G-code?

[BobO]

Let's see if I can stir up some discussion...

We are looking to add some coordinated motion to a future product. We have already implemented electronic gearing, axis following, and camming. Those that have played with it like what we have done...however...one area of weakness is in multi-axis paths. It can be done with multiple cams tied to a single virtual axis, but there is no good way to control the master axis in such a way as to optimize the slave execution time and/or tangental velocity (tool speed). So we need a multi-axis path instruction that can optimize for slave axis constraints.

We don't envision this being a G-code interpreter per se, but we do like the idea of taking G-code output from your favorite CNC path generator and post processing G0/1/2/3 to produce a time-based cam output that we can feed into our existing cam solver. The basic workflow we envision is a) make some G0-3 toolpath code, b) run through a DmD-based utility to produce a file we speak, and c) point a Do-more instruction to that file and 2 or 3 axes to do a coordinated path.

Questions:
1. Does that sound useful?
2. Is G-code the preferred source?
3. Are codes G0-3 adequate for most applications?
4. If this doesn't sound viable, what would make it so?

[Mike Nash]

You really are bored.

I had to go look up the G-codes just to recall that, yes, those are the ones I used when I was playing with CNC. I really need to finish those (personal) projects.

I have customers that run CNC machines every day, but they have industrial machine tool controls. I am drawing a blank at where I would need this capability in my day job type work. I reckon that means I won't be much help to you.

[BobO]

In discussions we've had with a few users, we feel that there is a middle ground between simple single axis stepper control and full multi-axis CNC where a PLC with a decent X/Y path control could play. So that's what we're trying to determine: if a PLC could do 2 or 3 axis coordinated path control, is that useful, and does our approach sound viable?

We don't want to waste time creating stuff folks don't want, nor do we want to miss the mark enough to render the feature unusable. We feel it is viable to implement though, and would love to do it if some of our resident motion experts tell us it is useful.

So sound off!

[PLCGuy]

that would really be helpful if you did the x/y coordinated motion. We do it all the time, but we us IAI actuators to do this. This can be very expensive. I certainly would be interested if Hosteng developed such a system.

[Controls Guy]

Let's see if I can stir up some discussion...

We are looking to add some coordinated motion to a future product. We have already implemented electronic gearing, axis following, and camming. Those that have played with it like what we have done...however...one area of weakness is in multi-axis paths. It can be done with multiple cams tied to a single virtual axis, but there is no good way to control the master axis in such a way as to optimize the slave execution time and/or tangental velocity (tool speed). So we need a multi-axis path instruction that can optimize for slave axis constraints.

We don't envision this being a G-code interpreter per se, but we do like the idea of taking G-code output from your favorite CNC path generator and post processing G0/1/2/3 to produce a time-based cam output that we can feed into our existing cam solver. The basic workflow we envision is a) make some G0-3 toolpath code, b) run through a DmD-based utility to produce a file we speak, and c) point a Do-more instruction to that file and 2 or 3 axes to do a coordinated path.

Questions:
1. Does that sound useful?
2. Is G-code the preferred source?
3. Are codes G0-3 adequate for most applications?
4. If this doesn't sound viable, what would make it so?

I haven't run into any of these apps myself, but I do see people doing things that, if I were doing them, I'd use a Do-More with this capability, so my guess is there is a market.  No idea of the size of the niche between independent or simply-cammed axes and full CNC.

[Controls Guy]

I have customers that run CNC machines every day, but they have industrial machine tool controls. I am drawing a blank at where I would need this capability in my day job type work. I reckon that means I won't be much help to you.

The first thing that occurred to me were machining-like operations in applications that don't demand the precision or robustness of an actual CNC machine.  Carving big letters out of styrofoam with a hot wire and two axes of motion for example.  Possibly CNC plasma cutting sheet metal for HVAC shops.  I'm sure they're using something proprietary for this now, but it looks like something where you could use a path-controlling Do-More if you're the OEM or want to build one in-house.

[BobO]

The first thing that occurred to me were machining-like operations in applications that don't demand the precision or robustness of an actual CNC machine.  Carving big letters out of styrofoam with a hot wire and two axes of motion for example.  Possibly CNC plasma cutting sheet metal for HVAC shops.  I'm sure they're using something proprietary for this now, but it looks like something where you could use a path-controlling Do-More if you're the OEM or want to build one in-house.

Plasma cutting was one of the specific requests mentioned. Fairly crunchy operations that don't require the precision of true CNC, but do require multi-axis path control. My gut tells me there are a bunch of apps like this that are being handled by much more expensive equipment than required.

[Controls Guy]

Ginormous 3D printer! 

[Garyhlucas]

 I think I have a little experience at this level of discussion. I am currently active on the Practical Machinist, Mach 3, Home Shop Machinist, CamBam, and Reprap forums, worked as CNC programmer, bought the CNC in our shop.

 I built my own 4 axis CNC in my garage running on Mach 3 with a motion card. A DL06 handles all the button functions, switches between the R8 spindle, router spindle, lathe spindle and 3D printer extruder. It also controls printer bed temperature and extruder head temperature.

A couple of weeks back I bought a Chinese 3020 router and built a test tube rack for it. I moddled the rack in SolidWorks, programmed the shop CNC using CamBam to cut out the 144 holes and then imported the same geometry into a cheap TopCNC TC55H CNC controller. 3 axis of control to move from a drain port to each tube and a 4th axis running a peristaltic pump to pull the sample and accurately dispense into the test tubes.  Works quite well except all programming is G-code only.

The problem with the inexpensive CNC controllers is that G-code does motion well but is very poor at user and I/O interfacing. The high end ones often have a completely independent PLC for I/O as well. I can't start the sampler from an external input. Recovering from a power failure is not possible. The only timing function is a pause. Things like tool changers need logic that G-code is poorly suited to and an axis or two of motion control makes the task much easier.

So I think if you offered some basic multi-axis motion commands there could be jobs I'd use it for. We built the sampler for inhouse use, but a number of people have already asked if we would consider selling them. We'd need better controller for that. It has to be standalone, can't run on a PC.

[Mike Nash]

Gary, I think you are going to have to bring a little more to the table than that...

 

[BobO]

I think I have a little experience at this level of discussion. I am currently active on the Practical Machinist, Mach 3, Home Shop Machinist, CamBam, and Reprap forums, worked as CNC programmer, bought the CNC in our shop.

 I built my own 4 axis CNC in my garage running on Mach 3 with a motion card. A DL06 handles all the button functions, switches between the R8 spindle, router spindle, lathe spindle and 3D printer extruder. It also controls printer bed temperature and extruder head temperature.

A couple of weeks back I bought a Chinese 3020 router and built a test tube rack for it. I moddled the rack in SolidWorks, programmed the shop CNC using CamBam to cut out the 144 holes and then imported the same geometry into a cheap TopCNC TC55H CNC controller. 3 axis of control to move from a drain port to each tube and a 4th axis running a peristaltic pump to pull the sample and accurately dispense into the test tubes.  Works quite well except all programming is G-code only.

The problem with the inexpensive CNC controllers is that G-code does motion well but is very poor at user and I/O interfacing. The high end ones often have a completely independent PLC for I/O as well. I can't start the sampler from an external input. Recovering from a power failure is not possible. The only timing function is a pause. Things like tool changers need logic that G-code is poorly suited to and an axis or two of motion control makes the task much easier.

So I think if you offered some basic multi-axis motion commands there could be jobs I'd use it for. We built the sampler for inhouse use, but a number of people have already asked if we would consider selling them. We'd need better controller for that. It has to be standalone, can't run on a PC.

So, yeah, I'd put you in the 'expert' class.

We have already implemented a few basic coordinated functions for the new platform, but we don't have a great approach for multi-axis path control, hence the query.

Currently support three physical axes and one virtual. Three coordinated instructions:
1. AXGEAR - simple electronic gearing with live variable ratio.
2. AXFOLLOW - slave follows master at a fixed ratio, and can do dynamic relative adjustments on the fly.
3. AXCAM - slave follows master as defined by a position cam. Up to 64 points on the cam, with cubic interpolation. Linear or rotary mode.

I want to add AXPATH, which is essentially G0, G1, G2, G3, and perhaps even G4...assuming that's enough to do what we need. Probably support feed rate code as well, so it can be changed on the fly.

[BobO]

I am a little confused about F code, though. I've seen some CAM software that generates F code on a per axis basis, i.e., F100XY or F30Z. I would think F would be defining tangential velocity (absolute velocity of the tool relative to the material), and not the velocity of a particular axis. Although I guess that could be describing tangential velocity of the XY plane, and then treating the Z tool plunge independently. Just trying to wrap my head around this, and hopefully come up with a useful expression on the PLC. Any feedback from our local experts welcome.

[Controls Guy]

I think I have a little experience at this level of discussion. I am currently active on the Practical Machinist, Mach 3, Home Shop Machinist, CamBam, and Reprap forums, worked as CNC programmer, bought the CNC in our shop.

I built my own 4 axis CNC in my garage running on Mach 3 with a motion card. A DL06 handles all the button functions, switches between the R8 spindle, router spindle, lathe spindle and 3D printer extruder. It also controls printer bed temperature and extruder head temperature.

A couple of weeks back I bought a Chinese 3020 router and built a test tube rack for it. I modeled the rack in SolidWorks, programmed the shop CNC using CamBam to cut out the 144 holes and then imported the same geometry into a cheap TopCNC TC55H CNC controller. 3 axis of control to move from a drain port to each tube and a 4th axis running a peristaltic pump to pull the sample and accurately dispense into the test tubes.  Works quite well except all programming is G-code only.

That's it??   

[Garyhlucas]

I am a little confused about F code, though. I've seen some CAM software that generates F code on a per axis basis, i.e., F100XY or F30Z. I would think F would be defining tangential velocity (absolute velocity of the tool relative to the material), and not the velocity of a particular axis. Although I guess that could be describing tangential velocity of the XY plane, and then treating the Z tool plunge independently. Just trying to wrap my head around this, and hopefully come up with a useful expression on the PLC. Any feedback from our local experts welcome.

F is just feed rate along a path whether it be 1,2, or 3 axis interpolated. An F5 G1 X1 Y1 Z1 move would be a path the diagonal of a cube at a rate of 5 along that line. Ussually if there is a separate feed rate for the Z axis it is a convenience thing for canned drill cycles so you don't have to keep changing a single feed rate.

[Garyhlucas]

Feed rate is a must for g-code motion. Also G0 rapid would be nice. Multiaxis rapid motion is ussually a dog leg. Each axis runs at the same high rate, so an axis with a shorter move will finish first, and the others will continue moving. A G4 on most system is a pause time not a move.

Someone mentioned 3D printers. The g-code for that far exceeds the memory of almost any PLC. I've cut some simple 3D shapes that took 80,000 lines of G-code, but a 3D printer program can easily run to millions of lines.