JDs BP J-Head CNC Conversion Blog

[lerman]

My understanding is that the speed is proportional to the voltage. The torque is proportional to the current.

So, running at a lower voltage will reduce the maximum speed, but should NOT affect the maximum torque (within some limits). Obviously, if the voltage is too low, it cannot drive enough current into the motor, but that is not usually the case.

Ken

[jdelaney44]

So here are some drawings of how I've envsisioned the X-Drive on the Bridgeport. These are exploded.

I did pencil and paper for the Quill. But after making several errors I think I could have avoided, I decided to invest the time with the CAD. I think in the long run it'll be faster.

You'll see a four bolt pattern on the inner and outer plates. Thisis for a couple of bearing blocks. There will be a 5/8 in. shaft that goes through those to a coupling that will join that to the ball screw. When I replaced the ball screw I got the one to go with the existing Bridgeport power feed. I did not want to disable the ability to put the machine back into the condition it was when I got it. Again, incase the CNC craps out I want to be able to use the machine to make repairs if needed.

The drawings don't look great in the browser. You might want to download and open with something else. My Turbo CAD's save as to JPG is less than great. I'll keep working on it.

Thanks,
-jd

[Hood]

JD hows your X and Y going????
I have got my Z completed and working and have actually made some money with it Will be a while before I get my costs covered but what a machine, still got limits to fit and cables to tidy but hope to be moving workshop soon so might wait until then to do the snagging. Also I am a very untidy person as you will see from the pics so I feel at home wth things being half finished
You can see my x y and z pics HERE

Hood

[jdelaney44]

Hood, I started the X today. Should have it done tomorrow. Going well. I'm trying to get the Y done by the fourth. I've got some personal stuff hitting the fan right now so that's slowed me down quite a bit. But, so far so good.

Thanks for the links. Looks good. Glad to hear it's working good too!

-jd

[Hood]

Look forward to pics of your finished X and Y. Today I hope to get my limits fitted so I will be happy to leave it alone and get on with other work while the Bridgie does its stuff.
Hood

[jdelaney44]

Here are the pics of the X-Drive. Just about ready to mount. Note the shaft. I have a screw designed for the Bridgeport power feed so I am making the x-Drive to fit that.

-jd

[jdelaney44]

Some more on the X-Drive construction.

(as always please pardon the spelling and typos)

The CAD exercise helped a great deal. I have previously just been slugging it our with pencil and paper. I use TurboCAD 11 (TC). Just got the upgrade in the mail yesterday. I've decided that CAD programs are a matter of personal choice. The 3D operations in TC make sense to me. They also have what, to me, is a pretty intuitive control for sizing and dimensioning objects. I haven't tried them yet, but the auto dimension features in 11 are supposed to be better. The dimension features in 10 allowed me to quickly set up the dimensions you see in the clips above. I need to reload those too. Just notices I managed to load the same drawings like four times. Nuts.

These drives are aluminum since I am not as concerned about the rigidity as I was on the z-drive. Actually I over killed on the z-drive. More on my observations there later.

So this time around I decided NOT to surface the stock like I have been doing. My rationale was a.) I see a lot of other people getting away without doing that, b.) the belt stretch should probably soak up and deviations created by the inconsistent plate thickness, c.) I can shim the drive and the motor to get perfect alignment if it does become an issue, and d.) I am using flanged pulleys so the belt coming off should not be a problem, and and e.) the material deflects when you surface it anyway so the true gain is really very questionable.

Still everything coming off the bands saw had to be squared up on the edges. I do this by picking the best looking side and putting it down in the vise. Then I mill the one edge sticking up with either a 0.5 or a 0.75 end mill. Then I flip it over and do the same. This time though I'm checking dimensions as I go. I currently over size the stock on the saw by about 0.25.

Once I have two sides parallel, I clamp the piece flat in the vice using a couple of steel parallels to hold it at the top of the vise. I got a set of these for $99.00. 0.5 thick by 6 inches long.

Once clamped I run the side of the end mill along it in the Y direction. One deep cut to get it close and and then one or two finishing passes to get it pretty. I have taken to running the finish cuts at about 0.002, dry no coolant, with the rpm at about 1500. Now turn it over and do the same this time working to the final dimension.

This go round included a run to Harbor Freight to get a set of 12 inch calipers.

Six finished parts one paperweight. I'm getting better! - This time the DRO was being squirrelly. I had to recalibrate it. Unfortunately I did not discover this until after I made the motor side plate and checked the dimensions. Everything grew by about .002 / inch. In the 12.75 inches of the plate everything looked perfectly proportional and accurately spaced. But it was WAY off by the time you got to the end of the plate.

I debated for a long time about not fixing the DRO for the balance of this build. It would be wrong, but it would all be the same wrong, right?

Well, then I figured if I ever want to make spare parts or retrofit something in there I'll be fighting these odd dimensions for the rest of my life. Forget that. So I remade the motor side plate.

The slots for the motor mounts screws are 1/4 inch. They are 2 inches center to center of the end radiuses. I did this to give me a lot of latitude in belt sizes. You never know when things break if you can get what you want, when you want it. This Just In Time world we live in doesn't always work out so good!

To cut those I plunged, slowly, the 4 flute HSS end mill all the way through the 0.5 at one end and ran it to the other end with the power feed at about .75 ipm. Real slow. I actually adjusted it by ear. When I heard a smooth cut I kept that speed. Oh, The end mill was in the collet right up to the top of the flutes. I got way too much chatter with it sticking out any more. The slots came out OK. They are smooth and the ends don't have too much of a deviation from what I will call recoil. Coolant was used, yes.

The slot for the motor shaft was made by running the same .25 end mill around the perimeter. I hand calc’ed the end points and used the DRO to tell where I was.

Note that I zeroed the DRO using an LED edge finder. It was zeroed at the upper left corner of the part, same as my drawing dimensions so I didn't get confused half way into the operation.

Note on using the edge finder. When I got it, the directions said not to have the machine turning when you used it. That made NO sense to me. I think they must mean high speeds.

I know I have about .0025 of runout in the spindle. So what I do is turn the spindle at like 10 rpms (using the VFD to slow it down). Then I crank the part over until I just get a little blink out of the LED as it grazes the edge. The tip of the finder is .20. So then I know I am .101 - .1015 or so away from the edge. This offset is then set in the DRO for X, repeat the process for Y. You now have a pretty good home setting in the DRO. I say pretty good ‘cause nothin’s perfect ever.

Hmmm……To make this better, I need to really put edge finder in the collet and mark the position somehow so I can get it in the same spot every time. Re-check the total runout periodically. Then use the same collet, the same edge finder, all lined up the same way every time. And make sure the R8 tapers are spotless of course. If I get a dog hair in there I’m off by .002.

So lets see. Do all that, then put in a three jaw drill chuck with a stock #10 drill bit in it and pray for +/- .005 as that wobbles around. Hmmmmm. Such is life.

See posting this stuff helps work things out!

The big deal is turning the spindle slowly and looking for the LED flash. That gets me a lot closer to compensating for the runout.

The back side was easier as it just has holes, no slots.

Then came the drilling of all those screw holes. Decided to use 10-32 x .75 screws. I thought that was going to take forever. It took about 45 minutes per big side and 15 minutes for the small sides. A #10 drill was used for the front and back plates where the screws would countersink flush.

I settled on drilling a line of holes, and then going back and counter sinking that line. At first I thought I might get lousy countersinks. But the machine and DRO seem to be holding +/- .002 or better so that's no problem for a drilling operation at all. So I drilled on side of holes, change to the counter sink bit, and went back to where I started that line, then went 90 degrees along the next side etc... I did get away with not locking the table as these were small holes with little chance of enough force being developed to move the table while drilling. This was done with an regular three jaw 0.5 in drill chuck on an R8 arbor. Cheap Chinese one. I need to get a Jacobs, they work so much better I can't even describe it. Actually I have the chuck for the late, just need the R8 arbor for it.

So now for the short sides of the box. Ran out of 0.5 plate, so I switched to some 0.375. The holes are slightly offset to make the outside smooth given the original design called for 0.5 through out. No big deal. Drill the holes as before just with a #21 drill this time for the 10-32 tap.

Tapping was accomplished by chucking the tap into my 1/2 VSR Milwaukee drill. Nice drill, lots of torque. Very carefully used the drill motor to drive the tap in. Reversing when the resistance got too tight. About 1/2 a turn to relieve the chips. Took some time, but not as long as it would have hand tapping them all. This also required that the tap be allowed to find it's way square in the first couple of threads. Which worked pretty well as long as I eye balled the drill to be square to the part. Some day I'll be able to afford a Tap Matic. Still, no busted taps this time. Yay!

Now everything is square, slotted, drilled, countersunk, and tapped. Assembly time! Much to my shock and awe, it all fits and ALL the screws go in easily. (The sun shines through the clouds as the organ plays and the choir sings for a miracle has truly just occurred! ) (Ok I just HAD to use the banana icon, sorry)

The only problem I found was that my pulley is hitting the bearing block on the inside, so just put the block on the outside and problem solved. I may still see if I can work that out somehow.

Onto the Y-Drive CAD drawings............

[Jamf2]

Congratulations.

Thanks for documenting what your doing. It has been very helpful.

[jdelaney44]

You are very welcome, thanks for the kind words.

[Hood]

Lookin good JD.
What software and drives etc are you using? Sorry if youve already said but I am too tired to read through the whole thread again, either that or just too lazy :-)
Ive been working on mine today, got all the limits working and also have a speed sensor tested, just got to make brackets for it and cut a slot out a pulley.
Hood

[jdelaney44]

Hood, well, I'm probably going to go with the flow and use Mach 2. I've got the Sound Logic break out board and I've got the Gecko G340 servo drives. Turbo CNC still looks interesting as do a couple of other DOS packages. But Mach 2 has grown to become almost a defacto standard in the DIY community. Almost.

Eventually I would like to revisit EMC which is Linux based and has the capacity to do true feed back to the software.

With the Geckos and Mach2 I get basically an error trip when and if the motor gets 128 steps behind. Then the Gecko faults. With the error signal I can trip a general e-stop in Mach 2. With either software you have to stop and re home and start the program over. I don't think EMC has the ability to dynamically adjust feed rates based on the ability of the servo to keep up. But that would be cool. Art Fenerty has talked about doing this with Mach, but I don't think he has yet.

Thanks to one and all for the positive remarks and replies.

Best,
-jd

[jdelaney44]

These are hopefully a little better than the last ones. I got the TurboCAD 11 upgrade and the JPG generator is supposed to be better. They still look bad on the browser. But they looked OK in a photo editor.

[Hood]

Hows it going JD got the Y started yet? Would love to see some pics of your completed setup.
I am amazed how easy and quick CNC is, obviously I have a lot to learn and I think I will always be learning, but I made a lid for a seacock for a boat the last day and had to cut a 1inch hex on the top for a spanner. Previously I would have hauled out the old Hartford Super Spacer and machined it that way but I had the code written, brass clamped to the table and the machining done before I would have had the Hartford set up I love CNC (i had to use the banana as well )

Hood

BTW have been building the control box for my next project, have been so impressed with the Bridgeport that my Student Lathe is next for the retro, will just have to wait until I can afford the steppers though.

[jdelaney44]

Hood,Cool beans! Well if you can sell a few parts you can get some more steppers!

Uploading a couple of DXFs that might be handy. One is for a ten tooth L series timing pulley. One is for a 30 tooth. These are not detailed, just the outlines. I'm using them this way so my puny Celeron doesn't get too slow.


-jd

[Hood]

JD, dont know if youve seen this but think it might be of interest
http://groups.yahoo.com/group/mach1m.../message/35796

Hood

[jdelaney44]

I did see that. Yes! Mach is going to be closed loop and function as a DRO too! Well with the addition of that board they are building. Sweet.

I'm really jazzed about that. There was a thread on that whole thing at CAD_CAM_EDM_DRO. A couple of people were tweaked that Mach wasn't closed loop. This has actually been debated a lot and the conclusion, I thought, was that it was no big deal 'cause it worked pretty good. But this is a big change.

I was beginning to think about trying EMC again after reading all that. But now it looks like I won't have to.

Thanks for the heads up.

[jdelaney44]

Here are the JPGs of the Y-Axis Drawings. Construction starts tomorrow. Maybe I'll get the box done tomorrow. I hope. I hid the side on that last isometric / exploded view to show the pulleys. I've used simplified block drawings of the pulleys to keep the load down on my PC. Same thing with the screw holes this time.

Once I get this built and both the X and Y fitted I'll post the .DXFs. Just don't want to set anyone off on a goose chase until I know all this fits and works well.

[jdelaney44]

Well, I did not make much progress today. The light switch in the garage broke so that turned into three little projects and a stop at my parents house to say HI! Hours latet I got started. THEN I crashed the vise into the column and threw that out of square. So that meant a vise remount. While I was at that, decided to cut some flats into the 5/8s bolts so they engaged the T-Solts. Etc........

But I did get the material rough cut for the front, back, and sides. I did a couple of unusual setups. So I figured I'd post pictures of thos activities.

-jd

[Hood]

Shame you never got started yesterday cant wait to see your finished Y.
I am hoping to get some MPGs fitted towards the end of the week so that will make edge finding a lot easier than jogging by keyboard. My plans are to build a seperate console and have a keyboard with a touch panel to do away with the mouse. Heres the keyboard I was thinking of getting http://www.kustompcs.co.uk/acatalog/info_3003.html

Keep us updated on your progress.
Hood

[jdelaney44]

Interesting keyboard. I am not a touch pad fan. I can never get the hang of them. I prefer the track balls like that Kensington track ball. Logitech had a wicked good track ball a while back too. They may still.

More on the pics above. I say unusual because the pieces are almost 24 inches long and .325 / .40 thick. So at the ends they get pretty wiggly. Also they were too long and too deep for my band saw. So I had to cut them with a .25 end mill.

Laying them flat on the table was no big deal.

But up on edge to start to true them up was due to all the material hanging out in space, I happen to have a couple of these angle irons that have been milled square, well pretty square. This shop in Costa Mesa sells them. They are good for structural welding & such. Not really square enough for layout. But, good enough to hold the ends of the material. So once I had the material in the vise I clamped the angle irons to the table and then the material to them. Did the trick. No chatter. Nice and quite. Lots of chips.

More later.