[691175002]

Are you planning to cut drain holes on each side (and maybe use standard sink drains)

I was planning on epoxying this hose attachment to the bottom of the pan. I don't like drains that have raised edges since it will make the coolant pool up.

Not sure if I need one or two drains. Coolant will be able to flow under the mill since there is a tiny gap around the sheet metal so I'll probably start with one.


Started working on the angled pieces. I think it looks much nicer, and the second bend makes it stronger as well.


I've been practising my welding and I've gotten a lot better at avoiding warping. As you suggested, it seems that waiting so that everything stays cool is very important.

[doorknob]

You certainly work quickly - in the time that it takes me to think about doing something, you actually get it done. Wish that I could be that industrious...

[manton]

If the pan is made of steel, and the part you want to attach is brass, instead of using epoxy to attach it, you could solder it on using a torch. This would be much stronger, since epoxy doesn't really bond to anything that well (it requires mechanical keying to stay attached). You shouldn't have any warping problems with this, since the temperatures are much lower.

[DRock]

Are you using a MIG for all this work?

[jid2]

Dont leave the legs of the stand just poking down unsupported, box them together at the base at least.

[691175002]

You could solder it on using a torch. This would be much stronger, since epoxy doesn't really bond to anything that well (it requires mechanical keying to stay attached).

Good idea, I've never soldered something that large but seems like a good time to learn.

Are you using a MIG for all this work?

Yeah, I picked up a lincoln 140 because they had a rebate for a free helmet.

Here is the new angled sheet metal:


Cut a plate for the front out of 16ga:


Bolted on:


Unfortunately I wasn't able to make everything line up perfectly but I guess its close enough.


Put some screws through the back:



Threw the table on for some photos:




It will actually look more like this but I didn't feel like mounting the sheet metal properly:


You may have noticed that the bellows are just sitting on top of the sheet metal instead of being attached to a dovetail. You can lift the bellows and take a look to see if chips are building up or shove a shop vac in there.

These bellows seem to really like staying straight so I expect them to stay where they should be in use. Only one way to find out.

[691175002]

Dont leave the legs of the stand just poking down unsupported, box them together at the base at least.

That was in the design but the metal place only sold 24' lengths.

I agree that it looks a little precarious so I'll try to find some 1" tube to add some reinforcement.

[bru102]

I'm intrigued with Mad Welder's process for sheet metal welding. The intent seems to be to localize and minimize heating of the metal to reduce the shrinkage in the weld zone.

I'm not sure I got it right... Seems like the idea is short (1") welds, spaced about 6", and in sequence either from the left or right end of the seam. I think you would then start over and weld the next inch of each weld, and repeat that sequence until the weld is continuous.

It seemed like there was some importance to the direction of the weld, and I don't get that. I'd love a clarification on that.

Ray B.

[Mad Welder]

I'm intrigued with Mad Welder's process for sheet metal welding. The intent seems to be to localize and minimize heating of the metal to reduce the shrinkage in the weld zone.

I'm not sure I got it right... Seems like the idea is short (1") welds, spaced about 6", and in sequence either from the left or right end of the seam. I think you would then start over and weld the next inch of each weld, and repeat that sequence until the weld is continuous.

It seemed like there was some importance to the direction of the weld, and I don't get that. I'd love a clarification on that.

Ray B.

Yippee some one is intrigued with a post I made......yeah Ray you're correct with the idea that you localize the heat, but really what you're doing is minimizing the heat input and hence the "warping" due to the weld cooling and causing "shrinkage" because when molten weld-metal cools it contracts within it's self and so as it contracts it pulls on the parent metal and this is extremely evident (as warping) with light gauge sheeting.......

and with the direction of weld travel my apologies here that I didn’t explain it more clearly but yes there is a huge importance to the direction of travel, so bear with me here and I hope I don't confuse you more either.....so lets say that we have a right handed welder and he has positioned himself directly in front of a horizontal seem (i.e. a horizontal weld seam is parallel to the floor, A vertical weld seam is perpendicular to the floor), and lets say that the horizontal seam is 3 foot in length.....so by measuring and marking the seam in one inch increments (I suggested the one inch increments in my previous post as the sheet gauge is quite thin, and therefore if the sheet was heavier the incremental measurements would be larger), so using our example here the welder would have 36 one inch divisions marked out on the seam, and as a typical right handed welder welds from left to right (i.e. his hand holding the electrode is moving "travelling" away from his body as he welds) therefore with this technique he would start welding the seam on the right hand end one inch from the end and as he is welding from left to right he would weld only one inch and then obviously he has reached the seam end..........so now his next one inch weld would be measured from the right end of the seam (i.e. where he has just completed his first weld) and so measuring six inches to the left from the first weld, he would weld another one inch weld pass from left to right (i.e. the direction of travel is still left to right or travelling towards the first one inch weld pass)......
and if a left handed welder was carrying out the same procedure he would start from the left hand side of the seam and his direction of travel would be from right to left........and yeah you're correct in the understanding that you would repeat the sequence until all the seam was welded, but in the case of light gauge sheeting you would also allow for cooling in-between passes as the residual heat would require time to dissipate……and never fast cool any weld (dunk it in water or throw a wet rag on the welded seam/structure) as you alter the grain structure within a weld and the heat affected zone causing a Martensite grain structure which is extremely brittle and susceptible to cracking…….

hopefully I haven’t confused you even more…….

[phomann]

Hi,

That's a really interesting post. Thanks for the information.

Cheers,

Peter

[Mad Welder]

you're more than welcome glad to be of some little help

[BobWarfield]

Great thread--thanks!

BW

[bru102]

Mad Welder... thanks for the thorough answer. I'm looking forward to trying out your technique on some 18 or 20 ga. steel. I've got a plasma cutter, but no bending brake... so continuous welding is probably my best shot at fabricating a pan like Ryan's.

Ray B.

[691175002]

Hi everyone, I had to put the project on hold since fall semester started up and I am studying away from home.

Anyways, I've got a month at home starting tomorrow and there are a few projects I would like to finish while I have access to CNC.

This essentially means that I hope to have my g0704 operating within two weeks from now.

This will mean rushing some stuff and putting in some temporary solutions. My plan is essentially to use the mill as-is (with a bit of work to mount limit switches and the bellows). I will need to get the spindle turning which is the hardest remaining task.

I hope to use a dmm-tech 1500W servo since I kind of want an encoder on the spindle.

[DRock]

I really like the idea of using a servo with encoder for the spindle...as always I will be watching this thread closely and I am happy to hear you will be updating it again soon!

I almost bumped this the other day to ask where you went haha, you have one of the best builds I have seen to date.

[691175002]

I've taken a day or so to reorganize and refamiliarize myself with all my stuff.

Unfortunately, Christmas is a bad time to be working on a project. Its cold outside and everything is closed so it will be hard to find material.

My basement shop is starting to get pretty crowded. I picked up a pair of glacern vices as an early present.


Rest of the mill is on the floor.


As I mentioned earlier, my goal is to have the machine cutting ASAP. The main tasks that must be completed are:

• Get the spindle running
• Finish welding the stand/tray and add reinforcement
• Finish mounting bellows/sheet metal
• Get all the wiring done
• Lap the 5th axis rotary table
• Get the electronics working

I've decided to take a stab at the electronics first. As mentioned way back, I plan to use the Dynomotion KFlop for this mill. It has some pretty nifty features (46 IO pins, 8 axis channels, 8 encoders) and is extremely customizable. It is fairly complicated to set up though and pretty much requires a working knowledge of C to do anything useful since it all has to be coded in.


Since I plan on adding a lot of features over time and want to be able to make changes easily, the electronics pose a bit of a problem. I've decided to try to make it as modular as possible by doing everything in simple sections that connect via cables so I can swap pieces in and out. Hopefully this will also keep things neat as well.

I made three breakout boards for the KFlop (Blue, Red and White). The blue board does the heavy lifting and handles 4 servos as well as three encoders (The linear encoders plug into the tan board) so it can pretty much manage a full machine on its own. It also has a serial port for communication with a microcontroller down the road if more IO is required.


The white board will provide 4 more step/dir channels for a total of 8 servos/steppers (5 axes, spindle and two more axes for ATC). Finally, the red board handles all the general purpose IO such as limit switches, coolant, MPG and so on.

Above is just a really quick show of how everything will be laid out. The white/red boards are still in progress but blue/tan are finished.


The top of the boards look very neat because the mess is underneath. I have tested all four step/dir(enable/fault) connections as well as all three linear encoder channels.

Everything works on the bench so hopefully there are no surprises.

[Mad Welder]

As always I'm fascinated with this build, and your pics and descriptions are fab......I too am looking into the Kflop boards......but alas I'd have to learn from scratch C programming. ..is there a link that you got the wiring schematics for making your boards or if you designed them yourself would you be able to post a pic/detailed of the wiring configuration? ?

[691175002]

I don't have much information on the wiring. I pretty much just put the connectors were they would be visually appealing and then added one wire at a time until I was finished. Most of the connections are just one pin to another pin, for example the step/dir lines from the servo drive need to connect to the step/dir pins on the kflop. Repeat 20 times and the board is done.

The linear encoders came with schematics for connecting to a TTL line that involves a couple resistors so that was all included on the tan board.

[691175002]

Similarly to how the breakout boards are modular, I also plan on separating each section of the electronics. I have no idea what size or shape of box I will need for the electronics, but I can't just leave everything on a table, so I'll go halfway and screw everything onto acrylic sheets. (And eventually mount the acrylic sheets in a box together.)

I started by measuring the hole patterns of all the components I need to mount, and then did a test cut.


Everything lined up pretty nicely.


Also tried putting some screws through.


The first "section" of electronics that I decided to make was the power supply. It has 3.3, 5, 12, and 48V as well as a pair of SSRs and a smaller 12V supply to power them. Its a lot of stuff but it means I can power pretty much anything.

Test cut with paper...

And the acrylic version.


Hopefully it still looks nice after wiring. I think I might stack the KFlop and servo drives on top of this. If I lay everything out flat it will be massive.

[Mad Welder]

....... Most of the connections are just one pin to another pin, for example the step/dir lines from the servo drive need to connect to the step/dir pins on the kflop. Repeat 20 times and the board is done........

Now why didn't I think of that before asking. .....:stickpoke