[Coco] OT - Re: George's CNC Machine (WAS: Something else)

Charlie chazbeenhad at hotmail.com
Wed Sep 24 23:37:28 EDT 2008


I work for Cozzoli Machine Company in NJ. 19 years now. I started as a 
trainee and now run the CNC department.
I took 6 years out of the shop for the company and did lots of IT work, 
Networking the PCs and RS6000 as well as connecting through leased lines to 
our sister company in WI. (now closed).
http://www.cozzoli.com

We are looking to buy Haas machines soon. We LOVE  our Cincinnati machines , 
but they have changed to a Fanuc control and getting parts and repair is 
becoming a pain in the ass. Not like the good old days. Now I know someone 
to bother with questions if/when we do get Haas machines ;-)

Charlie


"Ron Bihler" <rbihler at msn.com> wrote in message 
news:BAY118-DAV1030612BFD03A732D3024AAF440 at phx.gbl...
> Looks like a nice machine, thats one controller I have not worked with, I 
> have used an Ecmo 220 control in the past.  Emcotronic - For a slow system 
> it worked pretty well.
>
> My primary machines are a 2000 Haas VF0 with 4th axis.  This is a pretty 
> good control system and so much easier to operate over a Fanuc.
> With that said my other machine is a Hardinge with Fanuc control, to be 
> honest I do understand the joy people have with the Fanuc control, very 
> powerful.  But the Haas is simple to use and just as capable for every day 
> use.  This Fanuc has lost it's memory due to a power outage twice, and has 
> cost me much more time and money to keep running.
>
> I am the owner of a Scientific Glass Blowing company www.techglass.com if 
> you are interested.  We produce some interesting glassware that I sell all 
> over the world.   One of the best parts of the job is I have been creating 
> some simple automation to take over some of the hand operations.  And to 
> do things that we just are not able to to normally.
>
> We have started to create some very small coils out of Quartz material. 
> These things hold .5ml of liquid and contain 18 wraps of material.  The 
> interesting way we do this is to stretch out the material to create a 
> smaller diameter.  All automated as a glassblower just can't do this on a 
> consistant basis or at all for that matter.
>
> We have just started to work with an 1998 Control system (PC based) in a 
> laser system.  Now talk about re-learning how it operates.  Basic G-code, 
> but the PC interface (running under w95) very strange indeed.
>
> I would really like to undertand the Inverse kinetic better, the 6 axis 
> robots uses this.  Was a 68000 based system and now a PC with a custom IO 
> board - actually it has 2 pc's running.  One to handle the user interface 
> and send position commands, the other to handle all the math and related 
> details to the Robot.  Funny thing was the old 68k system had a bunch of 
> 6809 units to handle the axis control. One processor for each axis, this 
> took care of the step and direction control plus all the encoder feedback.
>
> I must thank much of my old time Coco programming and basic lack of fear 
> to mess with this equipment.  The guys at the shop have said that I will 
> fiddle with everything.  :)
>
> Anyway it is always interesting.
> BTW what company do you work at?  I deal with several pharmacuetical 
> companies for glassware needs.
>
> Ron
> Author of RiBBS
>
>
> ----- Original Message ----- 
> From: "Charlie" <chazbeenhad at hotmail.com>
> To: <coco at maltedmedia.com>
> Sent: Wednesday, September 24, 2008 8:28 PM
> Subject: Re: [Coco] George's CNC Machine (WAS: Something else)
>
>
>> Hi Ron. I'm a programmer and lead man of a CNC shop for a pharmacuetical 
>> machinery company.
>> Have you ever worked with Cincinnati A2100 controllers? I think Siemens 
>> has the rights to them now.
>> We have several of them and I'm a huge fan even if they are very 
>> "touchy".
>> They are a dual mobo 486 or pentium1 setup that run winnt4.0. Basicaly 1 
>> board runs Windows and the user interface/touchscreen and the other 
>> controls the machine. Very powerful system. Ive always been amazed at the 
>> speed of things even though the cpus are slow 486's with only 32 megs of 
>> ram. Some of our smaller Arrows move at 1400ipm and can stop and repeat 
>> within .00002
>>
>> Here are the insides of the controller:
>> http://www.geocities.com/chazbeenhad/MISC/2001arrow500back1.jpg
>>
>> And here is the front of the little 20x20 Arrow 500.
>> http://www.geocities.com/chazbeenhad/MISC/arrow500.jpg
>>
>> Charlie
>>
>>
>> "Ron Bihler" <rbihler at msn.com> wrote in message 
>> news:BAY118-DAV72C491AD46A2262C6C0FEAF450 at phx.gbl...
>>> The speed all depends on how fast you want it to interpret the moves and 
>>> how accurate you need this to be.  Trade off, accuracy requires more 
>>> math and processor time hence slower movements.
>>>
>>> Most programs are taking the circle movement and breaking it into small 
>>> line segments. Again depending on how accurate you need.
>>>
>>> I am also running a sherline mill system, I am using an old Fijitsu 
>>> tablet pc.  Granted it is a Pentium, but turbocnc is doing all the work 
>>> of handling the stepper motors.
>>> The Par port is connected to stepper controller, step and direction is 
>>> all controlled and it moves faster than I would have expected 
>>> considering how slow the system really is.  I suspect this is still 10 
>>> fold faster than a coco, but then again without all the messydos 
>>> overhead one can never tell.
>>> This system is pretty accurate, good within .001" or better for circle 
>>> movements.
>>>
>>> I also have an old American Robots Robot.  6-axis movement, this was 
>>> running on a 68k system.  Once the system would boot (Close to 10 
>>> minutes) it was very good.  But that's a 68k system, required to perform 
>>> reverse kinetics movements.  Major math here.  I upgraded for a Pentium 
>>> II system, night and day difference.  Bottom line, the coco should be 
>>> able to do some basic circle interpolation, but it will not be the 
>>> fastest system on the block :)
>>>
>>> I also used an vintage Emco lathe system, no idea what processor they 
>>> used. It was dedicated to the task and was plenty fast.  But considering 
>>> I was only able to transfer the code at 2400 baud with character spacing 
>>> it must have been a slow processor.  1984 vintage.  The point was this 
>>> system did work very well with a much slower processor.
>>>
>>> I am assuming he is not needing .0001" accuracy here, doable but not 
>>> fast. This also makes the assumption the coco is doing all the work, 
>>> with JRKerr modules they just receive some simple serial commands and 
>>> control the steppers or servo drives.  In this case the coco should be 
>>> able to do much of the work and pass the true movements onto another 
>>> processor.
>>>
>>> I am not involved with JRKerr in anyway, but I sure have used there 
>>> products to automate several task within my shop.  I own a scientific 
>>> glassblowing company, I can't find good help anymore and have had to 
>>> resort to automating many of the task. Honestly I do have some good 
>>> help, but they hate repetition and can't repeat like automation.
>>>
>>> Sounds like a very interesting project, thanks to Chuck I may include a 
>>> 6809 system into some of my operations.  Sure is easier to program than 
>>> the messydos systems. I can understand what's going on, and it doesn't 
>>> need 2gb of memory to operate.
>>>
>>> Ron
>>> Author RiBBS
>>>
>>> ----- Original Message ----- 
>>> From: "Gene Heskett" 
>>> <gene.heskett at verizon.net>
>>> To: "CoCoList for Color Computer Enthusiasts" 
>>> <coco at maltedmedia.com>
>>> Sent: Tuesday, September 23, 2008 9:29 PM
>>> Subject: Re: [Coco] George's CNC Machine (WAS: Something else)
>>>
>>>
>>>> I'm not Ron B., but I do have a small amount of cnc experience, running 
>>>> emc on
>>>> a linux box to run a Harbor Freight MicroMill that has been heavily 
>>>> modified.
>>>> emc can run up to 9 axis's, and I'm currently running 4 myself.  The
>>>> processing power required to do that math in real time needs a faster 
>>>> machine
>>>> than the coco, but it can do it nicely on an old xp-1400 athlon, 
>>>> drawing
>>>> characters on the workpiece at speeds in the 10 inches a minute range, 
>>>> at any
>>>> angle, to sub-micron accuracy if the mechanics are that good.  Mine 
>>>> aren't,
>>>> but can usually stay under a thou for error.
>>>>
>>>>>
>>>>>--
>>>>>Coco mailing list
>>>>>Coco at maltedmedia.com
>>>>>http://five.pairlist.net/mailman/listinfo/coco
>>>>
>>>>
>>>>
>>>> -- 
>>>> Cheers, Gene
>>>> "There are four boxes to be used in defense of liberty:
>>>> soap, ballot, jury, and ammo. Please use in that order."
>>>> -Ed Howdershelt (Author)
>>>> sticky bit has come loose
>>>>
>>>> --
>>>> Coco mailing list
>>>> Coco at maltedmedia.com
>>>> http://five.pairlist.net/mailman/listinfo/coco
>>>>
>>>
>>>
>>> --
>>> Coco mailing list
>>> Coco at maltedmedia.com
>>> http://five.pairlist.net/mailman/listinfo/coco
>>>
>>
>>
>>
>>
>> --
>> Coco mailing list
>> Coco at maltedmedia.com
>> http://five.pairlist.net/mailman/listinfo/coco
>>
>
>
> --
> Coco mailing list
> Coco at maltedmedia.com
> http://five.pairlist.net/mailman/listinfo/coco
> 






More information about the Coco mailing list