Vectorsc’s G0704 Conversion

Hey everyone! First of all I wanted to say thanks for the inspiration and information to get this done! I don’t think I could do this without you all.

6 months ago I was fortunate enough to score a like new g0704 with an aftermarket base for all of $600. It had none of the tooling, and was missing anything that might have come from the factory. It had been used to mill plastic based on the chips i found under the ways cover.

My units build date was 2010.

After finishing the remodel on my house, I decided to get after it. I checked that the machine ran OK and other than that didn’t do much inspection or trimming. I figured it was all coming apart anyway.

I did a lot of research on this topic here and elsewhere -

Decisions made already, but always open to input:

-UCCNC control software
-Fusion360 for CAD/CAM
-C76 BOB and UC300ETH control
-Building the control into the bottom flip up panel of a craftsman rolling tool box. Allows me to put my tool holders with tool mounted on a rack in the toolbox, and a PC on top.
-Double ball nut screws
-Moved motor controller to the control box
-TTS tool holders and R8 collet
-Energy chains from Igus on Head and X
-2 3nM hybrid steppers, nema 23, for X and Y. RATTM
-1 8nM Hybrid Steppers, nema 34, for Z. I did this because of the upcoming PDB, sub spindle, and AC motor upgrade.

Things I still need to decide:


-Which CNC kit, Procut, ArizonaVideo, or Automation Tech
-Which coolant mechanism, mist or flood
- How to do the enclosure- I may actually build the unit it’s own room based on a shower stall for chip control and flood coolant recirc


Round two:
-I’ll be machining my own one piece head plate to do a motor mount/belt drive/PDB.
-ATC
-Sliding ways covers
-Head plates for the front and sides

Quick pics of starting my control layout and the rear down.

Edited to add that I ordered my motors!

I’m going to be doing a post on the motor controller wiring and the control box soon.

Ordered some cheap optical home switches off ebay.

Does anyone know what the wiring is for the stock spindle encoder?

I have a G0704 with 570 oz-in motors on it and I'm also converting a RF-45 size mill using 12N-m closed loop steppers on all axes. so I will give my spin on things.

First, I'm not sure if you are referring to closed loop steppers when you say 2 3NM hybrid steppers, as the 570 oz-in motors are also "hybrid" steppers. Hybrid refers to a combination variable reluctance/permanent magnet motor and has nothing to do with closed loop. This is a common misconception. I'll assume that you mean "closed-loop".

Based on getting the table moving on my RF-45, the closed loop motors are nice. I like the idea of closed loop mainly for peace of mind as the drivers can provide a fault indication back to the controller to stop all motion if they get out of bounds.

Having said that, the open loop 570 oz-in work great on my G0704 mill. I could run them at around 180 IPM rapids, but I have it turned down to provide a buffer for lost steps and also because I like to run my double nuts and gibs a little tighter which causes some additional drag. If you decide on open loop then the 570 oz-in work great. I have them on X, Y and Z. I know many people use a 900+ NEMA 34 on Z but I haven't found it necessary and I don't have any counterbalance on Z

In any case, a 2.3N-m motor is 325 oz-in, and I personally would not use that low motor torque on a G0704 table whether it is closed or open loop. If going closed loop I'd look for at least 400 oz-in and I see 425 oz-in closed loop motor/drive combos on e-bay. That might be a good choice for closed loop. What I think would be ideal is the 570 oz-in motors fitted with encoders and hooked to a closed loop drive, but that would require more work

One thing I will mention is that my closed loop experience is limited to the 12N-m system I have from automation tech. I was a little concerned about using a motor with high inductance, but the driver that I have connects to 110 VAC. I'm not sure about the internals, but if it is simply rectifying that it would equate to 155 VDC which overcomes the motor's high inductance. All I know is I can rapid the table of my RF-45 size mill around at 225 IPM with very little actual tuning of the accel or velocity, and without touching any drive parameters. I bought one initially and them bought 2 more after some testing. I don't have any experience with the more standard DC-power supply/Driver/motor closed loop systems. The drives I have state that they can increase current for a short time when steps fall behind which I'm not sure if the other style drives do or not.

In terms of the conversion kit, I made my own, but if buying one I think I would go with the Arizona video kit. It looks to be top quality and from posts on here he pays attention to things that other may not, like repacking all of the ballscrews to reduce backlash.

Mist or flood. Well, Mist works good but flood works better. However you absolutely have to have an enclosure with flood. With mist you could do a less substantial enclosure without drains/filters/ recover tanks/ pumps, etc. My crude mist system is a valve, regulator and a air tool oiler, all of which I had lying around. I use WD-40 in it and it works great. I would rather have both mist and flood though, but I don't have any enclosure on the G0704. Something I plan to rectify on the RF-45.

Enclosure. I would not go with the own room idea because it still means chips are getting on the floor and they do get thrown a fair distance depending on the tool. I don't like walking on chips. I actually have similar to what you are talking about in that my G0704 I usually run up against the back wall of my shop with a couple pieces of extra steel barn siding standing up on each side. It is still a PITA compared to what a full enclosure with a door on front would be. I would even go mist with a cheap coroplast enclosure over the room idea myself.

Yes - I mean closed loop steppers.

Great feedback sir - and the steppers I'm using are 3NM or 428oz/in. I meant to say quantity 2 not 2.3NM. Oddly I have heard that due to the closed loop nature, even the 2NM steppers supposedly work on the G0704.

For the enclosure I'll look and see what I can do to build something a bit better than a room. I don't think I would like walking on the chips.

This would be so much easier if I had a second milling machine - makes me miss my old Lagun and Sharp mills.

So i ordered the conversion from Heavy Metal CNC oddly enough.

I liked some of the features, and I think the whole world must need something higher profit margin than these G0704 kits right now - HM was the only one to really have time to get me a kit.

SO. quick updates.

I made my first good parts this weekend. My commanded size on this part I made was .370...actual part was .37025. I'll live with that.

I have $3k total into this conversion at the end of the day it appears.

Lessons learned:

1. Get the azvideo kit. Its ballscrews are better than the heavy metal CNC ones. My backlashes are .0035, .004, and .004 respectively on XYZ . I hear the AZvideo kit will beat that a good bit.

2. Set backlash compensation to 150% of your base acceleration in UCCNC. If you don't, it will cause a thousandth or two of spurious movement when you move the axis as it tries to force out the backlash.

3. The CNC4PC breakout board with the 300ETH controller built in is REALLY fantastic. When you add up all the stuff it does, it just does a great job of controlling cost and reducing your work efforts.

4. Hybrid stepper motors were very good. I got mine from RATTM motor. You set the steps per revolution pretty high and the motors quiet down immensely. My machine runs almost quietly other than the spindle.

5. With the gibs locked, my spindle has .002 of motion under tension. .004 with the gibs adjusted pretty tight. This was one of the hardest things about this setup.

6. IGUS energy chain stuff off ebay = full of win.

7. Noga mist coolant very good. I used a 120v relay for the air and have it hooked to the contactor for the spindle. So spindle on = coolant on. I can turn the coolant off at the noga unit if needed. You can just buy more
tubing that is the same size as the Noga tubing to extend the lines - no need to buy their expensive kit. I run the coolant with a Kobalt "ultra quiet" compressor. 70DB of sound when it runs, its as quiet as the spindle.

The mist coolant does not fog up my shop or anything - you don't need a lot of if, I think the fogging issues come from people who try to treat it as flood coolant. Even though the coolant is water based I have no rusting issues or anything. It must have a non-corrosive additive to it. I filled it with distilled water as well, which should help.

8. The ebay TTS style tool holders were very very good in runout and execution. That said, the ER25 collets I bought from the same vendor sucked horribly. Every one ran out .003. So consistently that it had to be the setup of the surface grinder they used to make them. If they fixed that thing it could be a .0005 runout collet. I used the SYIC/Techniks ones after that and got the runout total down to .0002.

9. I bought a cheap 4" ebay vise. The one you can flip on its side too. This vise was quite well made, and the fixed non-moving jaw held very tight tolerances and the parallelism of everything was very very good. Super impressed for $80.

10. Instead of a touch probe I bought a TTS CScope. This is a high quality unit that is well designed for what we are trying to do here. It was INVALUABLE for tuning the machine as I can physically see the backlashes and such while jogging the machine and tune them out. Made in Canada at that.

Next steps are belt drive and bigger spindle motor. I could use more torque for drilling operations. And, as always, learning more about Fusion 360, lol.

Attachment 417534
Attachment 417536
Attachment 417538

I've got similar C7 DHU1605 ballscrews to the AZvid ones in the machine I'm building.
I split the nuts and added tin foil between the sleeve and spacer (0.03mm per piece) and used AC bearings.
Got them a little tight but hopefully be o.k.
Will see what lash it gives me in a day or two.

Are yours double or single nuts?

Single nuts unfortunately. The other main issue I have is that they don't have a wiper on them to wipe any chips away so you have to be super anal about guarding them.

Quote:

---

Originally Posted by vectorsc

Single nuts unfortunately. The other main issue I have is that they don't have a wiper on them to wipe any chips away so you have to be super anal about guarding them.

---

Couldn't be doing with having do be so anal :(

Get it to make some money before you put more back in it if you can.

I just managed to get a test done on my Y axis and got a 0.025mm baclash reading (0.00098") I think that'll do.
It's pretty tight where bearings/nuts are concerned so It'll either go well or eat itself!!
Been let down on material atm so I can't start on my X yet :argue: :rolleyes: