Same firmware, but I do have a different bootloader.
taylor@printerpi:~ $ vcgencmd version
Aug 31 2021 14:53:42
Copyright (c) 2012 Broadcom
version 67615e950e1e28b92dfae6303cf7a8b879a8908f (clean) (release) (start_x)
taylor@printerpi:~ $ vcgencmd bootloader_version
Jul 6 2021 11:44:53
version c258ef8fe1d2334a750078b17dab5e2c1a1787fc (release)
timestamp 1625568293
update-time 1631380726
capabilities 0x0000007f
Going to update bootloader to match yours and re-test
bootloader didn't make a difference, kind of expected that but was worth a shot... I'm running out of idea's
Blobs of that sort may indicate pauses in printing, which can easily be caused by something that interferes with the speed of data transfer between the server and the printer so the printer is waiting for another command.
On my daily driver it happened frequently when the display was going to sleep/blanking and has been eliminated by adding the Pi with octoprint to control the server. My Pi is booted with an hdmi monitor attached (won't boot without it), but then has the monitor removed and is controlled by vnc remotely.
It can also happen if there are delays and resends between the octoprint server and the printer.
Are there any errors shown in the Resend Ratio? My current print is showing 0 / 2668.4K which means no data resends have happened. I also run at 57600 baud.
@Ghost no resend ratio issues, both kernels show 0. I agree with what your saying, just nothing pointing to what in the new kernel is the issue. I've literally swapped everything, pi 4, USB cable (including brand change), with screen, without screen, fresh install on two different brand sdcards. All result in the same blob problem with the 8/31/2021+ kernels. So far the only common thing between those not having an issue and the one other person besides me that is, is the Creality board 4.3.x, which doesn't make sense, it's just a common USB connection.
You brought up baud rate which I didn't consider since I'm running 250000, I'm going to test when I can with lowering the rate with the new kernel and see if that changes anything.
There is one other thing I had to find proper adjustment for with trial & error.
My printer is a delta, HE3D K280 and not cartesian like the ender so movement is different.
I was having several issues with print quality as I worked through the proper settings and finally settled on a suitable layer height and temps as well as replacing the extruder that came with the unit.
With PLA the final temps that give excellent quality are 185 hot end, 50 bed. Speed is right now at 50 mm/s and layer height of 0.2. Using ABS the only difference is temps with 200 hot end and 90 bed.
The delta of course uses a bowden tube and quality and routing of that tube are also an issue. The extruder I finally settled on is a dual drive geared extruder as the other design was a problem in the torque and speed stability.
I also have to be very careful of filament quality and finally settled on mostly using sunlu filament as I was having repeated issues with nozzle plugging when using other brands.
There is one other thing that can cause hesitation and blobbing. Without errors in communication (you said the resend ratio was zero) that leaves movement. Anything that can cause slight hesitations in movement can contribute to the blobs. A tiny amount of slop in a drive (belt or unitversal or gearing) can cause hesitation with direction changes. A slight binding in guide rods can cause jerky travel. Even a slight wear on the drive screw rods can give some slack that shows up with direction changes. A tiny bit of movement in the frame can contribute as well, particularly if there is slight movement at structural joints.
In post 3 you noted that the blobs appeared in curved areas, which means direction & rate changes in X & Y. It might be possible to minimize the issue as well as see the actual cause if you do a test print at lower speeds.
@Ghost yeah I had tried slower speeds, nothing stood out that wasn't noticable at normal speeds, but again, same settings, everything is perfect when I use pre 8/31/2021 kernel
@notfixingit I don't have a screen attached to mine yet and I did try the new -3 kernel and had the same issue. Rolled back again for now. Sorry for the slow reply hun.
This problem is so unlikely yet the fact you are able to roll back & forth between the versions shows there's definitely some kind of problem here. There is something you can try to stress-test the connection, not sure if it will show any relevant data.
This would stream a local file to the SD card on the printer over USB/serial comm:
python -m octoprint.util.comm <port> <baudrate> <local path> <remote path>
Obviously that will need to be run either with OctoPrint's virtual environment activated, or using ~/oprint/bin/python or similar. Be warned, I don't know if that actually works or not, I haven't tested it.
20210928 kernel release from last week solves the problem, just finished my second print on it and no blobs, yay!