Capabilities of mid-90's controls:HSM, 3d contouring?

[JasonPAtkins]

Hi all,
In the next year or two as my new shop is built and machines moved in, I'll begin diving into the world of CNC machining. My background is computers, and I have a CNC plasma table, but these machines will be my first CNC machining experience. My machine shop up until now has been all manual.

I'm curious about the juxtaposition of old controllers and new software/techniques. I've watched lots of videos on Fusion and creating HSM toolpaths, but not many of those videos then export that CAM onto old machining centers/controllers.

My first CNC machine tool (which is already over in Africa where my shop is) is a mid 90's Tree J425, with the later Acramatic 2100 control (which is based on NT3.5 or so, I believe.) Am I being realistic in hoping that such a computer-based control (even though it's a *really* slow computer) should be capable of following HSM paths and doing 3d contouring?

The second machine is a 2011 or so Trak 1630SX lathe, but I don't see a Fusion post for it, so will probably just stick with its conversational programming. "Simple" programming seems fine for what I'll be doing on that machine anyway.

The third machine, which is the one causing me to ask this question, is an old (early 90's, I believe) Bridgeport Torq-cut 22. It's control is DOS based. I want to like the machine because it has a few things the Tree lacks (so it would complement nicely), including a tool changer, rigid tapping, and a spindle gearbox. However, before making the decision to send it to Africa, I'd like to be realistic about its capabilities. Would I, for example, be much better served to sell it and put the money toward something 10 years newer, assuming that the control is so old that while there is a Fusion post processor for it, perhaps its control won't be fast enough to keep up with an HSM tool path? Or are people doing just that with no trouble?

Thank you for any input you may have!
-Jason

 

[gregormarwick]

Hi all,
In the next year or two as my new shop is built and machines moved in, I'll begin diving into the world of CNC machining. My background is computers, and I have a CNC plasma table, but these machines will be my first CNC machining experience. My machine shop up until now has been all manual.

I'm curious about the juxtaposition of old controllers and new software/techniques. I've watched lots of videos on Fusion and creating HSM toolpaths, but not many of those videos then export that CAM onto old machining centers/controllers.

My first CNC machine tool (which is already over in Africa where my shop is) is a mid 90's Tree J425, with the later Acramatic 2100 control (which is based on NT3.5 or so, I believe.) Am I being realistic in hoping that such a computer-based control (even though it's a *really* slow computer) should be capable of following HSM paths and doing 3d contouring?

The second machine is a 2011 or so Trak 1630SX lathe, but I don't see a Fusion post for it, so will probably just stick with its conversational programming. "Simple" programming seems fine for what I'll be doing on that machine anyway.

The third machine, which is the one causing me to ask this question, is an old (early 90's, I believe) Bridgeport Torq-cut 22. It's control is DOS based. I want to like the machine because it has a few things the Tree lacks (so it would complement nicely), including a tool changer, rigid tapping, and a spindle gearbox. However, before making the decision to send it to Africa, I'd like to be realistic about its capabilities. Would I, for example, be much better served to sell it and put the money toward something 10 years newer, assuming that the control is so old that while there is a Fusion post processor for it, perhaps its control won't be fast enough to keep up with an HSM tool path? Or are people doing just that with no trouble?

Thank you for any input you may have!
-Jason

It varies wildly. You will have to solicit specific user experiences with each control/machine you consider.

For example, the oldest machine I have is a '92 Hurco with the Ultimax II control. It is a really slow control, built on the 8086 platform, but it does have dedicated cpus on each axis control board to offload motion from the main cpu. It has about 100KB of usable program memory, and can drip feed. It will follow HSM paths and 3D contouring with small segments just fine, just slowly. Programmable feedrate maxes out at 100ipm (which is annoying because it will go at least twice that fast in conversational mode), and at that speed the serial interface can more or less keep up so that dwells are rare. It has a pretty good amount of look ahead (for the time) so it doesn't stutter and pause too much either.

I run Vortex (Featurecam/Powermill HSM) toolpaths on that machine all the time. Other than being really slow, you still get the excellent tool life that goes with HSM.

In comparison, a Fanuc OM from the same period will be practically unusable with the same code.

Your bigger concern, given the machines you are discussing, should be how long and how affordably you can keep them running.... Little use worrying about whether that bridgeport can do HSM paths when the motherboard is dead and you can't get a replacement. That said, three axis mills are trivial to retrofit, and a LinuxCNC control can be as fast as you want it to be.

 

[JasonPAtkins]

It varies wildly. You will have to solicit specific user experiences with each control/machine you consider.

For example, the oldest machine I have is a '92 Hurco with the Ultimax II control. It is a really slow control, built on the 8086 platform, but it does have dedicated cpus on each axis control board to offload motion from the main cpu. It has about 100KB of usable program memory, and can drip feed. It will follow HSM paths and 3D contouring with small segments just fine, just slowly. Programmable feedrate maxes out at 100ipm (which is annoying because it will go at least twice that fast in conversational mode), and at that speed the serial interface can more or less keep up so that dwells are rare. It has a pretty good amount of look ahead (for the time) so it doesn't stutter and pause too much either.

I run Vortex (Featurecam/Powermill HSM) toolpaths on that machine all the time. Other than being really slow, you still get the excellent tool life that goes with HSM.

In comparison, a Fanuc OM from the same period will be practically unusable with the same code.

Your bigger concern, given the machines you are discussing, should be how long and how affordably you can keep them running.... Little use worrying about whether that bridgeport can do HSM paths when the motherboard is dead and you can't get a replacement. That said, three axis mills are trivial to retrofit, and a LinuxCNC control can be as fast as you want it to be.

Thanks for your advice. Yes, I do understand your concern. In this case, the Torqcut just got a new (and newer style backplane type) motherboard, so that part at least should be good to go for a while. I don't know about drive boards and such, but the computer part should be ok. The overall point is still taken, but in my specific application (non-profit shop in the middle of rural West Africa) the conversation is different. Here the States we weigh TCO over the life of the machine and in that case newer often ends up cheaper in the long run. In my case, it's old/donated or nothing, so the balance shifts a little.

Thank you for your input, I appreciate it. Soliciting specific user experiences on those two controllers using modern toolpaths is what I was attempting to do. Perhaps I should have made the title more specific.

 

[DouglasJRizzo]

Older machines and controls can make good parts still. You will find memory constraints to be the biggest road block. Not impossible, but something to work around.

 

[EmGo]

A quick read of the goldratt books might be worth some people's time ...

 

[BOB-OO]

The Bridgeport is very good and reliable machine, but that controller was not the best choice for 3d contouring, it does great in 2.5 milling. The same machine was offered w a Heidenhain TNC2500 which did much better in 3d applications. Make sure you have a backup of parameters & control software, it should do well in many applications but I would avoid high cutting feeds (HSM) and 100% Rapids. Treat an older machine like an 80 year old friend, they will usually treat you well in return.

 

[lumley32]

The A2100 control is one of the best you could have on a machine of that age. it will be quite happy with big files, probably got a 2Gb hard drive, networking is easy, has lots of features that some new controls don't come with.

of course, it's getting old and you might have problems, but that's the game you play, first step of troubleshooting is to reseat all the cards and connectors!

no idea if the iron will handle it, but at least you can get a modern tool path on the control, unlike fanuc of the time!

 

[PegroProX440]

At an old shop we had two torq Cut cncs. Never tried HSM on them because they were slow enough with legacy tool paths. Even our Fadals were crappy with crunching code. I'm sure you can figure out a retrofit to the control that will make it work better

 

[JasonPAtkins]

The A2100 control is one of the best you could have on a machine of that age. it will be quite happy with big files, probably got a 2Gb hard drive, networking is easy, has lots of features that some new controls don't come with.

of course, it's getting old and you might have problems, but that's the game you play, first step of troubleshooting is to reseat all the cards and connectors!

no idea if the iron will handle it, but at least you can get a modern tool path on the control, unlike fanuc of the time!

Thanks for the feedback! I only messed with the control for a couple of hours under power before I packed the machine up and shipped it to Africa. I've read the manual through a couple of times. Without much to compare it to, I was quite impressed with how capable the control felt, given its age! The computer hardware looked pretty standard, which is also an advantage in terms of repairability.

I have high hopes for that machine, actually. It has weaknesses (no automatic tool changer, no rigid tapping, limited programmable Z (though a big Z envelope because of the movable knee) but a lot of strengths, too. Box ways, should be very rigid. I like that it's open, too, since a lot of my work is onesie twosie stuff, being able to hang a part off the end of the table can be useful. Control, while being different than the Haas/Fadal/Fanuc/whatever people are used to, seemed pretty easy to understand. Looking forward to getting started on it! The next trick will be that being unpowered for the last several years probably means the CMOS battery is dead and I'll have to get BIOS settings figured out from scratch. That's just computer stuff though.

Thanks!

 

[JasonPAtkins]

Just a thought...TorqCuts have a gearbox in the head, but aren't very rigid machines.
As you're going to be a little "isolated" from service...incase of issues...would it be better to go for something different?
Perhaps a Fanuc control (mtbf for reliability and also for spares), and perhaps no tool changer and certainly not a gearbox head?
Less moving parts = more reliability.
And NOTHING is more frustrating in a shop looking at a machine that just won't run...

And ref the SX Prototrak - I had a SLX with a dxf converter. Never even used it - never needed to.
Conversational was superb.

Thanks for the feedback, that's helpful. I'm really excited to get into that machine. I've watched their demo videos and it seems extremely capable and pretty easy to program. I think it's going to be perfect for my application where I'm more likely to do 20 one-off jobs than a 20-part job.

Last year I had to make a round tube bending die on my manual lathe by interpolating the plunge depth every .010" across the 1" half circle. Jobs like that (or anything with curves, lol) are going to be so nice on CNC!! I also have limited threading capabilities (manual lathe doesn't do metric, but most of my needed threading is). I have high hopes for the new capabilities it will bring my shop. And now, I can cross the Holdridge radius cutter off of my wish list for the manual lathe, haha...

 

[JasonPAtkins]

Just a thought...TorqCuts have a gearbox in the head, but aren't very rigid machines.
As you're going to be a little "isolated" from service...incase of issues...would it be better to go for something different?
Perhaps a Fanuc control (mtbf for reliability and also for spares), and perhaps no tool changer and certainly not a gearbox head?
Less moving parts = more reliability.
And NOTHING is more frustrating in a shop looking at a machine that just won't run...

And ref the SX Prototrak - I had a SLX with a dxf converter. Never even used it - never needed to.
Conversational was superb.

The Bridgeport is very good and reliable machine, but that controller was not the best choice for 3d contouring, it does great in 2.5 milling. The same machine was offered w a Heidenhain TNC2500 which did much better in 3d applications. Make sure you have a backup of parameters & control software, it should do well in many applications but I would avoid high cutting feeds (HSM) and 100% Rapids. Treat an older machine like an 80 year old friend, they will usually treat you well in return.

At an old shop we had two torq Cut cncs. Never tried HSM on them because they were slow enough with legacy tool paths. Even our Fadals were crappy with crunching code. I'm sure you can figure out a retrofit to the control that will make it work better

Thank you guys for the specific feedback points on the Torqcut. I need to do a little thinking about it. It's at the very edge of my capabilities to move (barely fits in a shipping container, but I think it will, just, and is as heavy as my biggest pallet jack can move, but even then will have to be shimmed up in order to get the jack under it, and will be super tippy.) The gearbox is a concern, as actually the shop that's keeping the machine for me until I could send it next year just told me that it's stuck in low. I imagine it's just that the shot pin needs a little sandpaper or something, but the larger point of it being a fairly old and complicated machine is taken.

HSM and full 3d isn't absolutely necessary, a lot of the work I'd have for it would be done just fine 2.5d with simple toolpaths, and would be assisted by rigid tapping and good low end torque.

Thank you all for your input. Anyone have any concept of a fair value for a machine like this in Michigan if I did look to sell it and put the money toward something (perhaps a bit smaller and younger)? The computer was just replaced as well as one of the drives. Way covers were new a couple of years ago. I've read that these can run happily on 230v single phase, so maybe someone with a home garage shop would value it enough to justify the cost to move it?