[Dan2464]

Hi All,

Got the old Harbor Freight mini last year and the forever famous HOSS Freak conversion got me roped in.

So my take on this was to convert without using ball screws and go with 5/8 precision ACME 8-1from Mcmaster Carr and home made Delrin bushings and belt reduction. I had read a post on making these zero backlash bushings so I got a box of cuttoffs of 2" dia by around 3-4" long pieces cheap plus 6' of ACME for $43 from Mcmasters and away we went.

Machine specs are KL23H276-30-4A 282oz steppers with Leadshine DM856 drivers off of Ebay $42ea and a $12 BOB off ebay that seems to work fine. I have a PMDX-122 if it doesn't but I wanted to see if these cheap boards work out. I have a 70V toroidal power supply from Anek and for now I'm running the 5V and the 12v for the BOB off of wall warts. Anek modules for the 12V and 5V are on the way. Running 2.75 reduction on the X and Y with 2.25 on the Z. I got the longer Y base from LMS and spaced the head out 1" to match.

I have plastic gears on it for now as I couldn't find what I really need in metal since one of the points of this build was to be able to do it without a lathe to turn down the screws. Once I'm fully up and running I'll cut my own with a 4/5 axis I designed to prove the concept. The concept was could someone with just a drill press, and chop saw and a dremel build the CNC parts and have a good base to move forward after and then make the other parts.

So far everything has worked out very well and I'm at the stage of working through the electronics. The Delrin bushings are everything you could ask if they last in use. The only snag so far is the bushings are pretty tight and the motors are losing steps some of the time due to this.... I think..... but...... tonight while at work I figured out I don't have the settings for the steppers set right in Mach3 and that may be part of the problem. I was amazed that after doing the math that the X and Y are 44000 steps per inch with 10 micro steps set on this build. I didn't have anything near that in Mach so we'll see what the steppers do once I have the correct values for things on the drivers and in Mach.

Or it could be that the gear reduction doesn't give me the torque advantage I was hoping for and I'll have to upgrade the steppers from 282oz to something better or the drivers are cheap knockoffs and don't run the steppers correctly. I'll get it figured out.

Of course here are some pics

[Dan2464]

A few more pics of sorting out the build

[Dan2464]

Had some good results today. I got the drivers and Mach3 all set with the correct numbers and the steppers are doing much better. I still think I'll have to go bigger on the Z axis but we'll see. I think the combination of the tight bushings without a good lube and the ways adjusted tight are the problem. I don't want to chase the threads in the bushing as I like where they are at for fit right now. Very firmly tight but I can still turn the screw by hand but it's tough. I want to see if and how they wear in over time.

One of the benefits of using Delrin/Acetal bushings on the lead screws is they are tight with zero backlash. I tried a light oil on them for lubrication but they fit so well the oil is just wiped off as the screw runs through the bushing. The Z axis bushing is 1.75" so not much oil makes it to the other side with the bushing acting as a wiper. So I tried some anti seize compound and it works pretty good. Without the anti seize the friction of the Acetal on the lead screw would warm the screw. I picked up some good old powdered graphite and I'm going to add that to the mix.

So I got to use the Y and Z on the mill to face the plastic X drive gear and it worked very well. Had to go to work so hopefully tomorrow I'll get all the gears faced and all the drives on to cut a file for one of the pieces to this conversion.

Here are some pics of the tap you can make to cut the bushings from the lead screw you buy if you do this. For the 5/8 in lead screw on this project I used a 1/2 in hole then little by little run the tap through. Little bit in and then take it out, clean the slot in the tap then back in, rinse, repeat.

The second picture is cutting the threads in the Z bushing on the mill so the hole is perfect. I also used the top plate of the CNC bracket with a drill bushing to cut the 1/2"hole in the bushing with it in place at the top of the travel. That's the 3rd picture.

[Dan2464]

Got everything together and running today and air cut a 4" circle. Still losing steps here and there. Ordered three 425oz motors from stepperonline off ebay. Not a bad price for the 3 at $120 shipping included. I'll keep the 282's for a 3D printer. They should be more than enough for that.

I think the combination of running the gear reduction, tight ways and lead screw bushing and not having true bearings on the belt tension adjustment yet doesn't leave much to just drive the cut. If I run around 15 ipm it all works pretty good but I want to be able to run around 25 to 30 ipm reliably. So another confirmation that you should get the larger motors from the jump.

Now it could still be the drivers aren't performing (they seem to be fine) so I will be getting some Gecko 203V drivers to compare. I will need the Geckos for a router build anyway and it's been fun to try out the cheaper boards and drivers to see where this stuff has got to over the last year. I would have never thought the electronics would ever get to this level with the flood of china made components.

Once the new motors are on and performing as expected then I can cut all of the components of this build so I can incorporate design items I couldn't create on a drill press. CNC makes things like flange bearing pockets so much easier. And the custom gears for the gear reduction.

All of the electronics is in an ATX case with the motherboard. Need to tidy up the wiring but still waiting on the 12v and 5v modules to replace the wall warts.

[arizonavideo]

I made Delrin lead screw for a few axes on a IH mill and smaller ones too. You will get some flex and lash but they do last a long time after the screws smooth out.

That is one problem. The screws are some what ruff when new so will eat up the first set of nuts fairly fast. Once they smooth out they will have almost no ware and will last for years.

You still need oil on the screws.

High speeds will over heat the nuts and they can melt. You should not have that problem with the reduction your running but that is another problem.

With a 10 TPI screw and any stepper you should never gear them down. Your top rapid speed will be poor and even at 20 IPM your running the steppers way past there maxim torque output.

I did the same mistake with a CNC lathe attachment and 10 TPI ball screws. I wanted slow but I got really really slow.

You might think of changing the reduction to 1.5 to 1 at the most and even go direct drive if it fits the mounts.

[Dan2464]

Thanks for the info arizonavideo.

I was looking at the rpm scale and its 330 rpm at 15 IPM and 440 rpm at 20 IPM with the gear reduction I have That doesn't seem too high. I looked at a torque curve chart and you lose 1/3 of the torque from 100 rpm to 450 rpm. I'm thinking that with the larger motors it will be fine since these 282's almost get it done on the 8 TPI screws.

I was looking at a thread were the op was trying to cut a turbine blade for a model jet engine and the gear reduction plus the 4/5 axis I'm working on will give me the resolution to cut these.

[arizonavideo]

You have done a lot of work on the mounts. With larger motors you can go direct drive on the X and Y. If it works out better to leave belts the you should go closer to 1.5 to 1.

Larger steppers should have higher inductance so have peak power even lower RPM but might make a bit more at 400 RPM than the 282 but your spending a lot and not getting much.

The 282 with 2 to 1 would be about 560 oz with a 10 to 1 screw that's like 900 pounds of force.

If anything get the lowest inductance steppers or highest current steppers you can. Its OK to drive a 5A stepper at 4A. It will just run cooler and make less noise and be faster.

[hoecken]

The higher the RPM of the stepper the less power you get from them.
Bigger steppers might do it but you need the right drivers for them.
Gecko claims their drivers work for any stepper.

[RTP_Burnsville]

What process did you use to make your ball nuts? Did you split and melt the plastic or drill and tap? I have been messing around with the split and melt process but still don't have anything whic h is acceptable. BTW: I also have a X2 with a conversion about 1/4 complete.

Thanks,
Robert

[109jb]

What process did you use to make your ball nuts? Did you split and melt the plastic or drill and tap? I have been messing around with the split and melt process but still don't have anything whic h is acceptable. BTW: I also have a X2 with a conversion about 1/4 complete.

Thanks,
Robert

I would say he drilled and tapped since he shows the screw he modified into a tap and he talks about it in post #3:

Here are some pics of the tap you can make to cut the bushings from the lead screw you buy if you do this. For the 5/8 in lead screw on this project I used a 1/2 in hole then little by little run the tap through. Little bit in and then take it out, clean the slot in the tap then back in, rinse, repeat.

[RTP_Burnsville]

I would say he drilled and tapped since he shows the screw he modified into a tap and he talks about it in post #3:

That could be, but some have done the melt and form process followed by using a custom made tap.....

[109jb]

That could be, but some have done the melt and form process followed by using a custom made tap.....

OK. Here is exactly what you asked:

What process did you use to make your ball nuts? Did you split and melt the plastic or drill and tap?

And here is exactly what the OP said in post #3 which I quoted in my last post:

Here are some pics of the tap you can make to cut the bushings from the lead screw you buy if you do this. For the 5/8 in lead screw on this project I used a 1/2 in hole then little by little run the tap through. Little bit in and then take it out, clean the slot in the tap then back in, rinse, repeat.

So where is the "could be" in this? He absolutely used the tap method and he said so. Re-read the post and it is abundantly clear.

[RTP_Burnsville]

OK. Here is exactly what you asked:



And here is exactly what the OP said in post #3 which I quoted in my last post:



So where is the "could be" in this? He absolutely used the tap method and he said so. Re-read the post and it is abundantly clear.

What's with the attitude?? I just misread the original posting so take a breath or get a life!

Robert

[109jb]

What attitude?

You said "That could be", as if you doubted what was posted, so I pointed it out again. I think you are reading too much into a simple post. Thin skin?.

BTW, my life is just fine.

- - - Updated - - -

What attitude?

You said "That could be", as if you doubted what was posted, so I pointed it out again. I think you are reading too much into a simple post. Thin skin?.

BTW, my life is just fine.

[Dan2464]

Sorry I haven't checked in for awhile guys. Had to take care of some family stuff and it's all good.

We have made some great progress on the mill. Got the larger motors installed and running and that seems to have cured the stall issues I was having. Just to stress this point so folks that read this thread understand. I didn't do direct drive steppers for a reason. With belt reduction you get true resolution and the ultimate goal is to be able to cut the fan blades for model turbine engines on a 4/5 axis I designed. The first step is to have a mill to cut those other parts. I also set the stepper drivers to just 4 step micro stepping. With the belt reduction you don't need as much micro stepping and I read that it helps with the motor torque.

I also want to test out just how solid and accurate the Delrin/Acetal bushings are and if they will last in use. Total cost in my mini mill is maybe $90 for the screws and bushing material shipped and I have enough bushing material for 20 more nuts for next to nothing in cost.. With the way I implemented the nuts I have almost a 2" nut in the Z axis with a solid 1" in the Y axis. The X axis has the same thickness as the original so that one will get a real test on how well it wears. When the time comes I'll do a small mod to where it mounts and increase the size. I wanted to try to do minimal mods to the castings so anyone wanting to try this could do so without a ton of other tooling to make the nuts fit. It looks like the Anti-seize compound and good old Shell Rotella motor oil is the ticket for lube on those.

As far as the bushing question I did not use any heat to form them. Just the tap made from the same lead screw you would buy. Just take your time cutting the threads. The groove you put in the screw when you make a tap won't contain alot of chips. So turn it a little in and then take out rinse and repeat. Got a 6' chunk of precision 5/8 from McMaster for less than $50. It works good and you can make all of your screws and still have plenty left over for the tap. So far all of the testing with these plastic nuts is great. I clamped a digital set of calipers to the table and did a 1" and 2" jog back and forth and the X and Y come back to dead on zero every move. Since I don't have the final aluminum gears in the reduction drive yet I can't snug up the lash adjustment for the lead screw bearings as tight as I would like. The gears don't run true but even with that I have .001 back lash on the X and Y due to the gears and not the nuts. I found the gears I need so I can cut the bearing pockets the way I want and will get the final results of the backlash and movement ability. And guess what?? I now have an operational mill to do the work on the gears

So after much setup and checking everything out I got some chips flying today Attached is a picture of a test piece. It was 3 different operations with the engraving, hole cut out and drilling. Great fun. I think the hardest part of this so far is doing the drafting and learning CamBam to make the G code. It is starting to make more sense when you can see what happens in the final piece.

I still have some more testing and I would like to figure out a good test to pit these cheap Leadshine drivers against some V203 Gecko's that I'm going to get. Part of this project was to see what can be done with the China stuff but I still wonder if the motor issue is the drivers and not the motors. I can still stall these 425's easier than I would think BUT I do have very tight leadscrew nuts and the all of the ways are snug and the belt adjusters are not running on bearings yet. I know for a fact it's not the power supply as I put in a toroidal 70V 500watt power supply after playing around with the little 48V 7A regulated unit I got to test with. I got five DM 856 Leadshine drives on ebay for $42 each shipped. Gecko V203's are $130 each so I figured nothing to lose testing out these Leadshine's. Also I got four China "Mach3 Interface boards" for $14.78 ea shipped. I did burn one up messing around with some DIY drivers but the one running the mill has been plug and play. So I bought four boards for less than the cost of my PMDX 122 I have still in the wrapper.

SInce the CNC bug has bit I'm looking at building a Delta Printer from scratch since I now have the mill to cut the parts and a CNC router. One way or another all the electronics I'm testing out will be used I'm sure.

Later guys.

[Dan2464]

@arizonavideo As far as the steppes are concerned the 282 steppers are 4.7 inductance which according to Gecko's calculation comes out to 69.37 volts and they are 3 amp motors. I started with a 48V 7A regulated power supply that you see packaged with the kits. Once I had the problems I wanted to be sure the power wasn't the issue and installed the toroidal 70V power supply. Even with the belt reduction still had issues and I'm not spinning the motors over 400 rpms. According to the calculator with 1.8 steppers using 4 micro steps on the .125" per turn lead screws and 2.75 to 1 gear reduction 16T on the stepper and 44T on the screw the X and Y motors will be turning at 330 rpms at 15" per minute. With what little cutting I have done I would love to cut at that rate but haven't yet as 10 or lower seems to be pretty fast for right now while I'm learning. Plus I'm betting to get above 15" per minute I'll need a faster milling motor in aluminum.

Now the 425 steppers I just put in are 3.8 inductance which calculates out to 62.3 Volts and their 4.2 amp motors. Running them at 70 volts should still be fine as long as I watch the heat and current limit them if needed. So far I'm running the drivers at the highest amp setting which is 5.6A peak with 4.0A RMS. Hottest motor so far has been 115 F. We'll see what they do on a long cut one of these days.

[arizonavideo]

You could run lower than the 5.6A peak.

I would set current to have the steppers be at a temp you think is good. Power at speed will be determined by the voltage which you have already at a fairly high level. With reduction you will have ample power at lower RPM.

[Fastest1]

The pc itself can cause issues too. Especially if you are using a PP.

[Dan2464]

Got the aluminum gears for the Z axis in and used Fusion 360 to model and then CAM and post the gcode to face the hub off and pocket and machine the hole. Looks like a great CAD/CAM package. Waiting on the gears for the X and Y and then this belt drive conversion will be where I can test out all the tolerances. So far with the crappy plastic gears it has been great. Should be spot on once I can adjust the bearing preload on the screws with gears that run true.

First pic shows the gear with the pocket for the bearing and washers they run on. I had originally wanted to pocket both sides but the 1/4 - 20 set screws wouldn't have enough material if I did that so only one side will have a pocket the bearing will run in.

Second pic is the gear on the shaft and getting ready to run the set screws in.. Bearing stacks on both sides.

[Dan2464]

Got the Z axis put back together and very happy with the results. Ran 2 tests with it, a 1.500 move both ways and .001 steps. The 1.500 moves dead on to 0.0005 and then to 1.5000 every time and the .001 steps are pretty much right on the money. It's a mystery to me that the 1.500 move will get to the same readout every move but when you change direction doing .001 steps it looks like there is .0015 backlash somewhere in the system. It takes a .001 step to get the lash out then on the second step the dial indicator moves .0005. Wherever the backlash is, after it is taken out then each step is .001 on the money.

More pictures of the X axis parts and the video! https://youtu.be/UXK7ot34OPQ