[MrHorsepower]

Hi guys, My big mouth may have gotten me into trouble again. I'm Plant Superintendant for a large manufacturing firm in the Twin Cities. We do A LOT of drilling tapping and countersinking, most of which happens in one of our VMC's. I have been kicking around the idea of building an X-Y gantry table, like a router, that I could hang some pnumatic drilling and tapping heads on, to ease to pressure on the VMC's. I mentioned this to the powers that be and they said "go for it".

I am looking for some recomendations as far as controll goes. I will need an X axis, and 2 Y axis. I will also need to output to approx. 8 relays either directly or through a PLC. I would like to be able to add 2 Z axis in the future.

I've got a pretty good handle on the mechanical part of my design, but I could use some help with the control.

My gantry will wheigh in at about 325lbs when loaded up, however rapid travel is not that important.

Should I buy a control package or piece it together? I would take recomondations either way.

I have the company credit card, so to speak, but I don't want to get foolish with it incase it doesn't increase productivity like I think.

Any help would be great.

Thanks.-Dave

[ynneb]

I dont see why you couldnt use a PC and Mach3 ($160 bucks) http://www.artofcnc.ca/index.html to control your servos / steppers and also a this ($98 bucks)http://www.homanndesigns.com/ModIO.html to do what you wanted.

[Karl_T]

I'd suggest you take a look at Camsoft, CNClite. Its an industrial control package that includes the PLC functions right with the control. I'd suggest that you fully spec. your needs to them and have the custom logic written and debugged on site. If you're in production, time is money and you can't fart around piecing something together. Your people won't forgive you if your idea isn't well made and thought out.

I see your'e in my back yard. There's a fellow, Dan Gendreau, that works with Camsoft products. He's good people. I'm sure Camsoft can reccomend another integrator as well.

I'm sorry. I'm on vacation in the FL Keys or I'd email Dan's phone. Fishin' is great and the temperature should hit 80 today. How's it back home?

Karl

[MrHorsepower]

It's back home here, I was thinking about camsoft. Looks like they have got it figured out. Just so I can ponder this over the weekend, Any Idea's on price?

[Karl_T]

I was a manufacturing Engineer for years. Custom controls ain't cheap. If you don't have $10K in the piggy bank, you're in trouble. Without knowing your requirements, I can't do any more than guess.

Big ticket items are: Camsoft software; Galil control card; personal computer; Opto22 I/O card and modules; Servo motors,amps and drivers; and labor to design and install all this.

I do this stuff as a hobby - labor is free - and I buy used for between $0.10 and $0.35 on the dollar. I also already own the software. Anyway, with this route, your machine would cost 1K or so.

Karl

[NC Cams]

There's a real potential for a brush with horro when you take a hobby type approach into a business environment.

I tinker with electronics and electromechanical stuff. I also run my own business. I"m aways torn between the lure of doing hobby stuff (lower cost plus high satisfaction when it works) and facing the reality of having to pay bills and help at the end of the week/month. Ultimately, one has to pay the bills and, at that point, you have to "go commercial" even if it kills you do do so.

IF you buy a turn key system (you can define the modularity of the system), you simply install it and it should run. If it doesn't, you have the vendor come in an make it run because you are in business and do have to make $$$$'s - that is your primary job for your employer.

Yes, it can be less costly to DIY it and you WILL look good when it works. But, if it doesn't or it has interfacing problems or the software has a bug and you have to rely on message boards and hobby type support for help/service, management can get REAL tired of that REAL quick. The "gee, we never ran into that before" line doesn't work when sparks are flying on a home built machine in a business environment.

It could rapidly turn from "our new machine works great" to a "YOUR machine doesn't work and when are YOU going to fix it nad you better do it SOON" - things generally get real ugly shortly thereafter.

Consider the liability aspect as well. Home brew stuff at home is considered as a "caveat emptor" sort of thing by most courts. However, if something should go wrong and workmens comp and/or a lawyer gets wind of a poorly designed system in a business that hurt somebody, he's gonna try to get SOMEBODY to pay for a "defective, dangerous system". Always evaluate any caveats that a control supplier may have on a system they are selling....

Unless you're good with software writing/development and hardware integration, I'd look for a control system that is as bolt-on and fully developed as possible.

Remember, this is your LIVELIHOOD not your hobby - treat it as such. This would probably get you a system that will run with minimal hassle and risk to your employment.

Some folks swear by a Windows O/S - others swear at it. Linux seems to be popular in some circles and it supposedly does real time multitasking which is something that Win doesn't. Could be wrong - check it out. It will be YOUR job/reputation if the software turns into vaporware that doesn't do what its supposed to do.

I'm not enough of an expert to say which O/S is better/best but the neighbor is an electrical/electronics engineer and he sought out NON-Win O/S based equipment when he started a CNC business as a legacy for his family. His machines run, make money and he hardly ever has a service guy there. Meanwhile, I"m just getting my head above water after trying to get a bunch of legacy stuff that was cobbled by a DIY electronics "expert" up and running. If I had it to do over and the money to do it in the first place, I'd have bought a more robust COMMERCIAL system for my various CNC projects....

Good luck and be careful what you wish for - your wish may get granted.....

[alfalfa]

MR HP,

(Karl - thanks for rubbing it in. It is a balmy 20 degrees here under bright blue skies.)
Back to topic - NC Cams has a good point about using a turnkey type system, there are a lot more "unknowns". That being said, Karl's point on Camsoft is also legit as well. It sounds like this is a "try it and see if works, we really have nothing to lose by trying it" situation. That gives you more flexibilty toward the DIY side (IMHO). There are several of us in the TC area that use Camsoft (I am a couple hours west of you), and I'm sure would be willing to show you "our stuff" if you would be interested.

AL

De Oppresso Liber - Is 61:1-4

[MrHorsepower]

I would be very interested! I need this thing to work, but I have time. My I idea would be to "prove" the theory with the homebuilt stuff and then retro it up to industrial strength if necs. I would like to buy a motion package from someone and I'm leaning to Camsoft. I apprieciate and respect the opinions. We currently have several high doller pieces of automated equiptment (Mazak lasers with auto loaders, Amada Vipros turrets with auto loaders, etc.) and they seem to be as quircky as my Bobcad and Boss 5 at home. We also cannot get Mazak to return our calls anymore about a glitch that sure seems to be a firmware type issue. They say it must be in our Radan (also big money) software, Radan says it's Mazak. There is a movement here pushing for more automation, which I am all for. Being part of the production team, I see the need to become more and more self reliant a must. As the technology advances at an exponetial rate in this country it will be harder and harder to get factory support. We currently have a great (read educated) maintainance dept, they just don't get the production end of it. Again, thanks to all.-Dave

[wizard]

To comment a bit on custom tools and maintenance, from somebody that has handled the maintenance end of things, it really doesn't matter who builds the machine. Some home built hardware works as well or better than the factory supplied stuff. It is a matter of applying time to the engineering, frankly time just isn't there these days.

Generally it is my belief that off the shelf hardware, that is tooling that has not been customized and has been stable for awhile is hard to beat for quality. In many cases service after the sale makes or breaks the vendors reputation.

As to something like a home built router I find it hard to believe that you can build an industrial quality unit at a reasonable price. Here I'm including the engineering and development time and effort. There are just to many vendors to choose from that have alredy gone through that process. I mention routers because there might be one or two on the market that can do what you want. Have you considered a HAAS GR series machine? It sounds like you need a drill and tap machine for large sheet goods, something like a GR "router" might work. HAAS also has a low speed spindle that might be useful for harder materials.

I understand what you are tring to do with drilling and tapping heads, there is potential. Especially if you have a bone yard of equipment to reuse. The problem is that I see having to zero out two or more heads or coming up with a set of offsets for each head. Doable yes but a maintenance/setup nightmare even more. In the end if you don't have a bone yard to work from, and the engineering and technical organizations to build the machine you are very likely to spend alot of money for a fixed purpose machine.

Thanks
Dave

[Karl_T]

MrHP,

I'm right on your path out to Alfalfa's place, a little over an hour west of Minneapolis. If you decide to go with Camsoft, a field trip would be worth your while. Stop by.

Karl

[NC Cams]

HP's comments about the software/hardware guy's finger pointing at each other is SO true. I have this basic philosophy: if you can't see electrons, you can't trust them.

The key to me seems to be to take proven mechanicals, match it with proven software control and you'll stand a reasonable chance of getting it to do what you want it to do.

Thus, if you'r gonna retrofit, find a system that is running nearby and talk to the user. See if there are "issues". The hardware is usually gonna have a legacy. Then it becomes an issue of system integration.

If you have the time to develop even a bolt together retrofit, fine. However, the chances of somebody without the time or software expertise bolting together a system that runs as intended out of a bunch of unintegrated parts is gonna be tough.

Hobbies are one thing, production environments are something a bit more serious. I'm speaking as business owner who's an engineer who CONSTANTLY fights the urge to get involved in another science project. Ultimately, you gots to pay the bills.

[MrHorsepower]

Karl, if you are in Dassel, I have to make a run out to Twin Cities Willys end of next week anyway. Could I stop by and check it out?

[Karl_T]

You'll be 1 1/2 miles from my place at the Willys, but I'm not home right now. Sorry.

Email me [email protected] and we'll set up a time.

Karl

[MrHorsepower]

OK, Some of you know about the mess I've gotten myself into with this gantry table I'm going to build for my work, so I won't bore you all again. here's my question. How much torque do I need to accel/decel a gantry that will wheigh in at about 300 pounds when loaded up. My intention is to use steppers (please don't debate the stepper vs. servo thing here, i'm trying to do this cheap to prove some theory). I am going to use a lead screw with a pitch of 5 and gear down with a belt drive. I am not looking for the ultimate rapid but I want it to work at a respectable rate. Thank's again.-Dave

[Zach_G]

Well now, that's quite a general question to a specific problem, so I'll try to give an equally general answer. Taking a mathematical approach, you can choose how fast you want your gantry to attain a certain speed and then calculate the necessary acceleration using V=a*t. Since F=m*a you can calculate the force to attain this acceleration.

Next, figure out the torque-force conversion for your particular leadscrew setup by using the formula Thrust=Torque*(2*Pi)/(Lead+Pi*u*PD) where Lead is the distance between a thread profile, PD is the average contact diameter of the screw, and u is the coefficient of kinetic friction between your screw and nut materials. We're going to assume the friction loss of the rolling element linear rails you're using is negligible in comparison to the friction loss of the nut. After obtaining all the necessary values, you can work backwards through the calculations (watch those units) and arrive at the necessary torque to get your gantry up to speed. If you're gearing the stepper, make sure to multiply by your gear ratio.

Of course, the torque of a stepper motor is not constant over its speed, and the torque-speed curve varies for each type of stepper controller you use a stepper with. Picking a stepper with a holding torque of about 2-4x greater than your calculated torque should give a good ballpark estimate. Don't forget to make sure that you're within the operating speed of the stepper too with some torque left over to cut the material. The calculation for stepper speed based on gantry velocity should look something like angular velocity=(V*Gear Ratio)/Lead. Or maybe someone with actual experience can just tell you what they used...

[Karl_T]

Or if you want a SWAG (scientific wild ass guess), get a 1000 oz/in unit.

I have a couple steppers off a BOSS 3 bridgeport (2000 oz/in) I'd loan out to you. Big and bulky, may not fit (size 42). I've been meaning to sell them. But, last time I sold something, I needed it the next week. (Official pack rat here)

Karl

[MrHorsepower]

That's very kind of you to offer Karl! If you decide to sell I'm interested-but I don't want to borrow them. I didn't realize that those Boss steppers were that strong. Are they the same on my Boss 5? Anyway I ordered 1780oz. motors from the ebay guy and am going to use a dtllc board and Mach 3 for my testing. If we can then recognize the time savings the machine provides, I will then switch to Camsoft and servos. I am so busy in all asspects of life right now, but I would still love to take a drive to dassel in a few weeks if that's cool. Maybe I'll bring enough cash for the 2 motors-I've got a project at home I could use them on.-Dave

[ViperTX]

Okay....well spit on your hand and grab a ground and use your other hand to search for electrons.....I'm sure you'll find some....even if you can't see them.....bet you can't find the bits or bytes that way....*chuckle*

[mhalling]

That what I do for a living make specail machines give me a call 605-628-2700 Mike

Hi guys, My big mouth may have gotten me into trouble again. I'm Plant Superintendant for a large manufacturing firm in the Twin Cities. We do A LOT of drilling tapping and countersinking, most of which happens in one of our VMC's. I have been kicking around the idea of building an X-Y gantry table, like a router, that I could hang some pnumatic drilling and tapping heads on, to ease to pressure on the VMC's. I mentioned this to the powers that be and they said "go for it".

I am looking for some recomendations as far as controll goes. I will need an X axis, and 2 Y axis. I will also need to output to approx. 8 relays either directly or through a PLC. I would like to be able to add 2 Z axis in the future.

I've got a pretty good handle on the mechanical part of my design, but I could use some help with the control.

My gantry will wheigh in at about 325lbs when loaded up, however rapid travel is not that important.

Should I buy a control package or piece it together? I would take recomondations either way.

I have the company credit card, so to speak, but I don't want to get foolish with it incase it doesn't increase productivity like I think.

Any help would be great.

Thanks.-Dave

[MrHorsepower]

Okay....well spit on your hand and grab a ground and use your other hand to search for electrons.....I'm sure you'll find some....even if you can't see them.....bet you can't find the bits or bytes that way....*chuckle*

I DON"T GET IT!