[imprez55]

This post is kind of long, so if your lazy and still want to help there is a TLDR at the bottom.

My plan is to CNC a X3 from Grizzly while still keeping the stock ACME leadscrews. I will probably update this later with ballscrews but for now my plan is just using most of the stock components from the X3 and retaining manual use. (side note, would the opto-isolation on the Step/Dir pins of the G203V drivers make it so spinning the motors by hand and thereby producing a voltage would not damage the drivers or circuitry?)

First off is the PSU. I have three options here: converting either a 70V @ 17A or a 48V @ 12A transformer. Or I could purchase a 48Vdc 8.3A transformer PSU. The kelinginc power supplies (the cheap ones) do not have enough amperage for the motors I would like to get, but based on those motors the 70V power supply is too much voltage, although I could just add in some resistors to lower it and still have a great PSU for future use.

Next up is the BOB. I have seen the most common two recommendations, in order, are the PMDX-132, the C11G. It seems to me the PMDX board is better overall and I love the fact that I can use 110VAC for power. Its downfall is that it costs at least another $33 for a c6 board for spindle control and its already the most expensive. The C11G seems very up to the task and already has spindle control built it which is a big plus at a cheaper price. I am leaning towards the C11G but I would like any input on a recommendation for a BOB.

I have pretty much decided on the Gecko 203V drivers. After looking over the specs and comparing it with others it is by far the most capable driver and maximizes stepper output. Its downfall is the cost, but I have no problem spending that on a quality driver and supporting new innovation of that sort. Although it might be much more cost effective to use a G540 if I can find a proper Z-axis stepper...

I have picked out a few motors from Keling inc (who seem to have the best selection for the price and I am happy with previous purchases from them). For the Z-axis I have chosen the Nema34 KL34H280-45-8B which has 640oz/in or torque at 2.2mH of inductance, and for the X/Y-axises the KL23H284-35-4 which has 387oz/in torque at 2.8mH of inductance. Depending when they are available (hopefully soon) I would much rather have the KL23H2100-50 on the X and Y axises. The website says its in development and boasts 570oz/in of torque at 2.5mH of inductance all in a Nema23 frame!

To control the machine I am leaning towards using timing pulleys on all 3 axises. This will free me up to use the manual side of the mill without needing the high precision of the direct drive (and not needing to buy pre-made mounts). I will keep it at a 1:1 ratio because the steppers I chose will hopefully be able to push it well enough on the stock leadscrews (I know people have used them for ballscrews, not sure on ACME). The main advantage here is keeping the motors out of the way.

TLDR:
CNC'ing an X3 with planned:
48V@12A or 70V@17A (resisted to less than 50V)
PMDX-132 or C11G (prefer 110VAC of PMDX but want the spindle control of the C11G)
Gecko 203V or G504 (pending the finding of a 3.5A ~600oz/in stepper)
pulleys on all 3 axises with a 1:1 ratio
:TLDR


Please post any recommendations/preferences you guys might have on any of my choices, thanks!


List so far (in USD):
1x PSU_____________= 71 or 110 (depending on how powerful I go)
1x C11G/PMDX-132___= 109/183 respectively
3x 203v____________= 417 (289 for a G540)
1x 640 stepper______= 89 (what to do with this if using a G540?)
2x 387 steppers_____= 98
pullys/belts______~ 50? (no idea yet really)
===================================
__________________ = 834-947
P.S. Does anybody have any experience with homeshopcnc? They have $73/76 single/dual shaft Nema23 motors with 570oz/in of torque at 2.3mH with 203V's at $129. That looks like a better buy and they have the 570 steppers developed already.

[SSN Vet]

I considered the same "upgrade to ball screws later" approach but decided to bit the bullet and do it all at once. Good deals on nuts and screws can be had...

if you stay with your plan, your going to want to come up with motor mounts that you can use now and then re-use on ball screws later (unless you're into repeating the efforts)...

here's a suggestion for you for X & Y.....

http://www.jeffree.co.uk/pages/x3-to-cnc.html

[imprez55]

Sorry its taken so long to respond, I have been quite busy the past few weeks. Thats an interesting read you linked to Matt. In it he talks about how we overestimate the power of the steppers needed to drive even the x3. If I could move down in power and stay with the g540 with weaker steppers I might be able to justify another $949 (I have no way to turn ballscrews so the only option I am aware of would be to do the deluxe cnc fusion kit).

I am not totally against directly connecting the motors, I just thought it might be more difficult seeing as how this would be the only real metal working machine I would be able to use and I thought the direct connection required more precise drilling of the mounts into the base.

*Edit*
How are the other kits from CNC Fusion. I would prefer not to pay nearly the price of the mill to upgrade it to ballscrews, but the next step down at $569 seems more reasonable, in fact I might even want to just purchase the ballscrews and make the mounts myself (If they had a package with just the deluxe ballscrews with no mounting hardware that would seem right on the money)

[Teyber12]

Sorry its taken so long to respond, I have been quite busy the past few weeks. Thats an interesting read you linked to Matt. In it he talks about how we overestimate the power of the steppers needed to drive even the x3. If I could move down in power and stay with the g540 with weaker steppers I might be able to justify another $949 (I have no way to turn ballscrews so the only option I am aware of would be to do the deluxe cnc fusion kit).

I am not totally against directly connecting the motors, I just thought it might be more difficult seeing as how this would be the only real metal working machine I would be able to use and I thought the direct connection required more precise drilling of the mounts into the base.

*Edit*
How are the other kits from CNC Fusion. I would prefer not to pay nearly the price of the mill to upgrade it to ballscrews, but the next step down at $569 seems more reasonable, in fact I might even want to just purchase the ballscrews and make the mounts myself (If they had a package with just the deluxe ballscrews with no mounting hardware that would seem right on the money)

Just fine. If you get the #4 kit and decide you want a higher end kit later on, for a reasonable cost he should swap kits with you.

Hes a really good guy, i got the #4 kit and just sent it back to him for a reasonable cost to have my x and z double nutted and the y axis preloaded.

I can only say great things about cncfusion

[imprez55]

After looking into ballscrews more (i'll admit, i did not know much more than increased accuracy and increased costs before) I am definitly leaning towards the deluxe cncfusion kit. I already justified a small price increase to the base ballscrew kit they have, and considering the following logic the "small" $389 price increase does not seem bad at all.

Using the experiences from the site that Matt linked to and my knowledge on ballscrews, the G540 should have plenty of power available to run a X3. The steppers chosen already have a voltage cap between 47v (2.2mH) and 53 (2.8mH) so the 50v cap on the G540 should be plenty of volts. The amps are always the killer when thinking on this route, but because the X3 will already have a ballscrew conversion, the torque needed to move the axises is lower. So the 387 oz/in (KL34H280-45-8) will still work for the X/Y and for the Z an under-amped 906 oz/in (KL34H295-43-8) would seem to do the trick by producing 550 oz/in at 3.5A at only $10 more! This to me nearly negates a lowering amperage, fits the CNCFusion kit, and still applies the "over powered" notion to steppers (I am going off of that article still).

So monetarily the $417 for G203V, $135 for a BOB (I decided on the UBOB before I did some math with the G540) turns into $289, a reduction of $263. Now its only a $126 difference between the deluxe kit and the #4 kit.

Updated List:
1x PSU_____________= 59.99 shipped (DIY 48V 12A)
1x G540 ___________= 289
1x 906 stepper______= 99
2x 387 steppers_____= 98
1x CNCFusion #6 kit__= 949
1x HF X3___________= 879.99 (best conditions: 20% off coupon -8% sales tax and no shipping)
_____________________________
= 2374.98 as opposed to 2036 with previous list and grizzly mill (still deciding between companies). That means effectively the deluxe kit is only about $300, which I am very satisfied with.

[slashmaster]

Hey Imprez55, I too am planning to convert my X3 and also don't want to pay for new leadscrews so this thread interest me. Is your mill one of the older ones that has an air cylinder to push up the Z axis or one of the newer ones where you crank twice as much to move the Z axis the same distance? Have you considered Probotix? I don't know if there is anything wrong with them or not but looks to me like you could get everything you need from them way cheaper.

[imprez55]

I have not previously considered Probotix, and looking at their packages I still do not consider them a very good choice. I hold Gecko to a very high standard. They have great products that are backed up with quality and reasoning instead of just inflated stats. They are active on these boards and are very helpful.

The Probotix drivers have a max of only 2.5A and after only a brief look the motors are underpowered by the driver and have a high inductance. It might be a nice kit for an 8020 and MDF router build but its not very appropriate for an X3 conversion IMO. The z-axis on the X3 would not be very happy with an underpowered 400 oz/in stepper with 6.8 mH inductance anyhow.

Unfortunately I do not know if I will keeping the stock leadscrews anymore. After looking into cutting a few bucks from other places it doesn't seem that far out of reach.

[cyclestart]

Keeping the acme on the z would provide a lot of mechanical advantage over a ballscrew. Another option to consider. There is a speed penalty.

S J H builds beautiful machines. Pity some of the photos are missing in this thread
http://www.cnczone.com/forums/showthread.php?t=24983

[SSN Vet]

Using the experiences from the site that Matt linked to and my knowledge on ballscrews, the G540 should have plenty of power available to run a X3. The steppers chosen already have a voltage cap between 47v (2.2mH) and 53 (2.8mH) so the 50v cap on the G540 should be plenty of volts. The amps are always the killer when thinking on this route, but because the X3 will already have a ballscrew conversion, the torque needed to move the axises is lower. So the 387 oz/in (KL34H280-45-8) will still work for the X/Y and for the Z an under-amped 906 oz/in (KL34H295-43-8) would seem to do the trick by producing 550 oz/in at 3.5A at only $10 more! This to me nearly negates a lowering amperage, fits the CNCFusion kit, and still applies the "over powered" notion to steppers

Looks like you and I will have very similar builds... mine is green though, as I couldn't get either of the HF guys in CT to bite on the 20% coupon (which says right on it.. "not good for special orders"... which the mill is, as they don't stock it).

We'll be among the first to do an X3 conversion with the G540.

I need to update my build blog as I have finally made some good progress with my stand.

Good luck to ya'

[slashmaster]

Just what stats did Probotix inflate? You're saying the 400 oz in steppers are not actually 400 oz in? Do you have a link to the cnc fusion kit # 6 you're thinking about? I only see 1-5, are you talking about the deluxe kit that cost $949? You're spending almost 1k and still don't have stepper motors or any electronics for it. You sure it's worth it?

I have not previously considered Probotix, and looking at their packages I still do not consider them a very good choice. I hold Gecko to a very high standard. They have great products that are backed up with quality and reasoning instead of just inflated stats. They are active on these boards and are very helpful.

The Probotix drivers have a max of only 2.5A and after only a brief look the motors are underpowered by the driver and have a high inductance. It might be a nice kit for an 8020 and MDF router build but its not very appropriate for an X3 conversion IMO. The z-axis on the X3 would not be very happy with an underpowered 400 oz/in stepper with 6.8 mH inductance anyhow.

Unfortunately I do not know if I will keeping the stock leadscrews anymore. After looking into cutting a few bucks from other places it doesn't seem that far out of reach.

[imprez55]

Just what stats did Probotix inflate? You're saying the 400 oz in steppers are not actually 400 oz in? Do you have a link to the cnc fusion kit # 6 you're thinking about? I only see 1-5, are you talking about the deluxe kit that cost $949? You're spending almost 1k and still don't have stepper motors or any electronics for it. You sure it's worth it?

I was not necessarily saying the Probotix inflated any of their stats, that was just reasoning behind the choice for Gecko. That post was not meant to knock Probotix, as I have seen their Fireball in action and it seems perfect for what it was built for. THey are a good choice, but not for this build. However, their website does say the "green monster" steppers are rated at 3A while the kit only provides drivers with a max of 2.5A. So I do not see the 400 oz/in steppers being able to provide that much in the kit they made.

Sorry for the confusion, yes the #6 kit is the Deluxe kit. I was going off of memory and I just assumed since they had a 1-5, and the deluxe is still a kit, it would be the 6th. The price is what I am trying to balance now. The inherit properties of ballscrews might allow me to reduce the power needed to drive the axis so I can cut costs other places. I am still uncomfortable making that purchase quite yet. In the 5th post near the end I try to compare costs between getting the ballscrews and cutting corners other places. In the best conditions for purchasing the mill (which I still have not) it would only be roughly a $300 increase to get the deluxe kit as opposed to keeping stock leadscrews and the G203v's. So in short, I do not know if it is worth it yet, but I am leaning towards yes.

Keeping the acme on the z would provide a lot of mechanical advantage over a ballscrew. Another option to consider. There is a speed penalty.

S J H builds beautiful machines. Pity some of the photos are missing in this thread
http://www.cnczone.com/forums/showthread.php?t=24983

What kind of mechanical advantage are you talking about in regards to keeping the z-axis Acme screw? In the acrticle you linked to S J H says his IPMs are 55 less on the z-axis, and backlash is 4 - 6.7 times as much also. It would be an interesting thought if I knew the advantage.


Matt McColley: My planning is so far matching your build log to the T! I have not seen it before, but it interests me greatly. With how helpful the CNCFusion guys are they might even go for a switch on the pully ratio in the kit

[cyclestart]

What kind of mechanical advantage are you talking about in regards to keeping the z-axis Acme screw?

Mechanical advantage in the simple machine definition of the term. The ball screws are ~5 tpi and the acme is 10(?) tpi. However the ball screw has much higher mechanical efficiency due to less friction. Maybe someone will chip in with some real world numbers. The ball screw probably needs less power to raise the load but it would be close. The acme will have superior holding power.

In the acrticle you linked to S J H says his IPMs are 55 less on the z-axis, and backlash is 4 - 6.7 times as much also. It would be an interesting thought if I knew the advantage.

He stated a backlash of .002" or less. That's a very usable figure with software compensation. Frankly I'd be surprised if the best cncfusion kit can beat that by 6.7X or even 4X measured at the head.
btw I use a cncfusion kit on my X2 and have nothing but good things to say regarding the company.

It would be an interesting thought if I knew the advantage.

It would be interesting if I could explain it. Maybe it's just a red herring. If you don't mind splashing out for the deluxe kit and the 203's it should be an awesome combination.

[SSN Vet]

I picked up the X & Y CNC Fusion kit #4 components second hand from a guy off of these boards.....

My thoughts now are to set up X & Y immediately with these components and to either buy of fab. the Z axis.....

Once I've got a handle on "real life" backlash and travel performance, then I'll decide if I want to use the mill in that config. to make the upgrade components for a set up like CAD Monkeys.....

[imprez55]

I picked up the X & Y CNC Fusion kit #4 components second hand from a guy off of these boards.....

My thoughts now are to set up X & Y immediately with these components and to either buy of fab. the Z axis.....

Once I've got a handle on "real life" backlash and travel performance, then I'll decide if I want to use the mill in that config. to make the upgrade components for a set up like CAD Monkeys.....

What kind of time frame are you looking at to set everything up and test? That would be a good determination for me on what to purchase if it is relatively quick. I would like to purchase everything by the end of June and have everything running smoothly within maybe two weeks after. Also, what does CAD Monkey's setup look like?


One thing I have overlooked in this is an enclosure and how to protect everything from flood coolant. Hmmm, no real access to metalworking tools or welders so it has to be wood. I think a design similar to hoss's, with the bed liner and removable walls would be good. I would prefer to simplify it with only one door, possibly a roof too so coolant doesn't splash out. I just need a CAD drawing of the X3 so I can design some things, I wish I had the mill on hand to take make one.

[hoss2006]

One thing I have overlooked in this is an enclosure and how to protect everything from flood coolant. Hmmm, no real access to metalworking tools or welders so it has to be wood. I think a design similar to hoss's, with the bed liner and removable walls would be good. I would prefer to simplify it with only one door, possibly a roof too so coolant doesn't splash out. I just need a CAD drawing of the X3 so I can design some things, I wish I had the mill on hand to take make one.

The herculiner has held up well after a year and a half but it's messy messy stuff to apply and has a rough finish that holds onto the swarf.
I would look into something with a smoother finish that would clean up easier like this bathtub refinishing coating.
I used something like it from home depot on a laundry tub and it's very tough but smooth.
Hoss

[imprez55]

Thats not a bad idea with the glossy stuff. I did see in your videos how the coolant even seemed to pool up in places. I will look into that more, maybe just a high gloss paint could do the trick.

Another concern I have is if I put the stand on wheels, will that make it unstable during use and cause accuracy/precision problems? The base I am planning on doing is about 3x4 so that would also be the wheel base. It would be nice to be able to move it when necessary, but an extra $100 in leveling casters would just be an annoying expense. Anyone with rolling stands have any problems with it?

BTW, I love your idea with the power drawbar, Hoss. I am in the process of collecting parts for an x3 style one right now. Already have the air cylinder, solenoid and relays now with a running total at $41.05 with shipping.I wasn't feeling the lever arm and how much space it took up so I sprung for a larger 4" bore air cylinder for a whole $0.99 on ebay. The seller is selling some more at $19.99 if anyone is interested here. I'll report back to see how the quality is on it, but I have high hopes!

[SCzEngrgGroup]

What kind of time frame are you looking at to set everything up and test? That would be a good determination for me on what to purchase if it is relatively quick. I would like to purchase everything by the end of June and have everything running smoothly within maybe two weeks after. Also, what does CAD Monkey's setup look like?


One thing I have overlooked in this is an enclosure and how to protect everything from flood coolant. Hmmm, no real access to metalworking tools or welders so it has to be wood. I think a design similar to hoss's, with the bed liner and removable walls would be good. I would prefer to simplify it with only one door, possibly a roof too so coolant doesn't splash out. I just need a CAD drawing of the X3 so I can design some things, I wish I had the mill on hand to take make one.

Consider mist coolant using a FogBuster. A small machine doesn't need flood - it *can't* cut fast enough to require flood. I use mist on my knee mill, making cuts you could never make on a benchtop machine (1/2" endmill, 1/2" DOC, 30 IPM), and mist is more than adequate. Mist is clean, cheap, VERY effective, and makes almost no mess. The only downside is the noise.

Regards,
Ray L.

[imprez55]

Yes flood coolant is overkill for a small machine but in this case it is practical. I already need to set up a bench area so the only difference now is walls with a glossy coat of paint. Mist would be better if it wasn't so expensive. 300 dollars for the basic kit is not cheap versus $60 for paint, $20 for pump and then misc. piping that I already have.

I feel in this application it is simpler and cheaper to go this route. If you could prove me wrong then by all means do. It would be much less time consuming to just make a bench and set up the mister, but its not worth $200.

[SCzEngrgGroup]

Yes flood coolant is overkill for a small machine but in this case it is practical. I already need to set up a bench area so the only difference now is walls with a glossy coat of paint. Mist would be better if it wasn't so expensive. 300 dollars for the basic kit is not cheap versus $60 for paint, $20 for pump and then misc. piping that I already have.

I feel in this application it is simpler and cheaper to go this route. If you could prove me wrong then by all means do. It would be much less time consuming to just make a bench and set up the mister, but its not worth $200.

My FogBusters cost me less than $30, and a couple hours of my time - I make my own, working from their patent. They work great.

http://www.google.com/patents?vid=US...390854#PPP5,M1

This is NOT the same thing as the cheapo $60 mist sprayers! It puts virtually NO coolant in the air, cools extremely well, and uses very little coolant. It will create anything from an invisible mist to almost a full flood, though you'll never need that. Basically, if you can *see* the mist coming out of the nozzle, or you have coolant accumulating on your work, you've got it turned up too high. I rarely go through more than 1/2 gallon of coolant mix in a long day of very heavy cutting. I've been working my way through the same $20 gallon of KoolMist for several years now, and have created many hundreds of gallons of chips. On a benchtop mill, a gallon would be almost a lifetime supply. How much are you going to spend on a flood system in coolant, and filtering and skimming, aerating, and dealing with biologicals, etc.? That all goes away with mist.

Regards,
Ray L.

[hoss2006]

You don't have to worry about biologicals if you use the proper coolant.
I've been using the same gallon of Syn-Kool for almost 2 years now and still have 3/4 left,
I have yet to have to clean the filter as the strainer in the enclosure does most of the work.
The coolant tank is still fresh as a daisy and there is absolutely no odor.
The reality of things don't always match people's preconceived negative notions.
Hoss