Advanced Search

> Sometimes stuff shifts mechanically

I verified that the PCB is working as expected. No welded contacts, etc.

> Is the PC you use to control the machine "frozen in time" (not connected to the internet or outside world)

It is connected to the internet, but there are no automatic updates happening. The software hasn't changed in at least 6 months. I also haven't touched the relevant HAL sections in as long.

1. Alan is right, not the same problem/issue
2. Where are the videos ?

I think in one of his videos, the Velocity Anti Dive is always active and THC is never active despite being enabled, hence my procedure.
BTW, that procedure should be in the docs, preferably a bit better explained and formatted, and should be mandatory for properly setting up the machine initially. This would save a lot of headache later in daily use, although definitely the Phils Auto Volt thingy does a pretty damn good job given the variety of hardware used and the amount of things that can be wrong and still cutting decently.
And so should be the Cutting fine tuning, again not very important but can be the difference between 3 minutes of cleaning and 30 minutes of cleaning parts. I am sure i wrote about it, but can not find it, and again, this would also need a bit of fines as i am still not good at explaining stuff.

> I do have the same problem
If your torch moves at all you have a different problem than me. I'd recommend opening a new thread and ask for help there.

OK, so pretty much even, we have Bears (no Grizzly though), Wolves, Foxes, wild Bore, Eagles, Falcons, Hawks, and one of the more venomous snakes with hollow fangs looks same as Rattle Snake but smaller nose and no rattle so you do not hear it coming ... or going... , rare scorpions but never seen bigger than 5cm, plenty of non lethal spiders.
Those are all native here, so about 4-5 snake bites per year (this goes over 15 when summers are really hot), 1-2 bear attacks per year rarely lethal as the biggest bear i have seen here is about the size of a human when standing up, and about 1 dead every year or two from wolves.
What we got as presents from UN personnel after the war:
Real Rattle snakes
The redback spiders (found one on my bedroom window about 15 years ago and wondered how it got here, so thank you Australia )
Some strange weasel like creature with a very angry face
Some mostly black birds that imitate my whistling perfectly, messed with me for years as whenever i heard them i thought my brother was calling me, always at night hence the long period to figure out.
Flesh eating bacteria that unfortunately is still present, about two months ago last case
Some type of giant wasp, survived for 3 years here due to mild winters but vanished after 2 strong winters, about 4-5cm in size, extremely aggressive, got bitten by it once, got a 5mm big red itchy bump that went away after half hour.
And we got probably more California Trout than California, plenty of restaurants have their own small fish farms, and plenty have been released in the rivers, so now we can very rarely find our native Trout.

Esaber wrote:

Right now the Z is moving up and down only when sensing the metal sheet and then when the cut is finished. And if the metal sheet is slightly curved the Z does not account for that and we end up with a welded torch tip !

It should see that and move the torch up. This is a sign you are not reading the torch voltage.

Assuming you are using a thcad, look at the encoder velocity in halshow while not cutting. Is it bouncing around near the 0 volt frequency. (/32) 

I guess if the amperage is too high for thin sheet metal, you may not see enough variation in voltage... Tommy will know more than m.

Looking at the 'probe_basic.ui' file makes me think that XYZBC configurations are not supported:
github.com/kcjengr/probe_basic/blob/main...robe_basic.ui#L31027

tommylight wrote:

I am not in Australia.
I do like Australia, with all the creatures that can kill people easily, mostly because of people.

Nothing to worry about, we don't have any man eating lions, tigers bears, wolves and similar carnivores. The only potential maneater is our snapping handbags, the salt water crocodiles. But they typically live in less populated regions to the north. The freshwater Crocs are not maneaters and their teeth are designed to catch fish...

So that leaves a few of the most deadly snakes in the world that fortunately have a very poor venom delivery mechanism (grooved fangs instead of hollow) and are typically timid unless cornered, two spiders (funnel-web and redback) and the blue ringed octopus mentioned... Nothing to worry about!

Following on form this thread.
forum.linuxcnc.org/9-installing-linuxcnc...build-boot-initrd-gz

Please do not run dist-upgrade or attempt to upgrade the kernel, unless you know exactly you are doing and are using the RPi kernel sources from the github page, and be aware some branches do not build or install the device tree files. The kernel was chosen carefully when the image was being built.

The kernel sources are specific to the RPi platform, from memory a Debian Arm kernel didn't allow for device tree overlays, which becomes an issue for enabling I2C and SPI to begin with.

If you find yourself in this position, back up you configs and reburn the image to the SD card, and copy across your configs. If you find a way to fix it by any other means great, but this is not a subject I wish to get involved in nor support.

Yep and they can be found via using google, probably too many to list. I wouldn't expect a "plug & play" solution going this route.

Before you go this route beware that the software stepping isn't optimal and it may be worth spending some extra dollars on an Mesa card or other such external step\direction control card.

Why did you a dist-upgrade ?

The kernel is a custom kernel (this has been explained in the first thread of the Rpi image thread) and does not come from the debian repos (the kernel sources are from the RPi github repo and contain patches specific to the RPi platform), from memory the debian kernel packages don't have the RPi patches nor support for dymanic device, eg spi, I2C and such.

Save your configs, reinstall and don't do a dist upgrade, there is no need to upgrade the kernel, if you do have stability related issues related to the kernel please make a report in the forum.

Not being to turn off the Rpi could be related to a device tree issue, eg the GPIO that is used is not setup correctly.

Try removing the LIMIT entries from the JOINT_n sections. That should set the value to +/- 1e99 which no machine will ever reach.


 
linuxcnc.org/docs/2.8/html/config/ini-co...t__lt_num_gt_section

 

I tried G10 L20 P0 A0

This doesn't work. It's resetting the work coords to 0 so the G-code works at first, but continues incrementing A0 indefinitely in machine coords. After about 10 runs, the G-code faulted mid-run, says the next command would move joint 3 past machine limits.

At that point, A has reached 9800 or so, ini limited A to 10,000. I can put in a larger number. but we're just kicking this can down the road.

Can I reset the machine coord value, or is there a way to say the A axis has no limits on how many times it can turn?

My Maho homes off the Heindenhain LS-403 scale index marks. Works the same as homing to a a separate homing switch.

Having ome slight problems here and was looking to see if someone could give me a hand.  I am finishing up a build and the last piece of the puzzle is the spindle and some of the outputs.

I am using a 7i96 that was reused from another machine and was set up for 4 step 1 pwm.  not sure the version of the firmware, but it should not need to be changed.  The computer is a clean install of 2.9 and the original configs from the previous machine doesn't work.  It was made using jt's 7i96 tool and then heavily modified.  so as of right now, I have most everything set up using mesaCT with some generic general values just to make sure that i can move and home each axis. so far so good.  

The spindle is 2.2kw 220v using a huanyang ec01 vfd which takes a 10v pwm signal so i have it ran through a small analog converter from the mesa board.  Now when i try and add the spindle to the config in mesaCT, all of the spindle controls are greyed out even though in options i have it set as 4 step 1 spindle and in axis, there is no spindle option right now. 

also, for some reason i can not get axis to turn on my relays either using mesaCT.  I have them set as mist/flood right now for testing, but i get nothing.  are these some bugs with mesaCT or am i just missing something obvious?  Thanks!