Advanced Search

Hey thanks for the fast reply. I ended up powering the 7i77 with a dedicated 5v power supply. Flipped all the appropriate jumpers on the 7i77 and 6i25

i would also like to point out that I removed the terminal blocks for ALL the encoders, drivers on the previously mentioned TB…so it can’t be shorted or drawing too much power because everything is disconnected! 

I started to wonder if my 6i25 was dead. I don’t see any LEDs light up on it when I power up the PC. I also can’t find it with lspci 

yeah some video would be very helpful. so your probe length changing is not really gonna help your cause. it should be fine for x and y, but your z is gonna be all over the place. you will be better off using a non repeatable holder touchoff routine which will basically use the height between the tool setter trigger position and your finished spoilboard. and make z0 the spoilboard surface for your programs. then every tool touch off will reset the work offset based on a calculation between tool setter and spoilboard and new tool touch. other ways to do it also but im not really a router guy so maybe better posting a thread asking for router tool touchoff help from the other users doing that already.

Signaling usually from 100Hz to 1000Hz, some up to 100KHz, and very rarely up to 1MHz.
There is no recipe for it, depends on a lot of stuff, so not easily framed into an answer, but:
-switching power supplies from 15KHz (older) to 400KHz (new ones)
-CPU inside a PC nowadays 2-6GHz, but almost all use a 100MHz clock
-power lines in USA 60Hz, in Europe 50Hz
-Mesa boards can do up to 10MHz, but most i ever needed was 0.96Mhz or 960KHz
-etc etc

I've spent the last few days working on the backplot and simulation and just about have it sorted i think. but this will be in the pyside6 release only which is a few weeks out from being primetime ready after testing.

Here is simultaneous impeller simulation on youtube:



 

What range of frequencies are usually used by cnc machines?

This is a PrintNC setup so router style setup with ER20 collet spindle. Workflow wise, I was using a 3d probe for workpiece locating and a tool setter for determining the tool offset each time a tool is loaded. Like you mentioned, since this is a collet setup I touch off each time the tool is loaded. The Z=-7.4953 value is a G54 value that I assume came from the tool offset calculation, but I may be wrong. I can go try to capture some video. It very well may be operator error.

I appreciate all the help.

ok, i'm a bit confused, maybe something is wrong but mostly what i see looks ok. you said your z range is Z0 to Z-6.3, but i see your earlier post referencing Z=-7.4953 something doesn't seem to be adding up.

is this a bt30 or is30/20 machine with repeatable holders? or is it an er collet type machine?

i'm trying to figure out where the issue is. maybe a detail written walk through r a video would be easier, i don't see anything wrong off the top of my head but without seeing dro's at time of command or active work offsets etc. triggering a limit isn't hard to do. especially if you have an offset set super close to min or max travel etc.

im not getting any of that behavior nor other reports so i have to assume user error, not to be taken as picking on you, just trying to identify where the issue becomes the issue.

if you are suing non repeatable holders, it may be better to use that type of tool touch off. it basically resets the work coordinate system for each tool so it cuts at the same height as the previous tool without storing tool offsets in the table. this tends to work better for er collet type machines and i think there is a really good touchoff routine by another member on here maybe tootall18 from memory? but that is only if you are an er collet. if you have repeatable holders like bt or iso style then the stock touch off is the correct one to use.

grossm5000 wrote:

Anybody have any ideas??

Yes:

PCW wrote:

I would check the 7I77s 5V supply, my guess is that it droops
when you connect the MPGs, either from a wiring error or
too much current.

If you have the encoder power wires reversed, that will for sure make the voltage drop and cause the failure.
But, some parallel cables have so thin wires inside that might also be the reason for the failure.
My advice:
-set Mesa boards for each to have it's own power supply
-find a good quality USB cable and cut the USB-whatever connector leaving only USB-A side and the wire, wire it to Mesa 7i77 as the power supply
-do NOT do this before checking the wiring and making sure it is correct.

I have 5v power from the computer. I unplugged ALL the terminals on the 7i77 TB3,4,5,7,8) reinstalled LinuxCNC (worried about kernel issues) and the darn thing still dumps me when I launch the program. 

I feel like I’m getting to the limit of the things I can do to make this work. Any help is welcomed. Anybody have any ideas??

aikiaviator wrote:

I read somewhere about a need to set the config of the THCAD2 into the settings of the Software based on a frequency... Cant remember where I saw that but the config setting are my main concern at the moment. 
 

The thcad is a voltage to frequency converter. It has individual calibration frequencies at 0 volts and max scale volts (10v or 5v) on a sticker on the board. you enter this data into the QTplasmac interface. Read the docs and  it is well explained.

After another glance, 17.3MHz !!!
That is the realm of radio frequencies, not something you find on a machine.... usually.

Hi yes, I have a cut chart for the cutter. BUT it does not specify a variable air pressure chat such as the one you have presented. So I think this could be useful. Thanks.

Separately, I was concerned about the configuration settings: mostly about the arc voltage etc... I read somewhere about a need to set the config of the THCAD2 into the settings of the Software based on a frequency... Cant remember where I saw that but the config setting are my main concern at the moment. The other thickness settings are in line with the cut charts for the machine.

That being said, I have literally never used a cutter, so have zero experience on how this should be set up. I live in Melb, AU. So if anyone local, would put on beers coffee or whatever for some localised help. 

Thanks Tommylight.
-Set up the THC, any fine tuning without it is useless when THC is active.
Ok. Will note

-25A for 5mm thick is way to low, roughly 10A/mm is OK
Yes this is already set to 40A in the 5mm profile

-with working THC, you need to fine tune air pressure (exclusively while the air is ON), voltage=cut height, feed rate
Ok. This I have definitely NOT looked at. Will do. Thanks.

-cut quality will vary based on air moisture, material being cut, nozzle type and wear, so use new nozzle for fine tuning and after you have things working properly
I have a three stage Breathing Dryer rig from when painting, + a standard dryer, so dry air should not be a problem. But the rest of it I will need to work through. Thanks.

My main concern was the settings in the Configuration setting, specifically arc voltage setting etc... Any thoughts on this? Do you have some that you may be using?

Yup, that's noise. Your scope settings are so "zoomed in" in time and voltage you probably can't see the real signals.

• Set your scope's Y axis (voltage) range to more like 1V/division. More if the signals are (eg) 24V.
• Set a trigger on a rising or falling edge, with a threshold of something in the middle of your signal voltage range. Eg, if it's a 5V system, set the trigger at 3V or some such.
• Change your X axis (time) to much longer. This depends on how fast you're moving the motor, but 25 ns is way too fast to see the signals you probably care about. Something more in the 100 us or 1 ms range probably makes much more sense.

If you do the above, you'll probably see step/direction signals, or quadrature, or something. Either way, it's very likely something LinuxCNC supports "out of the box". Engineers are lazy and generally don't invent new protocols if they don't have to

Thank you for all of your help Peter, I have finished all of my testing now the conclusion is clear.