34 at 48v is going to generate a much lower torque. Rather than the 8.7nm its capable of, it's probably going to be running somewhere closer to 4.5-5nm. Speed however shouldn't change? Is that right?

Ok seen as nobody answered this here goes.! . . . . However first let me point out that suppliers will always want to sell you what they have in stock or highest profit margin they'll never recommend what's best for you if they don't have it on stock.! . . . So if today they have 48V psu's then "This what you need Sir".!!!

Torque and speed are both proportional to Voltage so both will be affected to some degree. However low voltage affects speed more than torque esp if Motor inductance is high, which is often the case with larger motors hence why 23's are often used.

The higher the Inductance of the motor the more voltage you'll require to get the same speed/torque than motor with lower inductance. Then we have how the motor is connected to factor into the mix.! . . Bi-polar Parallel or Series wired.?
Most large motors are 4 wire motors internally wired in series which means they'll require much more voltage than 8 wire motor which can be parallel wired.

Series wired motor will provide higher torque down low in the speed curve but torque will drop off quickly as the rpm rise. To get higher speeds out of series wired motor will require much much more voltage.
Parallel wired motors will have lower torque down low but carry it much higher up speed curve and in general give more balanced torque/speed curve using less voltage.

So in your case if you'd used nema 23 wired parallel with 48V and 2:1 you'd be better off than using series wired 8Nm 34 running on too low voltage.
Personaly I'd just fit the 34 with 2:1 with what you have got and see how it goes. IF doesn't work out just give it more volts.

Lloyd Barnes
11-02-2017, 08:22 PM
Thanks for the explanation Jazz, learning all the time. First off I have to absolve CNC4YOU of any blame. I watched a lot of videos of builds, then dived in and bought stuff myself. Nothing recommended by them. Fools rush in etc!
.
All my motors, NEMA34 included, are 8 wire, wired in parallel. At least I got that bit right then! Lol.
.
As you suggest I'm going to fit it and see what happens. If it doesn't work then more voltage will be fired at it.

I'm hoping to have it running in a couple of weeks so will be interesting to see what it can do. The columns rigid at least as that's bolted at the top to the enclosure frame.
.
Thanks for the advice. Appreciated. :thumsup:

Lloyd Barnes
14-02-2017, 04:53 PM
Quick questions folks
p
Just modelling up the Z motor mount plate to convert over to a 2:1 belt drive. Are you guys just running a pulley on the unsupported shaft of the stepper motor or are you adding some kind of additional bearing support? I'm conscious the belt is going to need to be pretty tight to avoid backlash, but worried about overcooking it and putting a tonne of side load on the motor shaft.
p
What have you chaps done?

Davek0974
14-02-2017, 04:57 PM
No, do not over tighten a toothed belt, it does not reduce BL and will overload the motor bearings. A properly tightened belt should have a small amount of give in it - you should not tighten it so you can play music on it :)

Lloyd Barnes
15-02-2017, 09:47 AM
No, do not over tighten a toothed belt, it does not reduce BL and will overload the motor bearings. A properly tightened belt should have a small amount of give in it - you should not tighten it so you can play music on it :)

Hi Dave,

Cheers for that. Are you running a belt drive? Do you find you need to run backlash compensation on Mach3 with it?

Lloyd

Davek0974
15-02-2017, 11:35 AM
I run toothed belt on both my mills, the bigger one has a heavy table and servo drives, if the belt is tightened just right there is no backlash to compensate for at all. The belt can move slightly but when that movement is converted to rotation and then reduced by the screw pitch, the actual distance linear is microscopic.

I have tried backlash comp and i would not recommend it to anyone - always get the system mechanically sound and unless you are working to nuclear industry standards, just learn to live with the tiny backlash we have - there will ALWAYS be backlash or the system would not move, its just a matter of knowing how much.

Lloyd Barnes
15-02-2017, 11:54 AM
I run toothed belt on both my mills, the bigger one has a heavy table and servo drives, if the belt is tightened just right there is no backlash to compensate for at all. The belt can move slightly but when that movement is converted to rotation and then reduced by the screw pitch, the actual distance linear is microscopic.

I have tried backlash comp and i would not recommend it to anyone - always get the system mechanically sound and unless you are working to nuclear industry standards, just learn to live with the tiny backlash we have - there will ALWAYS be backlash or the system would not move, its just a matter of knowing how much.

yep, couldn't agree more, do it right, do it once etc. I just wasn't sure whether switching from a direct drive to a belt drive inevitably introduced a hint more backlash regardless of how perfectly you tune the belt tension. Always faster/cheaper to learn from others mistakes etc. Should have the belt drive put together over the next few days so will be able to test backlash on all the axis.

As you say we're not working to nuclear standards. I'm generally machining car parts on the mills, and the tolerances don't need to be that high. Its just I'm tired of using the Bridgeport like an oversized etch a sketch to try to make curved parts! :smile:

Davek0974
15-02-2017, 12:21 PM
yep, couldn't agree more, do it right, do it once etc. I just wasn't sure whether switching from a direct drive to a belt drive inevitably introduced a hint more backlash regardless of how perfectly you tune the belt tension. Always faster/cheaper to learn from others mistakes etc. Should have the belt drive put together over the next few days so will be able to test backlash on all the axis.

As you say we're not working to nuclear standards. I'm generally machining car parts on the mills, and the tolerances don't need to be that high. Its just I'm tired of using the Bridgeport like an oversized etch a sketch to try to make curved parts! :smile:

I thought you were converting an AMAT25LV ?

My Bridgeport CNC conversion was the best thing i ever did, it makes it a much more useful machine. The backlash i do have comes mainly from the screw thrust bearings - in the end i bought new units here and it helped massively, not cheap though. The next is screw nut slop - I just fitted a standard Bridgeport conversion screw set from the USA, its good but not excellent. I have less than 0.1mm backlash in both axes and i don't think thats too bad for an old girl :)

Lloyd Barnes
15-02-2017, 12:38 PM
I thought you were converting an AMAT25LV ?

My Bridgeport CNC conversion was the best thing i ever did, it makes it a much more useful machine. The backlash i do have comes mainly from the screw thrust bearings - in the end i bought new units here and it helped massively, not cheap though. The next is screw nut slop - I just fitted a standard Bridgeport conversion screw set from the USA, its good but not excellent. I have less than 0.1mm backlash in both axes and i don't think thats too bad for an old girl :)

Sorry not being clear am I. I have a series 1 Bridgeport but have bought the AMAT to convert to CNC. Sadly my Bridgeport, even though its a 1981 model is a bit worn. I can't adjust the gibs to acceptable play for a CNC machine without the ways tightening at the outer ends. Also I really wanted something I could enclose, the Bridgeport fills the shop with chips, and I'd didn't want to add to it by getting the coolant working too. I'll certainly keep it for general use where the part is not complex or for sizing stock for the AMAT etc.
.
I read a load online and the general impression I got was converting a series 1 wasn't the best way to go, sounds like you disagree although perhaps you have a later model? I could certainly live with 0.1mm with the parts I make, not bad at all for a Series 1. I guess the bits were pretty pricey though?

Davek0974
15-02-2017, 01:26 PM
Ok, got it, makes sense now.

I think you did the right thing, my BP can fling the chips a good 8' away but its only me so who cares :)

I was strongly advised NOT to convert the BP and that advice was right. The advise was to sell it and get an old CNC VMC etc and retrofit the electrics - this would have been ideal BUT i could not physically fit anything but a BP in my shop - not even a proper CNC BP. I also do not have the power capacity for the big spindle motors on CNC stuff.

Now i have used it a few times, I REALLY want a proper CNC mill with an auto tool changer. ;)

Total cost was silly money but once you dive in you cant back out.

Lloyd Barnes
07-03-2017, 05:45 PM
Well, its about ready to go. I've taken the advice and setup the Z Axis with a 2:1 belt drive (very annoyingly I ordered the wrong belt length so just waiting now for the right one to come). Most of the rest of it is complete so once the belt is here its time to take a few trial cuts and see where it is. Y Axis NEMA23 with spindly legs is now mercifully covered up, cover plate for the X Axis next. Hopefully starting to cut some trial parts this weekend.
/p
Appreciate all the advice so far :thumsup:

209912099220993

Davek0974
07-03-2017, 06:41 PM
Looking good :)

Lloyd Barnes
13-03-2017, 08:51 PM
Well chaps, she runs like a good un! Many thanks for the advice. The Z NEMA34 motor didn't work with 48V and direct drive as you suspected, was losing steps. Switched it out to a 2:1 belt drive and all sorted.
P
Got Mach3 licenced today and ran a quick test part. Few screw ups here and there, didn't get the CAM quite right (Fusion 360) so the pocket bottoms didn't clean up, and like an idiot I managed to run the spindle in reverse for the first part of the chamfer op so that's a bit messy in parts, but that's just me being an idiot and easily fixed next time.
P
The machine itself I'm generally pleased with. It ran the part in around 40 minutes, and with fresh end mills that are a little sharper I'm pretty sure I can be a bit more aggressive if I needed to be. Dimensionally it seems to be cutting to around 0.1 to 0.075mm accuracy which I'm hoping I can improve by tweaking the jam nuts on the ballscrews to take out any backlash etc. What sort of accuracy do you think its practical to get on a small machine like this? 0.1mm would be fine for many of my parts I'll be making, some however I'd need a little better, for example where I'm after a sliding or press fit between two parts. I didn't put a lot of time into this part, it was more to get myself going on the basics so I'm sure with time I can improve.
P
Doubtless I'll have more dumb questions as I get the mill dialled in. For now though I'm planning on getting my tool library setup and sorted out in Fusion and get some parts through to get more comfortable with the CAM in particular. I'm still at the stage with Fusion CAM where I'm not 100% sure what I'm doing, particularly selecting the geometry to get it to do what I'm after but I'm sure it will get easier. Once again, thanks for the advice and encouragement. Much appreciated. :thumsup:
P
2108221083

Lloyd Barnes
13-03-2017, 09:00 PM
B8gger, I knew I had another question. I want to get myself a decent flycutter. I'm not good at grinding tooling, would prefer one with an indexable tip. I like the results the Tormach Superfly Cutter seems to get but are there any UK based similar tools? Or has anyone put an indexable lathe tool in a normal flycutter body and got good results?

Lloyd Barnes
18-03-2017, 12:55 PM
Slowly getting there.
p
Thought I'd make a tooling holder as a practice project. Haven't quite got the hang of the settings in Fusion 360 for chamfering yet, plus when I had Mach3 in feed hold and was restarting it did a preparatory move and milled a channel right across the piece (under my thumb in first pic) :upset:.
p
Fortunately I'm making 2 of them so this one can be the one at the bottom if I can get the second one bang on! :thumsup:
p
211462114721148

routercnc
18-03-2017, 03:03 PM
Hi Lloyd,

There are lots of examples of indexable lathe tools being used in home made flycutters. Here is one:
https://www.youtube.com/watch?v=QgvbVLfeig8

This Old Tony made one using high speed steel tools:
https://www.youtube.com/watch?v=JBDNu_ViHlU

I haven't used any of these as my WC spindle is too fast and the interrupted cut is a bit hard on the bearings. However, Arceurotrade have started selling these micro flycutters which look tempting and are low cost:
http://www.arceurotrade.co.uk/Catalogue/Cutting-Tools/Fly-Cutters/Micro-Fly-Cutter-Sets-Straight-Shank

As for the mess ups, this is how it goes when you start off. There is a list of them and you just have to work your way down them ticking them off and not doing them again. You are getting stuck in and the parts are looking good !

Edit: The preparatory move after a feed hold - was that using the 'run from here' button? I don't know if I'm doing this right but I've found that this moves the tool diagonally (across AND down) from where you are to where the cut wants to start. This can go through the work piece, clamps etc. So I always go to the prep move page which shows where it wants to move to in X and Y, then cancel out of that window and MDI the machine to that X Y position, then do run from here again with the spindle on. It then just goes straight down onto the cut and waits for you to hit cycle start to carry on the cut.

Davek0974
18-03-2017, 04:00 PM
Edit: The preparatory move after a feed hold - was that using the 'run from here' button? I don't know if I'm doing this right but I've found that this moves the tool diagonally (across AND down) from where you are to where the cut wants to start. This can go through the work piece, clamps etc. So I always go to the prep move page which shows where it wants to move to in X and Y, then cancel out of that window and MDI the machine to that X Y position, then do run from here again with the spindle on. It then just goes straight down onto the cut and waits for you to hit cycle start to carry on the cut.

Thats what i do too, got caught once, cutter bit the bullet.

Have to be really on the ball when using run-from-here.

Lloyd Barnes
18-03-2017, 04:30 PM
Yep, it was the "run from here button". Is the prep move page the little box that pops up when you hit cycle start for the first time afterwards? I'll try that next time, really frustrating to spoil a part when you think you've got everything right, but I guess like you say, we learn from our mistakes!
p
Cheers for the links, will have a nosey at those. I have a bog standard fly cutter for the Bridgeport but never really got it running sweetly. I've never taken the time to learn how to grind tooling properly. I guess I should but if I can just chuck an insert in it and get cutting then I'm always tempted by that!

Lloyd Barnes
02-05-2017, 08:21 PM
Well, bit the bullet in the end and grabbed the Superfly cutter from Tormach. Nice enough looking bit of kit and seems to cut very nicely in the little CNC mill although the unbalanced nature of a flycutter does seem to set up a bit of a vibration.

On another note, whizzed up a pair of brackets at the weekend. Am using Fusion 360. When I'm using the 2D pocket function it never seems to clean up the bottom of the pocket properly. Made 2 brackets at the same time from a single piece of Acetal and it did the exact same cut on both. Using the simulate tool in CAM it doesnt show this. I think the pic shows this ok. Am I doing something wrong?!

21484

komatias
02-05-2017, 10:24 PM
In the pocketing options, have a look at the setting on multiple depths. There is something that says stock to leave on walls and bottom. It usually defaults to 0.1mm or so.

For pockets in soft materials, have a play at adaptive clearing. It is quicker and really good fun :D

Lloyd Barnes
02-05-2017, 10:40 PM
Cheers fella, will check that out. I'm doing something wrong for sure! We never stop learning I guess.

The chamfers are also not coming out right, supposed to be just a 0.5mm edge break. I think its because I add a chamfer in the model, and then use chamfer option in the CAM but you end up selecting one of the edges of the 45 degree feature, rather than the now non existent 90 degree corner. That probably makes no sense at all!

komatias
03-05-2017, 12:22 AM
it does make sense. I have faced the same issue. To make it work, I usually make a sketch in the model mode and use that as the profile to mill from. Otherwise, select the inner most chamfer edge and set the heights accordingly