G code errors

[TDIPower]

did a simple design of a keyless chuck for a cordless/ screwdriver ahead of a project I am running this Friday. Had it running from 4pm to 8pm last night at an open evening.

The chuck was to be machined facing upwards.

I normally use raster x-y for work but last night I though I would try some of the other options to get a feel for what they do.

Roughing command was the first option (can't remember what that was here at home will update here)

The finishing command I used spiral as it looked to be the most logical from the little picture.
After lots of messing about with over travel alerts due to the program being written to go negative on the x command to machine the whole block I got it to run.

It completed a lot of the program but then hung/crashed (needed task manager to kill VR5) at the same place twice. Given issues I have had with VR5 in the past with this I decided to run the program in VR2. VR2 came up with an error when the cutter was in the same position as in VR5 and this is where there is a G19 in the code. A quick google told me what G19 (G18 & G17) are for but why is it crashing on this?

My only assumption is that QuickCAM 3D, VR 5 and newer machines understand G19. So when it is post processing it is creating a program which contains commands that the Micromill 2000 machines don't understand. This is why VR5 is crashing mid program as the micromill doesn't understand what the command is and hangs. However VR2 is a bit more intelligent and gives a report that it doesn't understand it.

Shouldn't VR5 take into account the limitations of the machine it is producing the code for and use commands that it understands or is there something else going wrong.

I have attached all the G code files I did last night, the last one almost ran to the end but I had to stop it as the Micromill lost position with 300lines of code left to run

Feedback Please

Pete

chuck.fnc

First program

(149.7 KiB) Downloaded 964 times

chuck1.fnc

2nd program

(139.92 KiB) Downloaded 991 times

chuck2.fnc

3rd program

(373.78 KiB) Downloaded 995 times

[Steve]

Hi,

The G19 is generated by the CAD engine automatically and is not used by the Denford machines. It is a plane select command. The engine that generates the code in QuickCAM Pro is an industrial program and has more features than we use. The G19 does not appear to be an issue.

The problem I suspect is more likely due to the working area of your machine. The roughing passes and approach moves can be outside the block and on a larger machine this does not cause problems.

I assume you are using QuickCAM PRO and not QuickCAM 3D? The spiral option is only in PRO. What version are you using?

What version of VR Milling 5 do you have?

The spiral mill cycle would not really be suitable for the job as the tool is constantly revolving away from the center so it is not possible to machine a true circle. I would try the raster waterline option as the waterline fill follow the vertical profiles and the raster finish the flatter surfaces.

On your computer can you simulate the cutter path in VR Milling without error?

[TDIPower]

the version of VR5 was not the most current, ill come back to this...

I dropped the latest version of VR2 on and this would let me run the files I produced with QuickCAM Pro (in 3d simulation). the earlier version hung at any G17, G18 or G19 command with an error code.

VR5 will run through the files in simulation.
I installed the latest version of VR5 and ran a file I produced using raster waterline as you suggested. I was doing some work on a PC next to the machine and I could hear a solenoid firing every so often from the micromill (sounded like the door lock) whilst it was machining. I gave the door handle a tug and it was locked, a moment later (while it was still machining) it clicked again, this time when I pulled the handle the door swung open!

This can't be right can it? It can't be anything in the G code so VR5 is doing something very odd.

I'm getting fed up with it all now, I have spent a lot of hours trying to get 2.5d and 3d work to be machined out and am now running out of time and contemplating ditching CNC work and the machines. I have 2 machines both miss behaving since moving to quickcam pro and vr5. I can't drop back to VR2 as I need to import files from techsoft 2D design, I get lock ups which require full reboots to clear when running those. For the 3D designs I don't know what happened to the original stl > g code post processor software so can't drop back to that (can't even remember what it was called now). We have never had these running properly since the upgraded software, we had had to change as the school moved to Win7, in 2012, I put it down initially to it being school network but we are now running XP on stand alone machines again as we had for 11years without issue. I had had to go through all the machine settings on vr2 and vr5 until I found one error in the setup file on VR5 which I posted about a while back. It just seems that either VR5 and the Micromill just don't get on OR both the Micromills developed the same faults at the same time on the same day VR5 was started to be used.

[Steve]

Hi. We still produce MicroMill machines and they go out with VR Milling 5 and can also be supplied with Quick CAM Pro.

As he MicroMill is quite a slow machine with limited spindle speed and working area most people use routers for 3 D work though there is no reason why the MicroMill will not machine 3D parts.

The error you describe with the door solenoid locking is not related to the software version as this is controlled by the machine firmware. The door locks when the spindle is running. If it unlocks when the spindle is enabled and there is a green indicator on the spindle control panel then there is a hardware fault.

We have seen some problems with the locking guard switches where the contacts bounce indicating the guard is open for a split second (this sometimes happens as the guard locks as the spindle starts or if the guard in knocked while machining). There is a fix for this in the current version of VR Milling 5. Can I suggest you download and install this?

Where are you located? Could you bring a machine and computer up to Brighouse and we could take a look at the problem there?

[TDIPower]

OK lots of tests done.

1st thing. I have used both the micromills successfully to produce 3D models in foam and MDF from day one (back in 2001). I am well aware of the slow speed (about 1hr per job), they are all we have so have to live with it. Yes I would love something faster but that is the long game.

The machines were fine running on P4 2ghz machines with 128mb ram VR2.8 and then VR2.14. I can't remember what the post processor was we used to get PROD STL files into G code but it was one supplied by Denford (this disk has disappeared). This was up to 2012.

The school upgraded the PCs and I had to have a new set all i3 windows 7. The combination of having to swap onto QuickCAM PRO and VR5 the new computer system and time meant the micromills were not used.

I picked up a couple of old P4 2ghz 512mb ram laptops and started to play. We needed to get Techsoft files engraved onto material. This was the start of getting the machines back up and running. After having to correct an issue in the machine file (see my other posts) to even get things to run We have issues with the machines crashing when running techsoft files just Text (imported as DXF) 1 in 4 programs will crash requiring a full reboot of both micromill and laptop.
Then onto the 3D machining. The same thing, programs crash locking up the machine and so on (see my other posts)

Well after last week with the door unlocking itself mid machining I have done a lot more testing.
I got another higher spec PC on Friday, did a full clean install of XP all the latest versions of software from the site.

Ran the program this morning (using your recommended machining methods) and got the same issue with it going crazy and then unlocking/locking the door.

I found the NC block start thing and have narrowed it down to between 5674 and 5923. it also seems to pause as if unsure what it needs to do at line 5604 5629

Up until this point it had been running fine. Every time I run from 5674 it makes really odd noises (not the normal tune it plays) but I can't pin it down more than that.

My next step was to clean install windows 7 and see if that made it run right.

So we are down to having tried 2 laptops and a Dell pc, all with the same issue, 2 micromills.

As for being able to bring a machine to you. I would love to but it would have to be in the October half term as I am teaching everyday.

See attached file

chuck3.fnc

(298.01 KiB) Downloaded 949 times

Pete

[Steve]

Hi,

I have downloaded the program and will try running it on a machine going through production. I cant see anything wrong with the code and wonder if it is to do with vibration causing contact bounce on the switch. Do both machines do exactly the same thing at the same point?
Could you give me the machine serial numbers from the label on the side of the machine also the year of manufacture? A photo of the electrical panel may also help. I am interested to see which type of guard interlock switch is fitted and the electronics that controls it.

Are the machines communicating with a USB cable? I don't think the computers are going to be the issue.

It would be interesting to see exactly what you are doing when exporting files from Techsoft. Text is quite an awkward thing to export as it will try and export the font data as well as the lines. Could you try selecting the Text in 2D Design then from the Edit menu select Explode. This gets rid of font data and converts the text to lines.

Then export the file as a DXF.

The older 3D software would have either been MiniCAM which is now no longer available (Supplied by NC Graphics) or the older Denford package QuickCAM 3D.

[Martin]

I have tried a machine here and starting at line 5600 I can see the change in noise you refer to coincides with the axes starting a lot of short moves. The machine here has a USB Communications cable and a NextMove Card. The NextMove allows move buffering to increase 3D throughput. If I turn off the Turbo in VR Milling this turns off move buffering and would simulate the program as run by a Denstep control card which you have.

The old MiniCAM and QuickCAM 3D software packages were both limited to raster output programs. I suspect the Control Card you have is having problems with 3 D contouring. I don't know why the spindle guard is locking and unlocking.

If you restrict QuickCAM PRO to just raster roughing and finishing is the machine any better?

The option may be to upgrade the control cards to USB which would allow faster manufacture. I will video the part of the program in turbo and non turbo mode to show the difference starting at 6570.

[TDIPower]

I did a raster rough then a raster finishing program. I did notice that the raster finishing in this version doesn't give the option of machine in X, Y or X & Y but only in X so the quality of finish is not going to be as good as i used to be able to get if I have to use this format.

Ran the program,

Between 4840 and 5150 it did the odd noise twice and messed up location. This is repeatable if I run using the NC Block tool (very useful to narrow this lot down.)

Running the programs in the 2d/3d sim on VR5 I have noticed that at various points the job flashes up in red, does this have any significance?

If it wasn't for needing VR5 for importing techsoft files I would be on the hunt for minicam so I could go back to running VR2.14. On that note and to prove a point, would you have a copy of minicam you could produce the NC file on for me to try with VR2.14 (yes I know there is a later version but I want to go back to everything the way it was to try and pin down the issue. Ive attached the chuck STL file.) From memory I used to do roughing at 80% step leaving 1mm for finishing. For a file like the chuck I would raster finish in X & Y. The cutter is 6mm Dia ball nose 55mm length 50mm flutes 2 flute. As it is only modeling foam I would rough cut at about 15mm per pass.

Im about to take off the back cover and check out this interlock issue as it now isn't locking at all with this file.

You mentioned about upgrading the control gear. What improvements would this make to the machine besides USB and better buffering? what would be reliable traverse rates. I used to run 300-350mm but have been testing on 200-300mm in case that has been an issue.

I would rather be able to get a Microrouter but the cost is way out of budget at the moment.

Ive not given up yet but have invested way to much time in these. I just want there to be a simple solution, Is there any information I could send you that might help?
When I was trying to get these running I found I was getting 'Access Violation at address 0040A72B in module 'VRMilling5.exe'. Write of address 00000190' So went through every detail of the machine set up compared to VR2 (viewtopic.php?f=2&t=4537) I had to change a VR5 setting to the Micromill 2000 Denstep > Communications > Control Type 20 Originally the control type was set to 0 and I got access violation errors.
The only other thing in that section was the Mint File, In VR5 it is listed as Micromill 2000 DENSTEP.mnt. In VR2 it was Mint File Micromill 2000.mnt.

chuck.stl

(202.03 KiB) Downloaded 1008 times

Cheers

Pete

[Steve]

Pete,

A few things.

If you set the raster path you can set the raster angle ie set 0 degrees or 90 degrees. Once you have the cutter path simply create another raster path with an angle 90 degrees different from the first you then get the raster in X and Y.

Mechanically adjusting the machine and lubricating would help.

We would reliably run at 600mm Min. The new card would buffer the moves so that the axis does not start and stop with ramping for every move. It ramps up to speed and buffers all the moves and streams them without deceleration until it sees an axis reversal.

In the simulation if the tool touches the job at rapid or without the spindle running or exceeds the axis travel it shows red.

[TDIPower]

Ahh ok I see how the raster part works now.

Ive attached 3 pics of the boards in the back of the machine listed above. (I had to rotate them as the site prefers portrait to landscape)
These were all taken while a program was running, The 7 seg display would swap between displaying P and 8.
The door was not locked either.
I won't be able to look at anything until Thurs or Fri now but will get the info off the other machine and run the same programs through it as well.

The files from techsoft are literally the kids names using the techsoft single line font, we found you must deselected the arc command as it really messes up otherwise. Ive attached 2 files 'John', one as the font as we have been dong, the other exploded.

John.dxf

(24.12 KiB) Downloaded 953 times

John explode.dxf

(26.4 KiB) Downloaded 957 times

I looked on the main site about upgrading but couldn't see anything for the Micromill, would it be a return to you? Could email me details and prices so I can look at possible routes forward. Would these have any trade in value on something like a microrouter? I doubt anything could happen until next financial year as that's when budget allocation happens but If I can put it all in the department development plan now the ball is rolling.

IMAG0006[1].jpg (1.18 MiB) Viewed 28075 times

IMAG0007.jpg (1.42 MiB) Viewed 28075 times

IMAG0008.jpg (1.21 MiB) Viewed 28075 times

[Martin]

The Guardmaster is controlled by the zero speed output from the Sprint 400 spindle drive. It is the 2 white wires on the top right of the drive card. It should control one of the relays next to the fan & outputs 24 volts DC so you should be able to see if the relay is changing over & measure the voltage.

[TDIPower]

Cheers Martin.

I hope that the guard is something simple like crappy contact. I will get to investigate on Thursday I hope.


Pete

[TDIPower]

Bit of an update. There was a meeting on after work so I could stop past 6pm, Ran chuck3 file via the old laptop again and the door was fine up until about line 6000. Then the door solenoid was firing open n shut almost all the time. Aborted the run and ran Chuck4 (the raster only program) The door would not lock at all. Couldn't understand how the heck the program code could mess up the lock as Martin had pointed out it is triggered by the Sprint 400 speed controller yet it had done the same with the raster program the other day too.

Whizzed the back off the machine while it was still running the program and grabbed my meter. Barely touched the terminal on the Sprint400 and the door relay fired. BINGO for this one... Dry joints on the connector on the PCB. So re-soldered and door lock is fine!!!!

Ill run some programs in the morning, I really hope this turns out to have been the fault and the arcing has been messing up the control. Ill need to check the other machine too now.

Pete