Advanced Search

You can try using a USB WiFi dongle, but that is also "hit or miss", so it might work fine or might not work at all or might work intermittently.
I've had very good luck with USB to LAN adapters, so far out of over 10, all have worked fine for connecting to Mesa boards.
This also gives you 2 options for Mesa so you can test what works better for you.

Thanks all. Ground the PSU but this wasn't the cause

I should have known this but turns out the pi cannot use Ethernet and wifi reliably at the same time

I am pretty sure MesaCT does not overwrite them, but been quite some time since i used it.
-
In case you want to prevent overwriting config files that are good and done with editing, you can use the file explorer and set the permissions to read only.

flyingbrick wrote:

 The 7I97T is powered by a Mean Well 5V PSU with no earth connection.

Must have earth connection, on all power supplies.

In general I lack experience in this stuff, ground loops especially are confusing. I didn't want to ground the 5v negative and create some other issue. According to what I could find online, the floating 5v is fairly normal (switch mode power supply).

Yes the case etc is ground, just 5v is floating.

Thanks pcw, will ground.

If you have the mind of an engineer, why don't you test the software's functionality instead of complaining you don't like the colour?
Its very comprehensive and the probing is first class.

No, once you get a working config stop using pncconf and edit by hand.

possibly but if you follow the updated how to sticky thread (1st post) , we don't use the igh repo anymore.

Is there a way to prevent the .hal file from being overwritten? If you manually edit the .hal file, everything gets overwritten by pncconf. Is there a setting to ensure that only the parts that were actually modified get overwritten in the .hal file?
Thanks,
Hans

So, is it best to ground the 5V and 24V negatives?

flyingbrick wrote:

... and I measure about 100VAC between 5V- and cabinet earth.Does anyone know what specifically causes CR7 to latch on like this?

The answer to your question is there in your question, 100VAC is way to much interference, so the grounding is non existent if that is happening inside a metal box.
Does the shop have proper grounding and a ground fault interrupter?
Is the machine grounded through the mains cable, properly on both sides?
Is the power supply grounded to machine/control box ground?

Just for safety, and noise I would ground the 5V common.

The Init LED can illuminate for power related issues (check the 5V at the 7I97T)
With power related issues, the Init LED typically blinks.

The other cause of the Init LED being illuminated is because the watchdog
has bitten. This usually indicates either a latency issue with  the host, some
general communication fault or the watchdog timeout set too low in the hal file.

I am configuring my CNC with a 7i97t, but I don't have that problem. I have a Mean Well 5V 10A power supply with a ground connection for the Mesa and the Raspberry Pi 5; the Mesa needs to be grounded near the RJ45 jack, where it is marked "frame ground."

 I'm having an intermittent issue with a Mesa 7I97T on a Raspberry Pi 5. Occasionally CR7 comes on and stays on until I reset the board.The 7I97T is powered by a Mean Well 5V PSU with no earth connection. The 5V negative is floating, and I measure about 100VAC between 5V- and cabinet earth.Does anyone know what specifically causes CR7 to latch on like this? Is there anything in HAL/HostMot2 I can log to find the cause?Also, should the 5V negative normally be bonded to cabinet earth, or left floating?
Thanks all!

tuxcnc wrote:

I simply have the mind of an engineer and I hate "artists" who spoil formally good designs to make them bad but nice...
What's the point of a small gray font on a gray background?
LinuxCNC is not a game that you can run in a dark room and then a dark interface actually makes sense.
LinuxCNC is a machine control that can be dangerous and there is no room for anything "nice".
It is no accident that the e-stop button has a different shape and color than the start button, so as not to look for it and not to confuse it with another one when the machine starts to live its own life...
 

All valid points, except the "nice" part, see new Mazak controls, they look like form a space ship. Did not use it yet, so no idea how usable it is.

tuxcnc wrote:

I know that someone will soon start explaining to me that the authors do it for free and I should be grateful to them.
 

OK, it is open source so you can modify it to however you like it, see bellow

tuxcnc wrote:

The rule is simple - if you do something, do it well, and then people will be grateful to you, even if they have to pay for your work...
 

"Do it well" for them might not always be well for you or me, and i am fine with it, as i can modify as i see fit, and be thankful i did not have to do the whole thing from 0.
-
Seeing as you are quite good at this, i wonder, would it have taken less time to change the colors instead of this?
I would really like to know, and i do not mean the time it took to make it work for you, i am aware that can be tricky sometimes with versions and stuff, and i had my fair share of numpy issues.
Thank you.