Other What did you fab up today? A thread for the home fabricators!

[beerdart]

 

[OCD Solutions]

We are building a new dry Kiln on this site as part of the upgrades and they have been letting me take some of the aluminum used for shipping braces.

It’s heavy stuff and in random lengths so I have no idea what to do with it, I just know it’s free.

This is the second batch I’ve gotten so far and there’s plenty more I will get once they finish up all the structural components.

2” angle, 3/8” thick and random lengths from 24” up to 48” long.

 

[OCD Solutions]

Brainstorming with ScottS and we came up with a perfect use for all this 3/8 material.

Scissor stands for the new SXR's.

 

[Pro-pulsion]

Brainstorming with ScottS and we came up with a perfect use for all this 3/8 material.

Scissor stands for the new SXR's.

Might want 1/2" steel for that

 

[Vumad]

Upping my mounting game today. Using tabs and made my first set of soft jaws. Found 2 nice matching pieces of bar in the scrap pile so my soft jaws are about $3 (not including machine time, but that’s all practice anyway. First time using the drill operation.)

Sent from my iPhone using Tapatalk

 

[Harbor]

Upping my mounting game today. Using tabs and made my first set of soft jaws. Found 2 nice matching pieces of bar in the scrap pile so my soft jaws are about $3 (not including machine time, but that’s all practice anyway. First time using the drill operation.)

Sent from my iPhone using Tapatalk

looks good... but what are the discs your holding again ?

 

[Vumad]

looks good... but what are the discs your holding again ?

Custom challenge coins for firefighters. Posts 861 and 889.

 

[Vumad]

Well, as you can see by the date on the coin (4/6/21), I’m out of time. It was my goal to give Scott the first coin I ever made and I gave it my all.

I tried really really hard to get tighter results. I spent a lot of time trying to get it dialed in but i haven’t been able to yet. It’s accurate to 1mm (0.040) but I expect tighter than that.

Sent from my iPhone using Tapatalk

 

[OCD Solutions]

It looks way better than the 3D printed version and I am sure he will still be very proud to receive it.

Where do you think your inaccuracy is coming from?
How often do you check and adjust your step settings?
What kind of backlash do you have and is it in the ways or the ball screws?

I recently replaced all 3 of my ball screws with double ball nut styles that hold much higher tolerance. It wasn't my main reason for upgrading but it was a nice bonus.

 

[Vumad]

It looks way better than the 3D printed version and I am sure he will still be very proud to receive it.

Where do you think your inaccuracy is coming from?
How often do you check and adjust your step settings?
What kind of backlash do you have and is it in the ways or the ball screws?

I recently replaced all 3 of my ball screws with double ball nut styles that hold much higher tolerance. It wasn't my main reason for upgrading but it was a nice bonus.

I'm not sure where it is coming from. I think the gauge I am using might be wrong. I was setting up my gauge and manually adjusting my steps per until the accuracy was down to .005 or less. I had done this before, after I got the machine, and my coin was coming out to 46mm on the X and 46.9mm on the Y.

Yesterday I spent a while trying to dial it in closer. The mach3 default units is in inches. I setup the gauge using the mighty magnet on the vice and the tip up against the spindle. I would g0 y0.05 then g0 y0. Every time I did this, I landed exactly back on 0 (no backlash) but I was off by roughly .010 on the move (short, 0.040 on the gauge). I adjusted the steps per until the move was exactly what the gauge said it was (0 to 0.50 every time, <0.005"). However, when I machine the circle, it was clearly much larger on the Y than it was before, so my adjustments were making it worse than the last time I did it despite being consistent in gauge readings. (I trust the caliper more than the gauge because I can visually see the 46mm part is out of round).

I don't think the issue is in the backlash since I always land on 0. I was thinking it is the gauge since I can set the steps to get consistent results where the move entry matches the gauge display but not be correct on the part. Except the inaccuracy on the Y is worse than the inaccuracy on the X using the same gauge. (Y is my short axis)

I also considered it could be my CAM settings. Maybe I moving too aggressively and getting tool deflection, but additional passes don't change the results.

I don't know for sure but I think ball screws. My steps per on all axis is 8300 presently and my backlash settings are .001 on 2 axis and .002 on the 3rd. My default feedrate on all 3 axis is 100.

 

[OCD Solutions]

I would focus on your steps and start doing tests over larger moves. It'll exaggerate your error and make it more obvious.
I use 1" moves for course adjustment and then use 3-4" on Y and up to 10" on X to really dial things in.
If your step counts are off, you will always return to zero as it's using the same scale to move in each direction.

 

[sturat]

no fabbing but I did get a TON of paintwork done this weekend, and what a relief. Had me tripping out with the weather and just life... but we got er done and it looks pretty damn good given the time constraints. PPG concept deltron pure white and nason dodge hydro blue pearl metallic cut heavy with PPG concept gloss clear. Ready to see them back on the water.

 

[Vumad]

View attachment 408989

View attachment 408990

no fabbing but I did get a TON of paintwork done this weekend, and what a relief. Had me tripping out with the weather and just life... but we got er done and it looks pretty damn good given the time constraints. PPG concept deltron pure white and nason dodge hydro blue pearl metallic cut heavy with PPG concept gloss clear. Ready to see them back on the water.

They look good but I’m waiting patiently for the X2 post.


Sent from my iPhone using Tapatalk

 

[sturat]

They look good but I’m waiting patiently for the X2 post.


Sent from my iPhone using Tapatalk

long gone, did the math, wasn't worth the investment, zero profit possible on a return

 

[Vumad]

@OCD Solutions

The enclosure I built has worked great for setup and practice but I think I am ready to start designing my permanent CNC enclosure. I think I would prefer a flood coolant system over a mist system. Have any recommendations on that? I'd also like to add spindle control to the Mach3 and a quick tool changer.

To put those into a priority list...

1) Enclosure with flood coolant
2) Spindle control Mach3
3) Quick tool changer

2 and 3 could be pretty time consuming and a distraction so I would probably delay those but a permanent enclosure with flood coolant seems like it could be a big and permanent solution to upgrading quality and work times. The machine did come with this doodad that hooks to the compressor and blows cold air so I will be setting that up soon. Right now I am using my ryobi bench blower by hand, lol.

 

[OCD Solutions]

I destroyed the OCD Garage this morning to build two simple camera brackets.

My mill is still in pieces so I had to do everything the hard way.

 

[X2Kid]

I imagine half the fun of being OCD is putting everything back in its place. Enjoy!

 

[OCD Solutions]

You are not far off. Plus, if everything has a place, it’s pretty quick and easy to square things away.

My OCD is calmed once again.

 

[Vumad]

Sent from my iPhone using Tapatalk

 

[te4250]

I'm not sure where it is coming from. I think the gauge I am using might be wrong. I was setting up my gauge and manually adjusting my steps per until the accuracy was down to .005 or less. I had done this before, after I got the machine, and my coin was coming out to 46mm on the X and 46.9mm on the Y.

Yesterday I spent a while trying to dial it in closer. The mach3 default units is in inches. I setup the gauge using the mighty magnet on the vice and the tip up against the spindle. I would g0 y0.05 then g0 y0. Every time I did this, I landed exactly back on 0 (no backlash) but I was off by roughly .010 on the move (short, 0.040 on the gauge). I adjusted the steps per until the move was exactly what the gauge said it was (0 to 0.50 every time, <0.005"). However, when I machine the circle, it was clearly much larger on the Y than it was before, so my adjustments were making it worse than the last time I did it despite being consistent in gauge readings. (I trust the caliper more than the gauge because I can visually see the 46mm part is out of round).

I don't think the issue is in the backlash since I always land on 0. I was thinking it is the gauge since I can set the steps to get consistent results where the move entry matches the gauge display but not be correct on the part. Except the inaccuracy on the Y is worse than the inaccuracy on the X using the same gauge. (Y is my short axis)

I also considered it could be my CAM settings. Maybe I moving too aggressively and getting tool deflection, but additional passes don't change the results.

I don't know for sure but I think ball screws. My steps per on all axis is 8300 presently and my backlash settings are .001 on 2 axis and .002 on the 3rd. My default feedrate on all 3 axis is 100.

Always landing on 0 moving the same direction does not indicate 0 backlash, set your steps/inch in mach3 using as close to the full travel of the machine as you can to minimize error. Setting steps/In over .05” and adjusting steps to compensate for backlash is for sure what’s throwing you off, .9mm of tool deflection would be a broken tool unless it’s a 6” long 1/4” end mill. Once you have steps/in close you can set backlash by moving against the lash one direction, and stepping until you just move your indicator the other direction, or by the method you already stated and using difference in commanded and actual position