The Issue with 5I25s (and 6I25s etc) is that the Xilinx Spartan 6 FPGA is not available
(and may not be available for quite some time if ever)
We have a replacement card coming out but it will...
Type: Posts; User: PCW_MESA
The Issue with 5I25s (and 6I25s etc) is that the Xilinx Spartan 6 FPGA is not available
(and may not be available for quite some time if ever)
We have a replacement card coming out but it will...
Exactly, these were Cymer US patents that ASML acquired so your "Not according to ASML they say there is no US patents used in there machine only in the equipment that is used to make their machine...
Nope 28 nm is heading for over supply.
The center of supply line issues is actually closer to 100 nm
Many Cymer US patents are used for the EUV light source in every ASML EUV machine...
Way to miss the point...
The current chip shortage is chiefly in 45 nm and older processes
ASML uses hundreds of US patents just for the EUV light source they acquired from Cymer (San Diego)
Certainly more diversity is a good thing, but the "eggs" dollar wise are mainly in high density chips which are not made in China.
Actually Chinese chip capacity for low tech legacy chips is...
Yep, Add FPGAs, optocouplers, almost all switching power supply chips, almost all Ethernet chips and even lowly Zener diodes... Its a real supply chain nightmare.
I don't see any evidence that it...
Not sure why you think there is editing of pins required
Stepgens and pwmgens are not addressed by pins and the GPIO pin numbers are the same for any configuration
The encoder in no way interferes with the use of those pins for home switches nor requires any hal changes
Not really, if you use the correct bitfile (A bitfile for a G540) the pinout is correct and
you have the proper resources for the G540: PWM and chargepump support
Since this is not a Tormach...
I think its valuable to divide jerk control into tangential jerk and centripetal jerk:
Tangential jerk (along the tool path) is full controllable by the machine.
Centripetal jerk is totally...
Pretty sure Tormach uses the 10.10.10.10 IP address so IP jumpers need to be
W5 DOWN, W6 UP
But beware, Path pilot will write it's bitfile to the 7I92, which is not what you want for a G540...
The G540x2 bitfile includes 2 encoders (one on P2 and one on P1):
(Though you would normally have some signal conditioning like a line
receiver or buffer for the encoder signals rather than use...
As long as you switch the AC side you should be fine.
In general its not a good idea to switch DC supplies
to electronic equipment with mechanical switches
due to the large surge currents...
If the driver fails to reverse you could get this issue.
Normal homing would
1. Move to the limit switch until it trips and then slow to a stop
2. Reverse and back-off the limit switch slowly...
Using a 7I92H would give you a PathPilot supported, Ethernet connected device that will connect directly to the Tormachs
breakout with a short 26 pin flat cable.
Note that if you are using a DPDT relay to swap the breakout board analog output pins, you may have problems if the breakout analog output
is not isolated (since you will be connecting the drives...
The difference between the 7I76E and 7I76ED is the outputs _not_ the inputs
The 7I76E has sourcing outputs (PNP type)
The 7I76ED has sinking outputs (NPN type)
The inputs on both are sinking so...
That's a very strange error especially since the following error limits are so wide
I recall this from the LinuxCNC forum. I could not duplicate your issue on my test
machine using your hal/ini...
The green trace looks like A and B (or /A and /B) are shorted together
The Gecko drive could not change an arc to a circle, that's more likely a controller or Gcode fault
LinuxCNC and EMC were derived from public domain CNC software from NIST:
https://www.nist.gov/publications/enhanced-machine-controller-emc-open-architecture-controller-machine-tools
Not from...
C79 was never installed
Another test option would be to replace the coolant pump with another load (say a desk lamp)
If the lamp works, the most likely culprit is the coolant pump motor. Its _very_ unlikely that
a relay...
Yes, with only one pulse per turn, the controller would have to estimate the spindle angle, this would get worse with larger taps
and larger load changes generating speed variations that the...
Have you considered using LinuxCNC+PlasmaC?
This would allow re-using your existing drives and having full feedback to the controller
(which you lose if you go to a step/dir system)
Different Ethernet cards will work with LinuxCNC and Mach but I am not aware of any single card that will work with both.
Unless you have very old x86 CPU, 64 bit is supported
(the comparatively ancient Core Duo is 64 bit)
That may have been parallel port interfaced originally but it has been upgraded with a Ethernet
Smoothstepper (the piggy-back board on top of he PMDX-126) so AFAIK it should work with modern
64...