Hoss's G0704

[skyguynca]

Ok, I did Hoss's cnc conversion 2 years ago. I have had no problem with it except wearing out gibbs. So I opted to do the linear rail conversion using my old column but to keep using my machine I purchased a new column and installed it so I could use my machine while I work thru the linear rail installation on the old column that has the worn out ways.

Well I took the whole machine apart and cleaned it, installed new gibbs with the re-installation of the saddle and the table.

In going thru the setup for movenment and backlash, as I have done every time I go thru gibb replacement, I noticed that my table was .002 Higher on one end.

I took it apart and checked to make sure under the saddle was clean and not causing the table to be at a very slight slant, but all was good.
I checked the ways on the table, all is good.

Now I know the column is within .0005" at the full travel of 14 inches true 90 degrees to the base.
the saddle is true to the base within .0005 at full travel of 9.7 inches. (yes I have both of Hoss's extension mods installed)

So with both so accurate how is it my table from one end the test indicator does not move more than .0005 over 8 inches, then climb .002 linearly over the next 8 inches?

I even tried reinstalling the tooling plate and faced it, the mirror finish on the aluminum tooling plate shows the head is correctly trammed, yet when I attached a test indicator to the head
and then run the table full length and still shows the first 8 inches is withing .0005 and then it starts a steady climb of .0005 every 1.33 inches of travel until it max out at the end of .002 rise in height.

Please throw all suggestions at me so I can try and fix this.



David M.
San Jose, CA

[skyguynca]

It is fixed now, I took it apart again, I worked the table back and forth and found binding in the last 8 inches of travel. It was not found during reassembly of the mill. That is because the gibbs were not fully seated. The final adjustments made to stop any play in the table, saddle or column had not been made yet. When they were made everything was moved with the steppers and the gibbs were adjusted to take all slop out.

Well it ended up being the gibbs I received from Grizzly. I put the new gibbs in with the steppers disconnected and slid the table end to end and after a adjustment I could only move the table half way before I could feel the binding starting and getting worse towards the end of the table. It would get so tight I would need a rubber mallet to move the table back past the half way point so it would be loose enough to move by hand.

The Y and Z gibbs are OK, but the X gibb is slightly greater in taper from thin to thick and in height.

I put the old X gibb back in and everything is fine, back to normal.

David M.
San Jose, CA

[MiniSoCalCNC]

I got an email from him Aug 4th.

Concerned ... somethings up - His domain expired 12/6/17 ...

Anyone heard from Hoss ?

[cd_edwards]

I emailed him beginning of dec for replacement files for his drawings and videos. I got a response immediately

[ger21]

Concerned ... somethings up - His domain expired 12/6/17 ...

This site is still up.
g0704.com

[hoss2006]

Concerned ... somethings up - His domain expired 12/6/17 ...

Anyone heard from Hoss ?

He's fine, busy gettin healthy. Working on combining the websites for convenience is all.
Hossmachine Store
Hoss

[MiniSoCalCNC]

He's fine, busy gettin healthy. Working on combining the websites for convenience is all.
Hossmachine Store
Hoss

Excellent news ... gave a brother a bit of a fright :-) Glad all is well - Merry Christmas / Happy Holidays my good man.

Thanks for the update ...

while we're here :-) got any links on any of your sites or others indicating how I can finish wiring this C11G BoB to a Solid State relay to control coolant 110v power plug? Ive seeming looked everywhere and was headed over to Hossmachine for some possible links and discovered it was - down ...

[Fastest1]

Excellent news ... gave a brother a bit of a fright :-) Glad all is well - Merry Christmas / Happy Holidays my good man.

Thanks for the update ...

while we're here :-) got any links on any of your sites or others indicating how I can finish wiring this C11G BoB to a Solid State relay to control coolant 110v power plug? Ive seeming looked everywhere and was headed over to Hossmachine for some possible links and discovered it was - down ...

Do you have the load connected to the relay?
Many SSR’s wont light up unless the load is connected. They use parasitic voltage from the wiring.


Sent from my iPhone using Tapatalk Pro

[kies277]

Can anyone tell me what the screw size for the Y-axis mount on the G0704 is? I need to buy some longer screws for my stepper mount and I thought I had the right ones, but I do not.

[cd_edwards]

Can anyone tell me what the screw size for the Y-axis mount on the G0704 is? I need to buy some longer screws for my stepper mount and I thought I had the right ones, but I do not.

if 1/4-20 doesn't work, try 8mm if I'm not mistaken. you can also re-tap to 1/4-20. this is what I did on my KC20-vs from King Industrial in Canada. made it easier than attempting to figure out the metric bolts on my mill which is a clone of the grizzly, but metric.

[CFLBob]

if 1/4-20 doesn't work, try 8mm if I'm not mistaken. you can also re-tap to 1/4-20. this is what I did on my KC20-vs from King Industrial in Canada. made it easier than attempting to figure out the metric bolts on my mill which is a clone of the grizzly, but metric.

I was reluctant to answer this because I didn't know which screws to talk about. The 704 itself is metric, tapped 6.0 x 1.0 mm, but that's not a motor mount, that's the extension that leadscrew goes through. The phase 3 motor mounts that are made from angle aluminum are tapped 8-32 on one end and 10-32 on the other. But the phase one and phase two mounts that are tapped, 1/2" aluminum bar are 10-32.

So the answer to what screw size is "which screw?"

[kies277]

I found it, they are M6. Thank you all for the help. Anyone know why my stepper motors are running dreadfully slow? I just got them hooked up through Mach 4 with an ESS board. I have a NEMA 34 and 2 NEMA 23's. I have the controllers set to 1600 pulse.

[diyengineer]

He's fine, busy gettin healthy. Working on combining the websites for convenience is all.
Hossmachine Store
Hoss

Hoss, tried to shoot you an email and it got returned. Hope all is well.

Sent from my SM-N910V using Tapatalk

[hoss2006]

hoss@g0704.com is back up and running now.
Hoss

[diyengineer]

hoss@g0704.com is back up and running now.
Hoss

Good to hear from ya! I'll get around to shooting you an email this week! Hope your making some chips!

Sent from my SM-N910V using Tapatalk

[kies277]

I am trying to get my Z axis dialed in and am having a problem. I had it at 1600 pulses/rev and got it moving the correct distance, but the problem is when i make small adjustments it moves .015 instead of .010 (and .007 instead of .005). So I can move it 6 inches up and it will be right on, but if I am at the work piece doing thousandths of inches, it is going 1.5X the amount I want. I tried on other pulses/rev settings also, from 800 up to 8000 and they all do 1.5X. I am using a smooth stepper and the KL8070D with a Nema 34 and the Ebay Z axis double ball screw. The X and Y work fine at 1600.

[ken226]

I am trying to get my Z axis dialed in and am having a problem. I had it at 1600 pulses/rev and got it moving the correct distance, but the problem is when i make small adjustments it moves .015 instead of .010 (and .007 instead of .005). So I can move it 6 inches up and it will be right on, but if I am at the work piece doing thousandths of inches, it is going 1.5X the amount I want. I tried on other pulses/rev settings also, from 800 up to 8000 and they all do 1.5X. I am using a smooth stepper and the KL8070D with a Nema 34 and the Ebay Z axis double ball screw. The X and Y work fine at 1600.

I'm not sure where your getting your steps per inch or steps per revolution information, or what specific motors controllers you have.

What you need to know is, first your stepper motors number of steps per revolution. Example, my nema 34 960oz motor has 200 steps per revolution.

2nd, you need to know how many microsteps your stepper driver is set to deliver. Example, my geckodrive g201x is set for 10 microsteps.

3rd, you need to know the pitch of your ballscrew. Example, mine is a 5mm pitch.

Now, for my setup. My mach3 native units are set for "inches". A little math: 25.4mm per inch divided by 5mm ballscrew pitch gives a ballscrew travel of 5.08 revolutions per inch.

So, 200 stepper motor steps per revolution, multiplied times 10 stepper driver microsteps, multiplied times 5.08 ballscrew revolutions per inch gives me a mach3 "steps per" (that's steps per inch when Mach 3 native units is set for inches) setting of 10160 (Mach 3 will interpret a microstep as a step).

10160 microsteps from the stepper driver = 1 inch of travel.

You said you have a 6mm pitch ballscrew. If I assume your native units in mach 3 is set in inches,and your stepper driver also set for 10 microsteps, and you have 200 steps per rev stepper motors, then you would need your mach3 motor tuning "steps per" set to 8166.666666667 "steps per". I doubt a stepper driver can deliver a fractional .6666666667 number of steps. I'd go for a whole number.

If you set your stepper driver to 12 microsteps, then you'd end up with a nice even whole number "steps per" setting of
(25.4 / 6) x (12 x 200) = 10160 "steps per"


My x and y axis stepper drivers are nema23 540oz motors, also 200 steps per revolution, and my x and y ballscrews are also 5mm pitch screws. My x and y stepper drivers are 8 microsteps. So, my x and y are 25.4 / 5 x 200 x 8 = 8128 "steps per"

It sounds like your "native units" may be set for mm, in which case, your 6mm screw and a 10 microstep driver and a 200 step stepper motor would need a "steps per" setting of 160 steps per mm.



I hope this helped.

[skrubol]

Do you have backlash compensation turned on?

[CFLBob]

I am trying to get my Z axis dialed in and am having a problem. I had it at 1600 pulses/rev and got it moving the correct distance, but the problem is when i make small adjustments it moves .015 instead of .010 (and .007 instead of .005). So I can move it 6 inches up and it will be right on, but if I am at the work piece doing thousandths of inches, it is going 1.5X the amount I want. I tried on other pulses/rev settings also, from 800 up to 8000 and they all do 1.5X. I am using a smooth stepper and the KL8070D with a Nema 34 and the Ebay Z axis double ball screw. The X and Y work fine at 1600.

First off, what Ken226 is saying is exactly right. I use the KL6050 drives that Hoss had in the parts list I bought, but the motors are the most common 200 steps per revolution, my drive is set for 8 microsteps, and the same ballscrews. it works out to 8128 steps per inch. Being a couple of steps different from the calculated value is normal, I think, and you verify that with a dial indicator moving a full inch. That's from small errors in the ballscrew that the Quality Control guys in the business call ballscrew drunkeness. If I recall correctly, all three of my screws are slightly different. What that means, though, is that if you set your table to go exactly one inch in the middle of your table, you might not go exactly 1.000 inch on any other place on your table.

When I saw this last night, I wondered how you knew it was going 6.000 inches instead of a few thousandths too much. It's not exactly easy to get micrometer accuracy over 6" (without a 6" micrometer). What your post is implying is that your system puts out too many step pulses if you go a short distance but it puts out the right number if you go farther. It really just multiplies steps per inch times the inches you're going and spits them out.

When I built my Sherline CNC system, it did some odd things, but just extra steps no matter what distance I was moving it, and just moving a couple of steps on its own. That was electrical noise pickup, but what you're talking about sounds different than that.

Sorry I have nothing more to suggest.

[kies277]

I am using a dial test indicator to measure. I start it at the table and move it up a known distance(a measured object such as a 1x2x3 block or parallel bar). The backlash is not relevant in this case, I am only moving it one direction. Probably does not help that automation technologies did not send one stitch of paperwork for the motors or stepper drivers.