share your refit stories?

Howdy,
I am in the process of fitting my manual BP J into cnc... and wanted to hear from others that have gone this route, or have refitted a nc/cnc machine to pc control.

In my case, I am taking a fairly accurate manual, that I had fitted a turret powerfeed to, replacing that powerfeed and couple with a much more accurate one, utilizing the acme screws, on the first attempt, and using the knee for Z.

Why knee? I have a right angle head and wish to be able to use it, and I am willing to "suffer" with slower feedrates

I bought John's 600oz servos and geckos, and will be doing a 2:1 on x/y and a 3:1 on Z, along with gas strut(s) on the knee to help lift it.

My goal it to have a machine that can make parts that basically can be held in one's hands... monoblocks, sights, "go kart" frame pieces, at this time, the biggest piece i can see machining, and that will be in fortal aluminum, is a bellhousing for a bmw v12 to a caddy northstar motor.

I'll be "tucking" the Y and Z motors, hopefully one a "one piece" frame across the front. Sure it'll be a couple pieces, but it will probably be .75 fortal, and then welded (tig) into one piece at the end of the day.

I am planning on the MINIMUM modifications to the BP as possible, as I know that I may refit with ballscrews at another date.

i'll keep you posted
jeffe

"Fortal" is 7075. It's not weldable unless you happen to have an electron beam welder handy.

stick to mild steel for motor mounts. Its cheaper, stronger, and more easily repaired than any of the aluminum alloys.

I retrofitted a Bridgeport BOSS4 with Gecko 201 drives, a Campbell breakout board and Mach2 software and am very pleased with the results.

Donald,
sounds like you reused the steppers... what sort of accuracy are you getting? I am certain, but woudl like to hear, that the geckos and a pc allow you to do 3 axis movements, and since you said boss 4, even multi axis movements?

jeffe

I did a large Beaver Mill about 8 years ago.
It was a CNC one but the old controller, a Posidata, was well past it's sell by date.
A pig to program, tape reader broken, a bigger pig to edit and no way to save a program and it occasionally lost position.

Had a look round and Ahha in the UK did a plug in conversion board to use the original bi level stepper drives.
Went to see a demo machine and bought the setup on the spot.
It took a week end to do the conversion from a Darlec to a PC.
Kept the original steppers and driver cards and power supply but scrapped all the wiring, the old oil filled VFD drive and fitted a new vector drive.

Ahha now runs the machine including a 4th axis, spindle start and speed, controls to air operated gearbox and can rum either the air clamps or the air over hydraulic vise.

In 8 years I have had various problems, some machine related, some computer related, most operator related [ ME !! ] but none down to Ahha.
I have looked at getting an upgrade but to be honest I can't see what improvement I can get for the trouble and expense of a change.

OK it's DOS and dated but it still works, it can run 30,000 line files with no problem.

John S.

Something that hasn't been touched on here is the issue of screws. I know from personal experience that a cnc mill with appreciable backlash in the screws is nothing more than a cnc drill. Backlash compensation can practically make up for a thousandth or two of backlash as long as the parts aren't critical, but the theory that backlash comp can make non-preloaded acme screws work in a cnc application is absolutely not valid. .005" backlash in one screw renders the machine useless from a practical standpoint for milling since every move of the program must be checked to assure no climb cutting occurs. This is far more difficult to do in practice than it sounds in theory. Also, since the comp move takes place when the axis reverses direction, even something as simple as milling a square shape and holding size becomes a problem. Say for example you mill in the X+ direction first. When the cutter gets to the location and the Y move begins, the cutter is pushed in the X+ direction further by the backlash amount. Only when a move begins in the X- direction does the comp move get added in. Same thing happens again at the opposite end of the part. Theoretically the oversize should be equal to twice the backlash. The truth is, the oversize will be some amount that you'll only know by measuring, and it won't be repeatable from one part to the next. With anything more complicated than a rectangular shape, the problems only get worse. I took a major league whipping on a job I took several years ago, knowing in advance I had a backlash problem in one screw but thinking I could work my way around it. If I'd had a manual mill with enough travel (other than a HBM) to do the job, I could have made the parts faster manually.

[ 09-28-2004, 12:26 PM: Message edited by: metlmunchr ]