SJDrummond's CNC Plasma Wiring Diagram

[SJDrummond]

Hi Acourtjester,

Thanks for attaching that document! It answered most of the questions!

So I think I’m right in thinking:
1 - Yes I do need a relay for the plasma start,
2 - No I don’t need a separate power supply because the 18v comes from the plasma itself, and goes through the relay.
3 - ENA+/- pins are correct in the diagram.
4 - have to admit I’m still a little unsure on this one…

[Plasma-art]

It's a good idea to use an external SSR instead of using the internal relay.
A major reason would be extra security when it comes to stray RF flowing back since an SSR is opticly isolated.
Just gives an extra layer of protection for the expensive controller.
Since you mention it uses relays internal you could choose either the 5v or the 24v to control that relays but do refer to the mesa manual for that.
Personally I would grab one of the Fotek SSR-10DD relays, useable with the 5 and 24v and overkill for firing the torch however its cheap.
I'll draw up a diagram later.

[acourtjester]

SSR are used for AC circuits they are triggered by a DC voltage. there are modules that have opto isolation for different voltage power like 5 volt and 12 volt and 24 volt, like this one can be triggered by a voltage or ground signal. I have used these for Ohmic sensor and torch fire circuits.
https://www.ebay.com/itm/253046084198?h ... R5zK4rqMYQ

[SJDrummond]

Thanks both for that information on SSR’s. I will be sure to pick a couple of those up, although I’ll probably look out for a DIN rail mount version as most of my electronics are already mounted on them. I think RS components sell one.

Here is my electronics up to now. Updated somewhat from my previous attempt - I finally gave in and accepted that I massively undersized my enclosure! Now it’s 50x50x20cm and I’ve still had to think carefully about placement!

High voltage runs down the top and left hand side only, whilst all lower voltage components and wires are kept towards the bottom right hand side. Hoping I won’t run into too many EMI issues this way.

E33F66F6-B310-4D9E-909E-921DD23E0B92.jpeg

[Rodw]

Sorry, not been on the forum for a long time.

I think stealing 5v power from the Mesa Stepgens for the encoder is OK. i've done it many times.
I would not use an external relay for the torch on signal. Reduce the failure points.The Mesa internal relays are robust enough. Just they are polarised so if wired the wrong way you will get the plasma torch starting on startup. Shut it down and swap the wires around and you'll be fine. I use the Mesa 7i76e on board spindle relay (similar to the 7i96 relay outputs) for my torch on and have never had an issue.
IF you do want to use a relay, I've had good results with solid state DC 6mm wide din rail relays. I was using one, but took it out in favour of the mesa spindle relay.

I would recommend you don't bother connecting the stepper drive ENA pins. These are actually a DISABLE signal. It might make sense to wire them to your estop output so an estop disables the drives. If your safety relay is dropping power to the motor power supply, the ENA is redundant.

[Plasma-art]

SSR are used for AC circuits they are triggered by a DC voltage.

The DC variants are getting used alot for DIY 3d printers these days, hence the price of them dropped dramatically and become more mainstream.
For switching the heated bed and enclosure heater.
Main reason to use an SSR is running the PID controlled side of things, a regular relay would wear out pretty fast in that setup.

[acourtjester]

I was not saying do not use SSR but that they are for AC switching power not DC, I use them in the controller I have built for monitor and power supply power that is controlled by selected circuits. I use the USB +5 volts to turn on a SSR for monitor power. and have a separate switch to trigger a SSR for the motor power supplies input. When using the table for a router I have another SSR that powers the router and Vacuum system inputs. They are great for removing the contacts of a relay for larger power inputs. Another use is power for the heating elements in my powder coating oven, being switched by the PID temp controller.

[Plasma-art]

I was not saying do not use SSR but that they are for AC switching power not DC.

It depends on the series what type of input or output you can use, there are literally a couple hundred different combinations possible.
The one I mentioned is specifically designed for DC - DC for example.

[acourtjester]

Ok it look like the designator in the model number tell if its AC or DC. It if is as you show DD DC trigger and DC output voltage type, and the DA is for DC trigger and AC output voltage type.

[Rodw]

@Plasma-Art, that is a terrible choice for torch activation. Totally unsuited. Its these kind of relays that gives SSR's a bad name for plasma.
You are only switching a few milliamps. Something like this (in a pack of 5)
https://au.rs-online.com/web/p/solid-st ... ys/8886843
This is what I use.

[Plasma-art]

@Plasma-Art, that is a terrible choice for torch activation. Totally unsuited. Its these kind of relays that gives SSR's a bad name for plasma.

Are they that bad??
I am using a whole lot of them for different applications and have had no issues yet. (bought 25 of them due to reasons so have a bunch in stock)

Do agree that the one you mentioned would be a more suitable option for the task of only switching a few mA with practicly no volts.
Especially when there is already a din rail present to mount it on.

[Rodw]

@Plasma-art, there was somebody on the linuxcnc forum who said that you had to use a mechanical relay to satisfy the Hypertherm dry contact requirement. (This is totally wrong BTW, dry contacts can be SSR's). The reason he said this is he used a inappropriate SSR and had problems with reliability. So therefore, he strongly advocated use of mechanical relays and avoid SSRs whether on board or external. When I looked at his specs, it was just wrong for the application. The SSRs you have are great to switch heavy loads like AC coolant pumps but not low voltage DC.

Mind you I am not immune to stupidity either having bought AC SSR's without looking at the specs and finding they don't switch DC!

Building a controller from scratch as you do with linuxcnc, there are so many options and descisions to make. I started when nobody knew anything about plasma so had some pretty brutal learnings so I try hard to stop people from making poor decisions as you end up going down a blind alley, and have no idea how to get back in track.

[SJDrummond]

Hi everyone,

Thank you all for you input on this and apologies for the radio silence on this. I have been working away in the background and I am pleased to announce I finished wiring the main unit up (pictures to follow).

I have three out of the four motors moving with QTPlasmaC which is super exciting. Im having a little trouble with the tandem Y axis for some reason, I can move them both independently but they don’t seem to want to move together. Some reading across the forums is leading me to think I the need to home all the axis before LinuxCNC sees Y1 & Y2 it as a joint axis? Seems strange to me, but I will keep you updated.

Anyway, here is the latest wiring diagram.
Again, please note this is version 0.6 and still in development. Once I have a fully working system I will promote the diagram to V1.0. Although Im hoping its not too far wrong now as my box is currently all wired up like this!! No explosions or fires yet…

All the best,
Sean

[SJDrummond]

Wiring Diagram.png

CNC Plasma Wiring Diagram V0.6.pdf

[kn612]

Thanks for the update and diagram! Did you set y2 as tandem in pncconf?

My progress on a diy 7i96s/qtplasmac controller is the same as yours. Been referencing your diagrams quite a bit. Before there is any motion on the y axis I have to click home in qtplasmac. I'm not running any limit switches on any axis yet so our pnc configurations are probably different but the tandem setting is the only thing I can think that's holding you up.

If you have gotten that far, how did you set up a 2nd thcad in pncconf?

[SJDrummond]

Glad they’ve been of use to somebody!

Yes I set Y2 as tandem. Fingers crossed it is simply the homing sequence that needs to be done first. The limit switches are my next connections to make so I will be able to find out over the next few days!

I haven’t quite got that far with the second THCAD, but I know you have to flash a different firmware to the mess board, then select the ‘dpl’ firmware option in PNCCONF. It’s available on the Mesa website > 7i96s product page > support software section. It’s the bit file you want as far as I’m aware. This enables the second and third encoder options in PNCConf.

[SJDrummond]

Photos of the build

CB1BF3D6-8117-4DDC-8953-2B462F58A961.jpeg

7F59ED63-F623-40AA-AA68-3BD4D0345452.jpeg

00212AB2-F0CF-454A-A1E3-41A0F4270E18.jpeg

[acourtjester]

Nice looking setup. Not sure if this will help I am using a 7I76E card with THCAD-10 for THC but did not use aTHCAD-5 for ohmic sensing. I used a small other module for the the Ohmic sensing much easier to wire and setup as the probe input. Also when setting up your home/limits X and Y home toward positive and Z homes toward negative. And you need to put a small distance for the axis to back away from the switches. X and Y move in a positive direction and the Z moves in a negative direction. Like 0.25" of 0.50". The home latch direction is opposite for all.

[Rodw]

Looking great!.
The single most important (and confusing) Linuxcnc document to read is homing configuraation http://linuxcnc.org/docs/2.9/html/confi ... oming.html

A gantry machine must have homing switches. Usually they are installed at the min travel end (eg. at X0 and Y0 etc)
Homing switches must stay triggered to the end of travel.
Homing switches can also be used as a shared home/limit switch but a seperate limit switch is supported
A limit switch at the max travel end is optional as they are located outside of the soft limits defined in the .ini file and should never get triggered.
If homing to a shared home/limit switch, its important to use a HOME_OFFSET to move off the home switch otherwise it will immediately trigger a limit error when it comes out of homing. (home and limits are shared right)

Linuxcnc seperates motors (joints) from axes. Some machines might not have any axes. (max 9 axes, 32 joints or motors) Axes are defined from the joints when the machine is homed. The order axes is homed is set by the HOMING_SEQUENCE in the ini file starting at 0. You want the Z axis homed first to get it over any rubbish left on the table, so number the rest of X =1, Y=Y2=2
Joint axes, (eg the gantry) are joined together by placing a negative sign in front of the HOME_SEQUENCE so Y = -2, Y2 = -2
This negative number tells Linuxcnc to treat these joints as if it is one axis.
If a joint goes the wrong way, you can either swap the phase wiring around or just simply change the sign of the STEP_SCALE in the ini file.
To square the gantry, you use HOME_OFFSET. If there is a 2mm error from side to side, just change the HOME_OFFSET so they are 2mm different.

It goes without saying at this stage its time to throw away the PCNCONF or STEPCONF configurators and use a text editor. I use Geany (sudo apt install geany).

Sean, send me an email when you are ready to get the ohmic probing ready and I'll share the config you need which will make it super simple to set up.

[SJDrummond]

Thank you Tom & Rod for that info on limit switches and homing, it will definitely save me a ton of time instead of trying to figure all that out!

I think with the 7i96s’s limited input pins I might just stick to shared limits, i.e X-Min/X-Home, Y1-Min/Y1-Home, Y2-Min/Y2-Home, Z-Min(or is it max?). Then use the soft limits for the max as you mentioned.

I will need to do some reading around the text editor before I delve too far into it. (I have next to no experience with Linux coding), so that will certainly be a learning curve. Thanks Rod for the offer of help with getting the ohmic probing set up, I really appreciate that and will take you up on that once I get there.

PS here are some images of the machine itself incase anybody is interested. Quite different than a lot of other builds I've seen. Still a lot of mechanical improvements to be made here but at least the electronics side will be sorted!

All the best,
Sean

IMG_9337.jpeg

IMG_0799.JPEG

IMG_0798.JPEG

IMG_0804.JPEG

IMG_9338.jpeg

[asuratman]

Did you experience any interference ? Is that hypertherm powermax 45 is non hf with a blowback torch do not make any interference ? I have a lot of interference with cheap cut50 with water switch ohmic sensor.

[acourtjester]

Nice looking table it will get the job done for, very handy tool to have in the shop

[SJDrummond]

Hi Asuratman,

Yes it’s a Hypertherm PM45xp and yes it does feature a blowback start torch, so doesn’t produce anywhere near as much interference as cutters like the Cut50L. There’s a video on YouTube by DrDFlo, where he makes a plasma cutter, and ends up swapping the Cut50 for interference reasons. I think it can be made to work okay by using good wiring techniques I.E using shielded wire, ferrites etc.

Have a look here see if any of this is useful, this is what I’ve followed to build my electronics:
https://linuxcnc.org/docs/html/integrator/wiring.html

All the best,
Sean

[SJDrummond]

Hi again,

I’m having some intermittent issues with my electronics. I’m getting 0.4V when I connect a multi meter across Earth and DC Ground, or DC Positive. It happens across any of the power supplies in the unit, lots of the IO pins on the Mesa board. Really struggling to isolate the issue. The only symptom it is causing at the moment is with the ohmic sensor. When touching the metal and clicking the ohmic test button in QTPlasmaC, the LED flashes slowly instead of being fully on.

When using continuity mode on the multimeter across the same pins, I am getting an inconsistent series of beeps.

Any ideas where to start?

I have tried disconnecting everything but struggled to isolate the fault to anywhere in particular.

Any suggestions would be greatly appreciated.

Thanks in advance!

[Rodw]

If its noise, its likely to be entering the control panel via the mains ground. Be sure you have a EMI filter on the Mains input.

If you have wires connected to the THCAD ground on the input side, please remove them

Monitor the ohmic voltage (and the torch voltage) in halshow (even at 0 volts). If a ground is connected, the mains noise can introduce a 50 or 60 Hz ripple from the mains which you can see in halshow (once you understand the scales).

I have some hard won experience here... I ended up buying a USB oscilloscope to see where the noise was coming from. The EMI/RFI filter solved most of it. I