RPI4 Raspbian 64 bit & LinuxCNC

More
3 years 4 months ago #273300 by drummond
Oh, I've got chromium installed it just won't run from a double click on the desktop.  It has to be run from a terminal and then you get a message from the Google gods that "this isn't awfully secure, but if you really want to risk life and limb to get online we can put you in sandbox mode"  if you answer "yes I'm an idiot" a chromium browser opens.

I really only need it to email stuff to myself because, once you put all your stuff on an SSD because it does everything *so much faster* then you have a wireless keyboard and mouse, the Pi runs out of juice on the USB bus and can't even run a stick in addition.

Everybody's struggling with this:

raspberrypi.stackexchange.com/questions/...irefox-on-rpi-zero-w

Please Log in or Create an account to join the conversation.

More
3 years 4 months ago - 3 years 4 months ago #273531 by MrRekt
I'm running Firefox ESR without any issue or hangup to speak of. Perhaps you can give that a go - but if you've already got a working solution, I might be a bit late with the suggestion.

- edit -
If you're struggling with powering it off USB (if I'm understanding correctly), I've heard that the Pi4 can power-spike to nearly 20 watts iirc - so far I haven't had any issue with the official USB power supply - but I can check back in when I start powering it via the 7c81 - I have a dedicated 5V/3A power supply ready for it, so I'm not really anticipating any issues.
 
Last edit: 3 years 4 months ago by MrRekt.

Please Log in or Create an account to join the conversation.

More
3 years 4 months ago #273545 by drummond
Yes, I'm running the pi through the 7c81 using an official 5V3A power supply.  I believe the max the 7c81 will take is 5A, but I don't want to go there yet.  I might try Firefox since you're having luck with it.

Thank you for the suggestion.

Please Log in or Create an account to join the conversation.

More
3 years 4 months ago - 3 years 4 months ago #273619 by MrRekt
Good news / bad news
Good news is that the image I downloaded from this thread seems to be playing nicely with the 7c81.
Code:
sudo mesaflash --device 7c81 --addr /dev/spidev0.0 --spi --readhmid gives:
Code:
unable to set bpw32, fallback to bpw8 Configuration Name: HOSTMOT2 General configuration information:   BoardName : MESA7C81   FPGA Size: 9 KGates   FPGA Pins: 144   Number of IO Ports: 3   Width of one I/O port: 19 ...
Bad news is that I bricked my 4GB Pi by giving it a poorly made cable while attaching it to the 7c81. There's a little more to the story than that, but I digress.

The Pi I'm having success with is a 2GB version which I mutilated during a previous project - but it's working well enough to SPI to the 7c81 - and I was able to make it work without a ribbon-cable (because it seems I can't be trusted to make those):
Last edit: 3 years 4 months ago by MrRekt. Reason: typo; weird nested code tags
The following user(s) said Thank You: cakeslob

Please Log in or Create an account to join the conversation.

More
3 years 4 months ago #273621 by Aaroncnc
There is a protection diode that dies if you hook up the gpio wrong. It heats up by the psu connector. If you remove it you may find your pi works again. But don't play with the gpio till you replace that diode.
Not always will it save you depending on how bad you messed the gpio up but it has saved me a few times.
The following user(s) said Thank You: MrRekt

Please Log in or Create an account to join the conversation.

More
3 years 4 months ago #274415 by MrRekt
I was having odd performance issues while trying to improve latency in different scenarios.

Ultimately, I was able to conclude that it was a matter of the Pi thermal-throttling the CPU.I found that overclocking the CPU to 2.0GHz had a small improvement in LinuxCNC's latency test, but would eventually overheat and ruin test results.


I ended up running a number of tests with a tool called "stressberry", which monopolizes all 4 cores with prime-finding and logs the internally-reported CPU temperatures. After seeing that reported temperatures were far higher than would be expected with the actively-cooled heatsink I have attached to the Pi, I swapped out the standoffs for the heatsink with ones that were about 1mm shorter to improve contact with the CPU.

This small change had a tremendous impact on the stressberry tests: 

This is the same test, run for the same amount of time, before and after.

I've been meaning to report back these finding, because I've been trying to get the fantastic results shown previously in this thread:
forum.linuxcnc.org/9-installing-linuxcnc...cnc?start=150#253623

There is a protection diode that dies if you hook up the gpio wrong. It heats up by the psu connector. If you remove it you may find your pi works again.
 
I don't suppose you have a link to something I can reference regarding this?

Please Log in or Create an account to join the conversation.

Time to create page: 0.103 seconds
Powered by Kunena Forum