[Willbird]

Well This is the start of all the fun, been reading here for a few weeks, and orderd some parts. I have 2 bridgeport mills, one dad bought new in 1972 when he had a shop in the garage, I still remeber the soapstone square on the floor where she was gonna set that he drew after he ordered her. She is a step pully 9x42...and she is gonna become a CNC machine. The other machine was bought new in 1986, vari-speed 9x42 and for now she is staying stock.

Coming via USPS and UPS is one 1160 oz/in stepper motor from keling, I got the dual shaft one because if I want to add encoders it gives me a place to do so. Also coming from keling is a 20 amp 72 volt power supply and a gecko G201.

Coming also is a C11A multi-function cnc board from CNC4PC and some timing pulleys and belt, I have the bracketry all drawn up in acad14 for X any Y, going 28-motor 56-screw on the pulleys....keeping the brackets fairly simple, I will post drawings from home later, basically a 1/4" steel plate behind the stock bport mounting, with another plate stacked on top to hold the motor, and some provisions to adjust belt tension.

I drew some long drawn out flight of fancy stuff to put in better thrust bearings than stock...but in thinking about it went simpler to get the machine moving with the acme screws...fun and fancy stuff it can make for itself and earn it's own keep instead of loafing while I turn the handles.

Hiwin ground ball screws are further down the road.

I'm not sure on the Z but leaning towards using the Knee to start anyway, I did some Boyles law work and it looks like (2) 2" bore air cylinders plumbed into a decent sized tank will allow counterbalancing the knee within + - 10 lbs at either end of the 16" stroke....I'm going to go 3:1 to 4:1 on the knee stepper once X and Y are done and use the acme setup at first...then if it looks decent, build a knee ballscrew.

I was taught to never mill with the quill down unless you had to, and dad would thump a boy if he caught him at that business hehe.

Bill

[mxtras]

Sounds like you have a plan!

One suggestion - run the G212s instead of the 201s - the 212s allow micro stepping and they run much smoother than the 201s. I used 201s for start-up of my conversion then switched to the 212s - big difference in low speed performance and smoothness of overall motor operation.

Scott

[Willbird]

Well it is too late because (1) 201 is on the way, I just looked at the Gecko webpage and the only differance I see between the 201 an the 212 is that the 212 can be set for full step, half step, 1/5 step, or 1/10 step but both drives are micro step drives ?? The only differance in the specs I can see between any of the drives are that some have short circuit protection and the 203v has all kinds of neat stuff ??

Bill

[mxtras]

The 212 also has a built in capacitor. From what I gather, there is also another difference, but I honestly can not recall what it is at the moment.....but I know you can buy a chip from Gecko to basically turn a 201 into a 212 and add your own Cap to smooth it out if you so desire. There is a huge difference in performance but it may not be a requirement for you - I don't know.

Perhaps someone else will chime in and either clarify or straighten this info out? Either that or I will do a little digging to refresh my feable memory and post an update.

Scott

[Willbird]

I do remember reading that in the manual for the drive, about upgrading it by pulling something out of a socket and plugging something in there :-) now that you mention it. I only have bought one so far so we can sure change what the others end up being.

I went to post a cad drawing last night and forget all about not being able to save it as .jpg or whatever from autocad....what a PITA :-)...remembered I used to just use a screen capture program to save the file as .jpg.

Bill

[mxtras]

You can 'BMPOUT' of ACAD. It's not that great, but it works.

Scott

[Adobe Machine]

Sounds like a good plan,will you use the present set up to mill the bearing mounts for the ballscrews once you get going ?

Frankley , it is just real nice to have minimal back lash or none for complicated parts..Doing a 12 bolt circle pattern, .001 clearence on each bolt, and the everything fits so good you can start and finger tighten just makes your day !

What software will you use ? Does your CAD program have CAM capability? You can not believe how good it is to CAD it then push a button and all the "G" code is done..now..fast. The only thing I have to do is add 2-3 M codes, add the Feed speed on the 1st G01 and Export to notepad..

I have had some 12,000 line programs that executed flawlessley using the CAM program , and truthfully have some MDI( Manual ) programs that were just horrible, so I draw everything up in CAD 1st,as there are a lot less mistakes, crashes and frustrations.

Most of all, have FUN..

Adobe (old as dirt )

[Willbird]

Yes I plan on making some booty kicking ABEC 7 bearing mounts when it is running and the ballscrews are in.

Yep 0.0000 backlash is the good stuff for sure, circle milling holes of any size beats boring them all hollow :-).

In a past job I used BobCad, some malign it, I never used it as a fully functioned CAM, just to draw difficult 2d toolpaths, cut and past them into subroutines and off we go. A good CAD/CAM will be another learning curve :-)..

Simple stuff I can think in G Code and it flies out of my fingers and onto the screen like magic...CAM can be a crutch to the point that I have seen people never learn to use it rathar than have it use them. Some of the shortcuts you learn from hacking G code by hand are good practice anyway, like using subroutines instead of repeating he same profiles over and over with several tools and thus making a longer program (all opinion on my part and worth what yaal paid for it )

I have ACAD 14 right now, and only work in 2d for the most part...have played in 3d but that isnt acad14's strong point from what I hear

Today the usps and ups truck brought a huge freaking power supply (22lbs worth) an 1160 oz/in stepper motor...a gecko 201(that stuff from Keling) a multi-function CNC board(from CNC4PC)...and some timing pulleys and a belt, I went with the GT2 pitch 15mm wide. And I stepped on my organ ordering the 56 tooth pully so if anybody needs a 50 tooth one with a 1/2" bore and no flanges I can hook you up :-) (A 6A55-050NF1516 is the part # of the extra timing pulley I have from stock drive products) I was jacking around with pulley sizes and got my numbers crossed, or just mis-clicked as it is the one above the one I actually wanted.


Bill

[Willbird]

OK lets try this :-) this drawing is just a rough out to give me some ideas on center distances and pulley sizes, etc.

Looks like it worked :-).

[Adobe Machine]

I would have to confess that I use CAM as a crutch..G code does not " fly off my fingers", in fact , not drawing the piece and using CAM, usually ends up a frustrating mess for me. I have several examples of some of my messes that as soon as I'm out of the wheel chair, they are going on a wall display, to constantley remind me slow down, draw it up 1st. In fact I sometimes print the part on a 1 to 1 basis, cut out the part and fit it up to ascertain that in the real world what I concieved fits and will function.
Couple of questions : What ratio are you using, servo to acme screw ? Can not tell from the drawing, but are you useing tensioner or slide to adjust the belt ? I have found from some past expensive experiments ( read screw ups) that if I did not have at least 50% of the drive sprocket engaged with the belt ( wraped) that it can slip.In fact I use a simple idler and a tensioner to both get good wrap and the correct tension, as opposed to a slide that you pry down..
Keep up the pictures, every one is intrested in your progress !

Adobe (old as dirt)

[RICHARD ZASTROW]

Adobe,
I hate to disagree but, (assuming the drive sprocket is on the motor) I believe you can have 60 degrees of belt engagement, as long as 6 or more teeth are engaged, and transmit the rated horsepower of the belt. At least I hope so or I'm in trouble. I got the above from a Godyear engineering manual.

[Willbird]

Here is what I have roughed in for adj. The plate with the motor on it can be moved 1/4" by the setsrew pushing against a pin or suck on a riser that is attached to the bottom plate.

I did see how you did your lathe Adobe,,,,I'm not an engineer by far, but to me it looks like the considerations the Konstructor must take into account are belt stretch, pulley wrap, and overall length of the belt that is in tension. (the shorter that distance the better for what we are doing with it IMHO) I did leave room there to put in 1 or two tensioners like you used on your lathe....and I intend to build a simple center distance test jig to see how much stretch takes place with my length of belt,

what is in my mind right now is to drill+ream a 1/2" hole in a piece of steel in the mill vise, and put a .500 gauge pin in the hole (.499 if the hole is tight on a .500) then put one pulley on it, and put another gage pin in a .500 collet in the spindle and slip the other pulley on it, and move the mill back in Y with the belt on the pulleys and see what center distance measures like with a decent belt tension...then build the system with the data thus gathered.

I went with 28 teeth on the motor and 56 on the leadscrew, 5mm GT2 pitch belt in 15mm width 80 teeth long....the center distance calculates to 3.6347 with the Stock Drive Componants center distance designer

https://sdp-si.com/Cd/default.htm


Before I had CAD I would get a mental idea and then start making chips....sometimes making paperweights too...........the CAD is a great tool for us, if your like me you play with an idea, and then get taken away to do other things...days, weeks, months, or years later I can open the drawing back up and my work is still there and I can pick up from it and go forward or apply the concept I was exploring to another plan, or finish it even.

[Willbird]

Here is a CPU heatsink/fan I found at the no speaka english computer store to mount the stepper drivers on, good heatsink I think for $12 including a fan :-). I have a 6-32 screw shown on one of the .075 wide slots, I think we can pull that off with some care, got (4) units of the heatsink/fan so I will set the control cabinet up for 4 axis minimum.

I decided to put my 72 volt DC power supply and the 12/5 volt DC power supply in it's own seperate box alongside the control cabinet, that will keep all the wiring in the control cabinet DC wiring, no AC allowed :-).

I'm using a 100 amp main panel with the main panel stuff stripped out for a control cabinet...nice and roomy and $58 with breakers which will find use elsewhere. I have another box big enough for the power supply cabinet.

I'm going to use an AT power supply for the 5 and 12 volt DC, I got some 10 watt 10 ohm resistors Radio Shack part #271-132 to load the AT power supply so the voltage does not drop if it runs without an external load on it. The link below describes how to do it with an ATX supply, but I have to have a neat small AT supply I'm going to use...some folks use the little plug DC power adaptors from Walmart, but I choked at paying $15 each for a 6 volt and a 12 volt, and they didnt HAVE a 5 volt anyway...the AT supply will have amps to spare for the job...

http://www.wikihow.com/Convert-a-Computer-ATX-Power-Supply-to-a-Lab-Power-Supplyfficeffice" />>>

I got some 3" square fans to move air in and out of the PS and Control cabinets. Got some heat sink paste too :-).

[Willbird]

This looks like it will mount up better with the gecko rotated 90 degrees. I see these same heatsink/fans on Ebay pretty cheap, might just be one of the cheapest things going :-)


Bill

[Adobe Machine]

Richard: I have seen the " 6 tooth" minimum also, but where a reduction gets higher, say 5 to 1, and the belt shorter, then you can have jump over. We had a belt engineer come and help us design a system with gilmer's that was replacing chains and sprockets..heavy duty stuff, but if the belt jumped once, it began comming apart..His advice in all circumstances was at least 50% engagement of both driven and driving pullys. We successfully used these gilmar systems in all of our products, and increased the time between changes 3 fold, as compared to chains.We also began using them in lower torque situations ( such as timed controls), and again the engineer advised to use 50% engagement. In all cases we used at least a tensioner, and in some cases we used as idler with a tensioner to correct the wrap angle.This advice from the factory engineer may have been CYA, as we used his product, but even in some real bad enviroments , never saw premature failure.

I looked at some factory built CNC equipment before even starting on the Tree and the Lathe, and they all seemed to have good wrap, most had a least an adjustable tensioner.I was suprized at some of the length ( long ) of the belts, but I guess Manufacturing space constraints ment they had to use a longer belt than I would have liked.

Willbird: I like those heat sink / fan for the amplifiers. I guess from what I've read here that by keeping these cool, you can run more substained amps to the stepper / servo, thereby reducing the chance of lost steps or shutdown.
Since I have not finalized my wiring system ( its higher than me and the wheel chair can get right now ) I'm going shopping for a similar system.
Thanks for the pictures

Adobe (old as dirt )

[NC Cams]

Belt wrap required is a function of the theoretical minimum tooth engagment AND (perhaps more importantly) how much power (torque) you'll be putting thru the system at ANY POINT IN TIME.

Simply put, as TORQUE goes up, MIN TOOTH engagement has to go UP as well. Otherwise, unless you use timing chains which won't stretch like belts, the belt will jump or break or prematurely wear.

It is amazing how folks will design to the minimum. Instead, they should read between the lines in the design specs and judge/consider the downside risk of what can/will happen if you apply way more than the minimum torque which is what the minimum tooth engagement will tolerate.

[Willbird]

Yup the minimum is not often optimum :-). I prefer to Konstruct things using Russian engineering, lots of cast iron and make it at least twice as strong as it needs to be :-). I did some measuring today and got the bolt centers for the bolts that hold the bearing housings for the X and Y leadscrews on the BPort....drew them up and printed 100% scale so I can have a double check and make sure things line up :-). Before putting cutter to steel.

Bill

[RICHARD ZASTROW]

I tend to over-engineer as well. I am not using the minimum 6 tooth engagement. I am using 11 teeth. There isn't enough room for idlers so that is where my "trouble" comes from. I checked with Gates and they do not reccomend idlers unless required for unusual circumstances. I didn't get 50 percent engagement only 37, but the belt width gives me a 100 percent overkill on torque capacity.

[Willbird]

Well on the idlers, I would guess that the reverse curve on the belt would lead to a shorter life than a belt that only got bent one way ?? :-). I have 11 teeth mesh on the motor 32 on the leadscrew.

I like the idea of idlers, but this setup is very russian, get it working, make some parts, among those parts will be some Campagnolo type bearing mounts/ motor mounts :-).

Looking at the X leadscrew bearing mounts, the right side looks as though they used the same casting same machining as the left (thrust bearing side)...but they simply pressed a single bearing into the housing instead of using a retainer like the left side uses to load it's two thrust bearings...the right side is drilled for the 3 screws to hold the retainer in. The whole point of this is that I need to regain the 1/4" from sandwiching the motor mount plate between the table and the bearing mount, it looks as tho I can simply press the bearing in the right end housing in another 1/4" and machine 1/4" off the end of the part the dial mounts on (maybe also make a 1/4" spacer...need to examine more)...the bearing in the right hand side could stay where it's at but the inner race would not be as well supported on the leadscrew as if I can press it deeper into it's bore.

Bill

[Willbird]

Yup the bearing will easily go 1/4" deeper in the bore so we're good to go, I can see no issues with moving the leadscrew 1/4" to the left and making everything work out right on the right hand end.