Advanced Search

Note that for Mesa Ethernet cards the latency test is not terribly useful.
What is much more important is network latency. Default Mesa driver settings
will drop (timeout) packets if the return data takes more than 80% of the
servo period. As long as you do not drop packets due to these timeouts, latency is
relatively unimportant/

if you have a mesa card, try this:

sudo chrt 99 ping -i .001 -q -c 60000 10.10.10.10

(this will run for 1 minute and print delay statistics)
 
Change the IP address if you use 192.168.1.121

 

Yes

Yep, that is 7I77 firmware.

latency does not matter for simulation

I made some time to look at this tonight.  I swear this worked the first time I tried it, but there's almost no way it could have ... so my apologies.

The reason it's not working is because you turn adaptive feed (M52) off:


Take out line N35 and it will work.  

Ah cool. I think this is a good example for the fact that you can get quality in China if you are willing to pay for it. Looks good.

I dont have a thread for my current project. It was a fast decision thing, because the price was low. Looks smaller on the picture than in real life. Weights 500kg (machine + table). Dont have the fancy pc table .

Old Berger Lahr 5 phase steppers. Drives invented 1987. Machine is build in 1992. I want to safe the old drives, because I like to maintain such old hardware. I already ordered some caps for replacement (known as fault reasons).

Machine is all cast iron! But with supported round rails instead of linear rails. But I will keep them. It will be sufficient for al lot of my work. Travel is 400x400x100. Machine is only 900x800xhight (didnt meassured it yet).

here is that, didnt think to look for a file, I am new to linux so the command didnt click till it wasnt doing anything lol.

got the Yaskawa encoder to work
www.youtube.com/shorts/YpUjbus9lQc

It's from LY CNC, which you can find on Alibaba. This is the 3040 model. I watched Stephan Gotteswinter's video about his new micro mill, and was inspired to do something similar. It was going to be difficult to get the same epoxy-granite frame he started with, here in the states. This cast-iron frame was a similar design, so I decided to try it out. There was a fair amount of waiting involved but the machine seems fairly sound.

With the spindle half-way down, I'm getting 0.000,5 in (13 microns) deflection with 20 lbf (88 N) pressing against the spindle nose. I think those are similar numbers to Gotteswinter's thumb test.

I got out the master square to check the perpendicularity of the X and Y axes. They're very square--within a few tenths (0.000,1 inch = 3 microns) over 10 inches (250 mm). The YZ plane leans back about 0.002" over the 7 inch travel of Z. In the XZ plane, there's 0.003" deviation over the same distance. I might eventually try to correct that but the parts I'm making are fairly small, so maybe it won't be a problem.

Do you have a thread about your retrofit?

Thanks! I'll give it a try tomorrow. I just got back from a wedding so I haven't tried your updated file yet.

Hello y'all!
I just received my new mesa 7i76eu, and I have a huge CNC with a durapulse GS10 vfd (until now it was runnnig with grblhal).

Does "pktUart" allow me to connect my GS10 vfd directly to the mesa without a usb dongle?

Regards!

Try removing the comments that are on the same line.
Also confirm that the right INI file is being looked at. More then one of us have edited the wrong version

Either change all the HAL pin names in the Hal files from qtdragon to inova or add -c qtdragon to the load line in the INI.

DISPLAY = qtvcp -d -c qtdragon inova

One of the safest way to have the probe and the tool touch sensor on the same hight would be to install a rigid piece of metal and facing it with an meassured tool. After that calibration of the probe with this new face.

Btw. Cool machine. Do you have somewhere more info about it? Right now I retrofitting a Kuhlmann P20. Looks quite similar.

Oops.  I'll have to add an interlock for the probe when it's in storage, on the right column. 

 

Getting the gage length of the probe isn't as easy as normal tooling.  It's a dual-contact spindle nose, so normally you can measure the length of tools on the surface plate.  But with the added variable of Z travel in the probe, you can't measure directly.  I ended up touching off on the 1-2-3 block and doing the guess-and-check on a gage-block stack to find the length.  A big gage block, adjustable parallels, and a micrometer would have been less dramatic but I didn't think of that until later.