Setting up a new machine.

[regpye]

I am new to this board and have explored Sheetcam a bit and find it a nice program and easy to understand, no real problems with the software yet.
I am using Corel Draw to produce the DXF files, not problems there either.

I am building a new plasma machine and am having a few problems that I have not been able to sort out myself.
Have downloaded Sheetcam and got it running OK, and feel that I may be able to work out how to use the software well enough.
The machine will be using an Arduino nano, GRBL 1.1h, and UGCS. No THC at this stage, although I have designed the machine with a Z axis and also a floating head.
Running the softwares on the screen everything seems to go smoothly as it should, but this may be different in real life when the machine is actually using it.
I realize that this is not strictly a Sheetcam issue, it is really a GRBL/Arduino issue, but I have not been successful in finding any help elsewhere, so I hope I am not treading on any toes by asking here.

I have tried using the GRBL post processor and that appears to work fine on the screen. I also tried the GRBL post processor with THC and had a problem because there was no real probe switch, so I will try that processor again when the machine is complete.
The problems I am having are with the M3, M4 codes. I have a relay module connected to D13 on the nano for switching the plasma trigger circuit, but it starts in the switched on mode and I have to send a M4 to turn it off. This would be very dangerous with a torch connected and firing on startup.
In the post processor I had to make some changes as I cannot get M5 to work. M4 seems to be the only way I can turn the plasma off, so I modified the post processor by changing the 7 occurrences of M5 to M4 and it all seems to work fine on the screen.
I have also tried using D11, and for some reason it stopped working, maybe I have damaged the nano?? but while it was working I had the same issues.
I think the problem may be that the relay module is a low level type, and is on when starting up. I have ordered some high/low level relay modules to see if I can correct this problem that I have.
I also noticed that using the same relay module for the coolant, using A3 pin, the relay works in reverse to what it should, M9 to turn on and M8 to turn off, but in UGCS the buttons show the opposite (correct way) So I am thinking that maybe it is the same issue with both relays.
After getting this all sorted out, I am sure that I will also need some help in fine tuning Sheetcam as I am only guessing what is needed for the settings, no information on using my plasma came with it, so it is all hit or miss at present. The plasma cutter is a CUT50 with pilot arc (Chinese machine)

[regpye]

I have been trying to work out this problem because no-one else seems to have any idea about how to solve it.

So far I have determined that I should send a M4 code to GRBL on startup, this can be done by making a startup block $N0=M4
This puts the relay module off before it has even started. Problem is that there is a bit of time before the system is up and running, so I have put a delay relay in line with D13 on the nano and the signal pin on the relay module. This seems to work OK when trying it out on the desk.

As I am using Windows, I need to start Windows first, then run UGCS so that it is ready and then turn on the power for GRBL (nano board).
As GRBL initializes it will send a short pulse to the relay module, but as the delay timer has broken the line between the relay timer and GRBL nothing happens. The M4 code sets the relay module off. After the predetermined time (I found it only needs to be a few seconds) the relay module is then connected to GRBL and all is ready for operation, no false triggering of the plasma torch.
I had hoped that there would be a simpler way of doing this as now there are three boards in use, the nano board, delay relay board, and the relay module, just to turn on/off a simple switch.
Has anyone used a better method, one that uses less resources and is more straight forward?

[djreiswig]

It sounds like you need to implement a Charge Pump. It is a signal that is sent from the control software telling the controller that everything is started up and it's safe to power up.
Does your software have this sort of signal? If it does, you could drive a relay that cuts power to the outputs until it's safe.

[regpye]

It sounds like you need to implement a Charge Pump. It is a signal that is sent from the control software telling the controller that everything is started up and it's safe to power up.
Does your software have this sort of signal? If it does, you could drive a relay that cuts power to the outputs until it's safe.

Thanks for the reply djreiswig
I have not heard of this, and am not aware of in any of the software that I am using.
Software I will be using will be GBRL 1.1h, Universal G-code sender, and sheetcam, all running under Windows 10.

I thought there must be others that are using GRBL and UGCS and are triggering the plasma torch though sheetcam, how are they going about this function?
Using D13 on the nano, I need M3 to fire the plasma and M4 to stop the plasma. I have managed to modify the post processor to do that and it works fine on the screen and the dummy setup I have on the desk. The relay module is a low trigger type, so I have ordered a high trigger type to test out and hope that I can stop using the delay relay. I have tried using pull ups and pull downs with no success.

[djreiswig]

It looks like there isn't a charge pump on that setup.
Maybe this will give you some ideas.
https://www.reddit.com/r/CNC/comments/5 ... h_grbl_to/?

[Les Newell]

The best solution is to change your relay modules to high level type rather than low level. Everything would then run as expected.

[regpye]

Thank you Les,
I have some high level modules on order as I suspected that was the case. Will try them out as soon as they arrive.

[regpye]

The best solution is to change your relay modules to high level type rather than low level. Everything would then run as expected.

The high/low relay modules came today and I set both relays to high trigger mode.
Both relays, plasma trigger and coolant are now working as they should, no delay module needed.
Also M5 is also working, so I changed the post processor to M4 on and M5 off. I also added M8 to the M4 and M9 to the M5 so that the second relay can turn on and off the water coolant at the same time as the cutting. I am trying this experiment to see if I can spray the work piece with water while cutting to minimize any distortion. All is working on the desk setup.
Thanks for all that tried to help me, much appreciated.