Custom Community Firmware for A5S & A1

Samuel Pinches

Apart from mesh bed levelling, the other main advantages are

• Reliable octoprint printing
• Ability to set and save print settings (EEPROM, good for extruder steps tuning, acceleration etc.
• More menu options
  

At the moment you give up

• Filament run out detection (shouldn't be hard to fix, but I haven't looked at it yet..)
  
• Power outage resume (probably quite a bit harder to fix, but I haven't looked at it yet...)
  

Robert

Filament run out detection is mostly useless unless you are there  when it happen,  I doubt anyone would miss it.

Power outage resume is equally useless unless the power resume like immediately,  once the bed get cold the print will pop up and you have to start it over.  I would not waste time on it. 

BTW,  one of the A5S I upgrade the firmware is having the LCD screen issue, I am waiting for the foil tape to arrive  so I can do the  trick on  your video to fix this issue.

Why didn't the stock firmware has any problem with the touch screen?

Samuel Pinches

I agree with your assessment - that's why they haven't been a priority. Filament run out can be a useful hack as an easy code-free-way to allow changing colour etc, but still low on the priority list.

Sorry, I really don't know why the stock firmware doesn't have problems - it could be a different SPI speed, or perhaps just a marginal signal. I tried a few things, but had no luck until I added the foil tape.

Robert

From what I can see the problem with the touch screen is only happen when the print starts, once the print is finish the touch screen return to normal.   So far only one the A5S is having this problem, there is still one more A5S still in stock firmware, I shall see what happen with the last one after I upgrade the firmware.  

NoFutureKid

Hi,

I tried to compile the firmware myself on Windows with VSC but it fails.

In file included from c:\users\thomas\.platformio\packages\toolchain-gccarmnoneeabi\arm-none-eabi\include\stdlib.h:19,
from c:\users\thomas\.platformio\packages\toolchain-gccarmnoneeabi\arm-none-eabi\include\c++\8.2.1\cstdlib:75,
from c:\users\thomas\.platformio\packages\toolchain-gccarmnoneeabi\arm-none-eabi\include\c++\8.2.1\stdlib.h:36,
from C:\users\thomas\.platformio\packages\framework-arduinoststm32\STM32F1\cores\maple/wirish.h:41,
from C:\users\thomas\.platformio\packages\framework-arduinoststm32\STM32F1\cores\maple/Arduino.h:30,
from Marlin\src\HAL\HAL_STM32F1\HAL_sdio_STM32F1.h:28,
from Marlin\src\HAL\HAL_STM32F1\HAL_sdio_STM32F1.cpp:26:
C:\users\thomas\.platformio\packages\framework-arduinoststm32\STM32F1\system\libmaple/stm32f1/include/series/gpio.h:489:15: error: duplicate 'inline'
static inline __always_inline void afio_exti_select(exti_num exti, exti_cfg port) {
^~~~~~~~~~~~~~~
C:\users\thomas\.platformio\packages\framework-arduinoststm32\STM32F1\system\libmaple\include/libmaple/nvic.h:112:15: error: duplicate 'inline'
static inline __always_inline void nvic_globalirq_enable() {
^~~~~~~~~~~~~~~
C:\users\thomas\.platformio\packages\framework-arduinoststm32\STM32F1\system\libmaple\include/libmaple/nvic.h:119:15: error: duplicate 'inline'
static inline __always_inline void nvic_globalirq_disable() {
^~~~~~~~~~~~~~~
In file included from c:\users\thomas\.platformio\packages\toolchain-gccarmnoneeabi\arm-none-eabi\include\stdlib.h:19,
from c:\users\thomas\.platformio\packages\toolchain-gccarmnoneeabi\arm-none-eabi\include\c++\8.2.1\cstdlib:75,
from c:\users\thomas\.platformio\packages\toolchain-gccarmnoneeabi\arm-none-eabi\include\c++\8.2.1\stdlib.h:36,
from C:\users\thomas\.platformio\packages\framework-arduinoststm32\STM32F1\cores\maple/wirish.h:41,
from C:\users\thomas\.platformio\packages\framework-arduinoststm32\STM32F1\cores\maple/Arduino.h:30,
from Marlin\src\HAL\HAL_STM32F1\../../inc/../HAL/HAL_STM32F1/HAL.h:41,
from Marlin\src\HAL\HAL_STM32F1\../../inc/MarlinConfig.h:30,
from Marlin\src\HAL\HAL_STM32F1\HAL_Servo_STM32F1.cpp:26:
C:\users\thomas\.platformio\packages\framework-arduinoststm32\STM32F1\system\libmaple/stm32f1/include/series/gpio.h:489:15: error: duplicate 'inline'
static inline __always_inline void afio_exti_select(exti_num exti, exti_cfg port) {

And many more of these error messages. Maybe you have an idea.

Samuel Pinches

Welcome @NoFutureKid !

Thank you for pointing this out - it looks like some changes in some of the libraries have broken things recently. You can download a working build here - with libraries included - that you can compile in VSC.

https://www.dropbox.com/s/m3nra4wcmiuhjzz/A5S_A1_Marlin2.0_latest_beta.zip?dl=0

I am short on time, but my next plan is to create a new A5S/A1 build from the latest Marlin 2.0 build, but I haven't been able to get around to that.

There's been some layer shifting problems that people have reported, and from what I have been following on the Marlin github it sounds like there may be some timing issues in 32bit that need to be fixed.

NoFutureKid

Sadly this did not work. I think the archive is missing all the libraries from your "platform" library and uses the automatically installed ones.

I also tried to uninstall all VSC and PlatformIO stuff (including manually deleting leftovers from my profile folder) and
reinstall VSC + PIO again. No luck.

Samuel Pinches

Drat, I was hoping that would have worked! sorry for wasting your time, I'll have to look into it further....

NoFutureKid

No problem. It was just a try, because I wanted to persist some stepper settings in firmware, instead of eeprom and rising the nozzle temperature a bit. I can only reach 235C and hoped to get up to 240 maybe 245C.

Samuel Pinches

This one should let you go up to 245C.

Simulanon

Could i put in a request for filament change out in the menus? I cannot seem to find anything

Samuel Pinches

@Simulanon - once I get the filament run out sensor working, that is a feature I definitely plan to add. However, for now you'll have to do it manually unfortunately. Thanks for the suggestion.

Samuel Pinches

New beta firmware for test - the goal of this one is to fix the layer shifting problems that have been occurring. After this is stable I'll be adding the new features like filament change, and filament run out sensor.

This is for A5S:

SomeoneYouLike

Standby before updating to the firmware. Layer shifting seems to still be an issue however, it has been significantly reduced.

aestrems

Hi Samuel, thank you for the update. I don't have this layer shifting problem, maybe someone can forward a stl or gcode and I can also try. 

Samuel Pinches

@aestrems - thank you for the offer. I've attached a problematic gcode file as an example. If you have time, that would be appreciated.

aestrems

Great, thank you for the file, I will try as soon as possible and I will let you know my results. 

SomeoneYouLike

@aestrems the problem arises consistently with a large print. The smaller size prints don’t usually have the problem. 

aestrems

Hi, here my results. First I’ve tried with the previous firmware and I got layer shifting. Today I’ve updated the firmware and I reprint the file, please take a look on the pictures below 

As you can see the problem has been solved in this update 

SomeoneYouLike

Strange... 

as shown with the photo, I still have layer shifting with the newest firmware update   @aestrems

Samuel Pinches

The challenging thing is, that there are many factors that can cause layer shifting. For example, the stock bearings that come with the A5S are 💩. I replaced mine with these, and my printer moves more smoothly now. I have a little more testing to do, but at least seeing some improvement is very encouraging.

aestrems

I had the layer shift around 12mm height and only on the Y axis. I also replaced the bearings because the noise was just horrible, so my X and Y bearings aren't the originals. 

This firmware update is an improvement so yes is very encouraging.

@Samuel Pinches during my test I've noticed that when you stop/pause your print, the nozzle stay at the same position. On the stock firmware, after stopping the print the nozzle lift and move to X home. Is it possible to modify the firmware to change that thing?  

Thank you for all

chinabean

Hey guys. Just joined the forum after having received my first 3D printer (an A5S). 

After doing lots of research, i definitely want use OctoPi to run the printer which was a nightmare with the stock firmware. I was getting difficulty connecting, and all kinds of weirdness going on that never allowed me to even start a print. Upgrading to the latest stock firmware resolved connectivity problems, but still no print.

That's when I started looking around for firmware and found this thread. I flashed the Beta 5 bin, did the multi point bed leveling, and was able to connect and print with OctoPi with no hiccups in connectivity or starting a print. 

I did have pretty severe layer shifting during the print, so it ultimately failed, but I have not tried any of the aforementioned tweaks yet. Would the ribbon shielding help with that at all? Or should I try the firmware from Samuel's post on May 19th? (I can't tell if that's a newer build than Beta 5)

Samuel Pinches

Please do try that may 19th firmware - it is newer. I haven’t yet had time (I have no spare time at the moment) to make a proper new build for A1 either, but I do recommend try that quick-and-dirty build on the A5S, as it seems to have managed to fix the layer shift issue for some people.

The ribbon shielding will only help with LCD problems, will not affect printing.

Frezap

I am pretty sure that this is a real fix for the layershifts:

https://github.com/MarlinFirmware/Marlin/pull/14030

It even made linear advance possible for me, but I didn't get any print improvement with it...

Samuel Pinches

Yep I’ve been following along with this work, very grateful for those who found the cause. I patched the code with an earlier version of their fix, but I need to redo with a fresh marlin 2.0 build.

Samuel Pinches

Beta #6 posted today! 

• Should fix layer shifting problems.
• More responsive touch screen.
• Entire new build of Marlin!
• Hidden easter eggs too

Download here

Spark

Did this end up fixing the layer shift?  I have not upgraded because the layer shift would drive me mad! I really miss my octoprint though :-( 

Samuel Pinches

@Spark - yes, There was a bug in the firmware that was causing shifting. While layer shifting is still a possibility for many reasons, at least that major bug was fixed definitively.

Spark

I just got HUGE layer shift on this print. Can't figure out why though!  I will Load this firmware tomorrow awesome thanks!