[Starleper1]

What program are you modeling with?

[djbird3]

What program are you modeling with?

I used AutoCAD 2011.
~Don~

[LatheMaster]

Hey DJ , Got all my stuff now . Just need a box , e-stop and home switches . Glad I didnt get a box yet . Didnt realize how tiny everything is .

Picked up a P4 Hyperthread 3gig computer with 17" monitor and keyboard n mouse for $170 with a warranty . Glad im almost done buying stuff . Probly get the LMS 4" vise , clamp kit and collets next .

Ill start my own thread and shoot you a link so we can compare notes .

[djbird3]

Hey DJ , Got all my stuff now . Just need a box , e-stop and home switches . Glad I didnt get a box yet . Didnt realize how tiny everything is .

Picked up a P4 Hyperthread 3gig computer with 17" monitor and keyboard n mouse for $170 with a warranty . Glad im almost done buying stuff . Probly get the LMS 4" vise , clamp kit and collets next .

Ill start my own thread and shoot you a link so we can compare notes .

Hey, today I ordered one of those Keling KL-5056D digital drivers to test out. I'll probably have it around tuesday. Hopefully I'll have all of the parts made by then and will begin mounting everything on the mill. I worked allday today and got a few more parts finished but had to redo a few things because of a few errors in my original model but I'm getting there little by little. Tomorrow I'll probably finish cutting the ballscrews, real pain to cut on this little 9x19 Grizzly I have, think it's time for a "big boy" lathe like that G4003G. Anyway I'll be posting more pics as I get stuff done and explaining what I'm doing. BTW, have you started converting your mill yet or are you still rounding up parts? If I were you I wouldn't worry about mounting any electronics in a box until you get everything operating, by then you'll probably have changed a few things. Anyway, have fun.

~Don~

[LatheMaster]

Still rounding stuff up. I have a nice 1 and a half inch thick piece of plywood that I am gonna make a bench out of. I will get it built monday. Still debating on the oiler. Thinking of trying a gravity feed setup . Will probly implement that later on. The little machine shop manual says to use lithium grease on the lead screw . Is this acceptable on the ball screws or do I use oil when its converted to ball screws?

I have a small craftsman toolbox I was gonna use for the electronics . That should be nostalgic .

Let us know how the digital driver works . I would have liked to be gone that route but I need to be more educated first . Im thinking it will make the steppers run quieter cause of the higher hz. It runs at. Is this correct?

[djbird3]

Still rounding stuff up. I have a nice 1 and a half inch thick piece of plywood that I am gonna make a bench out of. I will get it built monday. Still debating on the oiler. Thinking of trying a gravity feed setup . Will probly implement that later on. The little machine shop manual says to use lithium grease on the lead screw . Is this acceptable on the ball screws or do I use oil when its converted to ball screws?

I have a small craftsman toolbox I was gonna use for the electronics . That should be nostalgic .

Let us know how the digital driver works . I would have liked to be gone that route but I need to be more educated first . Im thinking it will make the steppers run quieter cause of the higher hz. It runs at. Is this correct?

I was thinking about doing an oiler also, something like the one on the Syil CNC mills. Their lube system has a manual pump though, I'd rather something automatic, gravity feed isn't a bad idea. I noticed these REX ballnuts I got have oil ports on them, all I need to do is make some small fittings so I can attach hoses to them to feed them oil. Also need to drill oil ports in the machine to get oil to the dove tails, this is what needs the most lubrication. I'll get around to working on the oiler eventually, not a big priority at the moment, still making parts for the conversion.

Your question about what makes the digital controller smoother, quieter, etc. has to do with the way it processes the signal that drives the coils in the stepper not frequency. With any controller analog or digital smoother operation can be achieved by using microstepping, because the more steps per revolution the smoother the motion. The downside to microstepping is that it is not as accurate as single stepping and can cause the stepper to get out of sync under certain conditions. A digital controller has many advantages because it utilizes a microprocessor to process the drive signal in a much more sophisticated way than can be feasibly done with analog circuitry. To understand what I'm talking about in regard to a microprcessors advantage just look at how powerfull your average handheld device is these days, these are only possible because of small high speed processors. So essentially these type of controllers are software driven which means they can be improved by simply changing their software, another advantage.

I should be getting that controller this week so I'll be testing it on some different motors I have to see how it performs. Hopefully I'll be testing it on my mill, I just have one more ballscrew to do and I'll be done, I hope.

~Don~

[sansbury]

The little machine shop manual says to use lithium grease on the lead screw . Is this acceptable on the ball screws or do I use oil when its converted to ball screws?

I use oil on mine. Grease tends to retain more debris which isn't good for the screws or nuts.

[djbird3]

Here's a few more pics of my current progress. So, far I've completed the y axis components, as well as the Z axis parts and all the X axis parts with the exception of the ballscrew. Unfortunately I discovered a serious wobble in my Z axis screw. Apparently the 5/8 collet I was using to hold the shaft during machining had poor parallelism and I didn't notice because the end of the shaft was hidden inside the spindle. I assumed everything was ok because the shaft was perfectly concentric, oh well, thats what I get for assuming. Fortunately I made the shaft longer than it needed to be just in case of a problem, so I'll just cut it back and remachine it. I tried to tweak it with my lathe chuck but this steel is way to hard and it's not worth screwing up my chuck just to avoid about an hour worth of work. Anyway thats it for now.

~Don~

[kawazuki]

I think that (refering to the last photo above) if you do not mill out the ballscrew slot in the base at the front behind the stepper mount you will not gain the extra Y axis movement you have machined into the stepper mount! Or maybe you have already - just looks about stock width? in the piccy.

[djbird3]

I think that (refering to the last photo above) if you do not mill out the ballscrew slot in the base at the front behind the stepper mount you will not gain the extra Y axis movement you have machined into the stepper mount! Or maybe you have already - just looks about stock width? in the piccy.

Yeah, I have to mill out about 3/8 or so to get the travel I intended, might mill some out the other end also.

~Don~

[djbird3]

I got my KL-5056D driver and KL23H286-20-8B 425 oz/in motor the other day and had a chance to test them today. Of course, the first thing I had to do was remove the cover to see what processor they used in this unit, for those interested it's a Texas Instruments TMS320F2806PZA 32 bit signal processor. Also you'll notice in the pic it has a rs232 port, Hmm, maybe for upgrading firmware?, since I have no documentation I can only guess.

Anyway, I started by testing the combo on the bench with a variable Labratory power supply and a function generator pictured below. I connected the motor with the coils in parallel, set the driver current to 2.7A RMS, 1600 steps/rev, and I enabled half power mode(this feature reduces the current draw while the motor is idle). When powered up at 40v set on the supply it draws about .2A when idle, and the current draw varied between .3A and 1.6A depending on the RPM. The unit seems to have the least current draw around the middle of the motors RPM range. I was able to vary the step frequency between a few hundred Hz and 50KHz as quickly as I could turn the knob on the generator and the motor never once lost sync. I also tested to see how fast I could get this motor(425oz/in) to run with this driver and was able to reach a step frequency of 144KHz or 5400 RPM's. Thats not bad considering the inductance this motor has. This driver didn't even get warm during the 20 minutes I ran the motor nor did the motor get hot like they usually do. Also this driver does run much quieter and smoother the anything I've had experience with.

Next I installed it on my x2 recently completed Z axis to see how it would perform. This is the purpose I got this motor driver combo for. I used the same settings as with the bench test and was able to get rapids in excess of 335 in/min and this is with the gibs tight. The only reason I could only get 335 was because I had the kernal set to 44KHz. I'm sure it will do in excess of 350 if I change the kernal speed to 60KHZ. Since my goal was only 250 I think the motor/driver combo is a winner and definitely would recommend the new Keling digital drivers to anyone in the market for a stepper driver. I'm probably going to get a couple more of these for the x and y axis's since they work so well.

~Don~

[LatheMaster]

Looks good Don . Those are some awesome results . Makes me bummed that I already have the G540 . Ohh well , I need simplicity for my first build .

Heres a link to my thread . Check in once in a while . I appreciate input from all directions .

http://www.cnczone.com/forums/bencht...cnc_build.html

I had a question on wiring . I heard of a need for shielded wire on the limit and home switches so they dont catch interference from the steppers . What is the preference for stepper and limit/home switch wire ? 22ga. good ? Shielded/Unshielded ?

Also I see you did a design on the Y to allow the table to come towards you more . Wish the CNCFusion kit did that . I am going to make a new stepper/bearing block when I get running . If I do it like yours I will get an extra .400-.500 in the Y . If I got a longer lead screw too I could get another .75 .

Just to let you know , if your happy with the Y just moving till it is flush with the front of the base I got 4.5" of travel with no interference with the portion everyone is cutting out . I dont understand why CNCFusion didnt make the Y screw longer . I paid extra for it . Theres plenty of room for .5 more of screw . Thers an extra .25 towards the column it could have gone . Check out pic .

[djbird3]

Looks good Don . Those are some awesome results . Makes me bummed that I already have the G540 . Ohh well , I need simplicity for my first build .

Heres a link to my thread . Check in once in a while . I appreciate input from all directions .

http://www.cnczone.com/forums/bencht...cnc_build.html

I had a question on wiring . I heard of a need for shielded wire on the limit and home switches so they dont catch interference from the steppers . What is the preference for stepper and limit/home switch wire ? 22ga. good ? Shielded/Unshielded ?

Also I see you did a design on the Y to allow the table to come towards you more . Wish the CNCFusion kit did that . I am going to make a new stepper/bearing block when I get running . If I do it like yours I will get an extra .400-.500 in the Y . If I got a longer lead screw too I could get another .75 .

Just to let you know , if your happy with the Y just moving till it is flush with the front of the base I got 4.5" of travel with no interference with the portion everyone is cutting out . I dont understand why CNCFusion didnt make the Y screw longer . I paid extra for it . Theres plenty of room for .5 more of screw . Thers an extra .25 towards the column it could have gone . Check out pic .

Hey,
I checked out your thread and subscribed to it to stay up to date with your progress, looks like you're well underway. I was looking at you're bench, looks pretty sturdy but if I were you I'd put a few cross braces on the legs because that machine will yank that table back and forth when it gets moving. The guage wire that you use for the limit switchs isn't really important because it only has to handle a signal thats in the micro-amps. I'd say anywhere from 26ga. to 22ga. shielded would be best. If you go thinner than 26 the wire is too fragile, any thicker than 22 is overkill. Shielded isn't always necessary but I would recommended using it to prevent any mistriggering on your limit inputs caused by crosstalk between wires, especially when there bundled together alongside the motor wires.

There's a few things CNC Fusion could do to improve their kit. The main thing would be to change the way the Z axis ballnut mount attaches to the head. If you want to adjust your gibs you have to take everything apart to get to the screws, kinda dumb. Why didn't they do like I did and make a threaded insert that presses from the inside of the head allowing the main mount screw to be on the outside where it is easy to access instead of it being on the inside of the head where it's a b1tch to get to? This would a big improvement that would cost hardly anything for them to do. All other aspects of their design looks fine to me but I don't understand the deal with the Z ballnut mount.

I guess because their kit is meant to be a bolt on type of solution they decided to keep the screw length on the Y just long enough to give the normal range of travel and avoid any necessary modifications to the mill. Also, they're probably trying to save on materials, they are trying to make a buck and thats not easy considering the time it takes to make this stuff. My Y axis screw goes all the way to the back of the casting so the ballnut can't go past the end of the shaft. I was originally going to mill about 1/4" out the back so I could get that extra travel but decided not to do it. I still need to mill about a 1/4" or so out the front to get the travel I intended. I'm s probably going to fall about .2" short of my goal of 6" because I didn't mill out the back like I was supposed too but it's too late now because I already cut the screw to fit. The front bearing block for the Y axis is .8" long allowing for the extra travel, I could have gone longer but decided that ws adequate and didn't want the saddle to travel any further off the base dovetail.

I put a couple links below for reference regarding the shielded wire suitable for wiring the limit switches.

~Don~

2-CONDUCTOR SHIELDED CABLE W/ DRAIN | AllElectronics.com

RG-174 MINI CO-AXIAL CABLE | AllElectronics.com

[fatal-exception]

There's a few things CNC Fusion could do to improve their kit. The main thing would be to change the way the Z axis ballnut mount attaches to the head. If you want to adjust your gibs you have to take everything apart to get to the screws, kinda dumb.

I second that...

I don't know if they are still shipping 'self locking' nuts for the ends of the ball screws, but they don't stay tight and need to be drilled and tapped for a 6-32 setscrew to lock them in place.

Paul

[djbird3]

I just wanted to let everyone know today I retested the Z axis setup with Mach 3 kernal speed set to 60K so that I could get faster step speeds and was able achieve reliable rapids of 415 in/min with the KL-5056D driver/425 oz-in stepper combo. To recap the settings on the driver were: 2.7A (RMS) and 1600 steps/rev., the supply voltage was 48V, which is as high as the driver will allow before it errors-out. I have no intention of operating at this speed but just wanted to see what was possible and I'm rather impressed. I think the days of analog controllers are numbered when you compare the performance of the two.

~Don~

[LatheMaster]

Your right about my bench . I slapped it together in a couple hours . Still have some reinforcing to do .

CNCFusion has been following along and offered to send me a longer Y screw . That was pretty nice of them . There looking for feedback on the LMS machines conversions cause table size and travel are more than other versions of the X2 .

400in/min is impressive . Thats blinding and bench rattling speed on a small machine . Thanks for the links .

[Diginanda]

You helped inspire me to do my own build! http://www.cnczone.com/forums/bencht...tml#post879304 I am using a Hon 2-door locking file cabinet with a 2' x 4' x 3/4" piece of birch ply as a top as my base. Since I'm 6'4" tall it puts the mill at the perfect height for me.

[djbird3]

You helped inspire me to do my own build! http://www.cnczone.com/forums/bencht...tml#post879304 I am using a Hon 2-door locking file cabinet with a 2' x 4' x 3/4" piece of birch ply as a top as my base. Since I'm 6'4" tall it puts the mill at the perfect height for me.

Hey, glad to see you got yours up and running. Looks like it's a perfect fit for that small room you've got it in. What kind of rapid speeds are you getting with that Probotix setup, the Z axis in particular?
~Don~

[Diginanda]

I am very pleased with the speed, definitely not servo speeds, but quick enough for a small bench top machine. I'll see if I can put together a couple of YouTube videos doing some real cutting.

[djbird3]

It's been a while since I posted so here's a few new pics of what I've accomplished so far. The first pic shows the completed parts I designed for the CNC conversion. The second pic shows the modification made to the saddle to allow the ballnut to fit under the table. The third pic shows the modification made to the base to allow greater Y axis travel, about .350" was milled out. I decided to change a few dimensions on a few of the Y axis parts to allow a little more travel but haven't remade the parts yet. The rest of the pics show everything installed on the mill including the motors. I'm currently in the process of dealing with mounting the limit switches and rounding up the rest of the parts for the control box.

The motors I decided to use for the X and Y axises are the Keling 282 oz/in single shaft steppers along with KL-5056D Drivers. I tested the steppers with the current set to 2.7A RMS and 1600 steps/Rev. with a supply voltage of 45V and was able to get 450 in/min., which is as high as mach3 would allow with the kernal set to 60K. I then put a 35 lb. bar of aluminum on the table to see if it could handle the extra weight at that speed and it managed with no problem, and this is with the gibs very tight. Since I only intend to operate the mill feed rates at a max. of about 250 I think I have more than enough toque to break any tool I'll be putting into the spindle. I think the KL-5056D drivers made the difference with this setup and would recommend anyone looking for a performance increase to try one out. I'll be putting up some more pics showing how I mount the limt switches as well as the controller box construction, diagrams, etc. in good time, untill then...

~Don~