Messages

1

Brains Development / Re: Enable-signal programming?

« on: June 13, 2010, 11:59:58 AM »

Just to clarify I'm not a machine builder delivering a CNC-router to a public school.
I'm a automation student in a secondary school here in Norway. A group of last years students started a CNC-router build as their main project. This year I and two others have continued on this build.
We are by no means finished and next years students will continue on the build and continue to improve it.

The machine does know the difference between limit trip and estop by visual indicator lamps on the control panel. Also a secondary pole on each limit switch and is wired to the Grex to give indication.

You are probably right about the shunting. I would have to confirm this with Geckodrive however. If everything is ok next years students could wire in a shunting relay (probably needs a small resistor in series?).

I belive i have read that the Grex do not support soft limits. I realize that the Grex is old and abandoned technology, but that is what we have to work with (tight economy).
However i will try to setup soft limits on Wednesday to see if they work. If they do problem is solved

By all means i appreaciate your input, and as said shunting the 48V PSU should probably be done. 
However i as far as i can see the machine will stop immediately on an estop (by disabling drivers) and can not be restarted until human action is taken.

2

Brains Development / Re: Enable-signal programming?

« on: June 12, 2010, 10:09:05 PM »

I use Gecko G210 drivers. These drivers have a disable terminal. When the safety relay cuts it also cuts the control power to miniature relay placed on the same board as the stepper drivers.
When this miniature relay looses power it will give all G210's disable signal (ground disable terminal).
This is in another schematic that i have at school. I could post it on Wednesday.

Really i would have preferred that the drivers had an enable signal instead of a disable signal, but the relay is placed as close to the drivers as possible, with direct wiring.

The drivers are disabled to stop all movement immediately, then they loose 48VDC power.

Stepper safety function #1: Disable all drivers to stop movement.
Stepper "safety function" #2: Cut incoming power to PSU.
Avoid lost steps function": Disable Mach3

NOTE: If the situation is sever enought to need to ESTOP why are you worried about loosing position??(;-)

During an ESTOP this is no worry. The case is that the limit switches are also wired to the safety relay. If a limit switch is triggered and not manually overridden the relay cuts out just like an ESTOP.
The reason for limit switch cut off is not for "human protection" but rather "machine protection".

To meet that qualification you would need to open the incoming power supply and then SHUNT across the PSU to effectively drain down the caps to stop the machine dead in its tracks.

It is clearly stated in G210 manual that the drives will suffer damage if power is cut on DC side of PSU ("fast" powerloss).

Using the brains is like using the Macropump it is an endless loop so if you program a saftey function that depends on an input to function it is possible that if the right conditions exist it could auto restart with movement.

As said the machine can absolutely not automatically restart. To restart the G9D relay requires a impuls action (non latching) key switch and a impuls action (non latching) start switch to be actuated at the same time by a human on the control panel. This can only be done if all ESTOP switches are out and all limit switches are overridden or not actuated.

Really this brain i request is by no means a safety function. It is simply a function to avoid loosing steps if a limit switch is reached.

I have attached the datasheet for the G9D safety relay.

3

Brains Development / Re: Enable-signal programming?

« on: June 12, 2010, 11:51:06 AM »

The safety relay cuts control voltage to a 3kW AC3 24V-coil contactor that controls 230V to VFD and 48VDC PSU. Since it cuts the  230V side of the 48V PSU the PSU will deliver power to the drivers for probably 10 seconds after power has been cut.
The safety relay however also cuts power to another control relay that disables the geckodrivers and gives a signal to input #1 on the Grex controller. This miniature relay should preferably be a MBB (make before break), but i think the inputs should be fast enough so that no steps are lost. The only way to find out is to test with the Mach3 brain

As said the safety relay also cuts power to the VFD. The VFD has been programmed so that it will brake the spindle with maximum level of DC injection if safety relay K1 is out, even if power is lost.

To reset the safety relay all EMSTOP-switches has to be reset. Then a non latching key switch has to be actuated the same time as the non latching start switch is actuated.
As you can see this is in no way a full auto reset 

Of course if Mach3 is running a program the program should be stopped so that cycle start has to be pushed after the reset. This is easy to do in brains.
However i can not figure out how to program a brain so that all actions including jogging is stopped in Mach3. Any suggestions? 

I'm attaching the schematics of the main controller board.

4

Brains Development / Re: Enable-signal programming?

« on: June 11, 2010, 08:41:48 AM »

Thank you for your reply Amir,

That brain would stop the G-code running. But i also want all jogging to stop (if not running code), and prefferably all "button presses" inhibited while input is low.
This is a machine to be used at a school so it needs to be "fool-proof".

The relay cuts power to the input side of the PSU, so the capacitance in the PSU should be enough to drive the steppers the time it takes for MACH3 to stop movement.

5

Brains Development / Enable-signal programming?

« on: June 09, 2010, 04:28:20 AM »

Hi guys,

In my control cabinet i have a safety relay circuit. The relay is active when everything is OK to run.
The relay cuts  power to the stepper drivers when it is disabled.

I have a signal to my Grex motion controller from the safety relay. I want Mach 3 to stop all program translation when this signal is not present (so that position is not lost when powes supply cuts out).
I could use the ESTOP input, but then i would have to press the reset-button every time the signal goes on (the relay is reset externally).

Is there a Mach button i could use in brains for this? Any suggestions how this could be done?

6

General Mach Discussion / Re: Extremely fast rapid movement

« on: June 07, 2010, 12:27:44 PM »

We have milled some more today and i have discovered where the problem occur.

In VcarvePro i have set plunge rate to 60mm/min and feed rate to 1630mm/min
The following example is a rectangular profile. The "very fast feed" occurs when the tool plunge down and start feeding in Y+.

As you can see (bold text) the tool plunge down 1mm at 60mm/min.
At the next line it does an arc move at 1630mm/min.

I have discovered that when a line with a feed rate change is followed by another line with a feed rate change  the G100 goes crazy.

(Rectangular profile)
()
G00X41.215Y40.556Z5.000
G00Z2.000
G1Z-1.000F60.0
G2X38.715Y43.056I0.000J2.500F1630.0
G1Y165.502
G2X41.215Y168.002I2.500J0.000
G1X290.384
G2X292.884Y165.502I0.000J-2.500
0G1Y43.056
G2X290.384Y40.556I-2.500J0.000
G1X41.215
%

However, if i put in a very small pause between the lines it all works fine, example

(Rectangular profile)
()
G00X41.215Y40.556Z5.000
G00Z2.000
G1Z-1.000F60.0

G04 P.001

G2X38.715Y43.056I0.000J2.500F1630.0
G1Y165.502
G2X41.215Y168.002I2.500J0.000
G1X290.384
G2X292.884Y165.502I0.000J-2.500
0G1Y43.056
G2X290.384Y40.556I-2.500J0.000
G1X41.215
%

If i set the plunge rate higher (ex. 110mm/min) the speed / accel is less extreme when the problem occurs. It may seem that it is the difference between the two feed rates, or the relatively low plunge rate, that causes the problem to occur.

All this i find very strange 

It seems weird no other G100 users (that i can find) have reported this problem?
Anybody have a smart solution to fool the G100 to work in this situation?

7

General Mach Discussion / Re: Extremely fast rapid movement

« on: June 05, 2010, 08:00:04 PM »

Last years students bought the G100 when it was still in normal sales from Geckodrive.
If I'd known what i know today i would have thrown it away and used another solution. However the control cabinet is finished and re-doing it for another motion controller would require a lot of work.
I think the G100 is actually a really nice concept, but as you say it has a whole lot of problems.

I have tried searching for other people having this problem on the G100 but I haven't really found anyone. It would be sad if this is a problem that can not be resolved. 

8

General Mach Discussion / Extremely fast rapid movement

« on: June 05, 2010, 07:13:52 PM »

Hi guys,

We have built a router at school and we're just now to the point that we can start using it 
However i have discovered one problem with it when running certain programs.

Example:
When i run a drilling cycle it will go down and drill just fine, but when the drilling is done it rapids up at an extremely high rapid feedrate. This causes the machine to loose all the steps in Z axis movement. This of course snapped the drill right off when the machine started moving in X-Y direction

Another example:
When running a circular pocket created on Mach3 wizards the machine goes down to the center of the pocket and starts plunging in to the material. When the machine then starts to move in X-Y direction the feed is extremely high and a few steps are lost until the diameter of movement is starting to get higher.

- Motor tuning is set and works fine (these rapid movements by far exceed both speed and acceleration i motor tuning).
- There are no rapid FRO active.
- FRO is 100 %
- Machine jogs fine.
- This does not happen on all rapid feeding, so far just the examples above.

Machine information:
GREX motion controller
Dual X-axis steppers (A-axis is slaved to X-axis). A-axis is set to not angular in general config.

Unfortunately i don't have the G-code programs now, but i could post it on Monday when I'm back at school.
However i don't really think the problem is in the program. This happens programs created by both the Mach3 wizard and on VcarvePro.