CFW for KSGER/Quicko STM32 Soldering Stations

[DavidAlfa]

Nice! Did you do anything special? The changes I made were minimal and I don't think they made any difference about the hardfault.
Did you check if you have a genuine STM32 after all?
Custom boards might be flawed, causing all kind of weird issues. Ensure 3.3V (VDD) is stable and have at least 10uF at the regulator output (This will depend on the regulator, check the datasheet) + several 100nF cap at every VDD/VDDA pin. .
A good design must have some extra filtering for VDDA, the current is very low so it can be easily done with VDD->10-22 Ohms->10uF cap->VDDA, or a 1K @100MHZ ferrite bead instead of the resistor.

[oleg401169]

The voltage is 3.3 volts stable, I’ll look into everything about the board, although before this everything worked stably.

[DavidAlfa]

Well, it took quite a lot of time (It's always the small things causing havoc ), but finally the the new flash storing method used in the temperature was extended to also the tips and profile.
It will wipe the flash page every 512 temperature and/or every 128 tip/profile changes, greatly reducing the stress on the flash.
I tested it pretty extensively, I hope it's not bugged!

Release  v1.11.6.

Edit:

Ouch! I'm forgetting how everything works...

- Small fixes here and there...

- The new way of storing settings caused the initial setup screen to never be shown. Fixed.

- Add Cold Boost function. Option added in SYSTEM options.
  A primitive booster, enabled when entering Run mode, if the tip is 100ºC colder than target, it will automatically boost +50ºC for 12 seconds. 
  It seems to work pretty well, the real tip temperature doesn't get that high, but heats up noticeably faster.
  This is intended for cheap T12 tips, genuine ones will heat up much faster and don't need this
Release  v1.11.7.

[elektryk]

Thanks for another release, will try it soon.

BTW, is it intentional that in stand mode System->Boot option disappears?
I've just commented out line 29 in https://github.com/elektryk2/stm32_soldering_iron_controller/blob/master/Drivers/graphics/gui/screens/system_screen.c and it seems to work but please note that I'm not using the newest version. In the new version it works the same.

And also what should be the behavior when the handle is out of the stand during booting and Boot->STBY is set?

[DavidAlfa]

Of course it is.
Why would you want a boot mode when using stand mode?
It's the stand what sets run mode, remove the handle and it activates, put it in and goes to sleep.
Maybe it could be allowed to boot in standby so it warms up but I don't think it's worth it.

The station boots in sleep or run depending on the stand status, enabling boot mode widget won't make any change, as this option is ignored elsewhere.
Check this.

[elektryk]

It turns out that for some reason this setCurrentMode() call overwrited the previous one that you had highlited https://github.com/elektryk2/stm32_soldering_iron_controller/blob/master/Core/Src/iron.c#L707

After some investigation I've seen that before first call of checkIronError() last_NTC_C value was -999 because readColdJunctionSensorTemp_x10() was called to late and that tripped Err.Flags.
I've solved it crude way by adding firstRun flag https://github.com/elektryk2/stm32_soldering_iron_controller/blob/master/Core/Src/iron.c#L665

If I don't miss something, it should explain why this hidden option took an effect in my case, after adding mentioned flag it should be useless IMHO.

[DavidAlfa]

That behavior is completely intentional, triggering an error at the start, so CheckIronError loads the boot profile later on when everything settled down.
Any error is supposed to clear up during the initial splash screen (1 second), thus the original code assumed this.
But it will set  Error Resume mode if cleared later.

Anyways, let's assume it took only slightly longer than the boot screen, this should fix it.
Extended the time to 3 seconds since boot start, if the error doesn't clear in time then the resume mode will be used.

I hope this is not caused by your own changes, breaking my head for no reason...

[elektryk]

I hope this is not caused by your own changes, breaking my head for no reason...

I really hope not to waste your time.

Just compiled latest release (only board dependent files were mine for different pinout and ADC channels) and it still goes there.

My "testing scenario" is:
-set boot mode to sleep
-set wake mode to stand
-set stand mode to stby
-set resume to last (this is the default option)
-do off/on

And the result is:
-station in sleep mode

I would be very grateful if someone could repeat this and share the result.

BTW Is my mentioned approach really bad? Because it seems to work for me.
https://github.com/elektryk2/stm32_soldering_iron_controller/blob/master/Core/Src/iron.c#L665

[asupsp700]

Switched from V1.11.5 to V1.11.8. Got this... While I rolled back to V1.11.5.....

[lordstein]

I have 2 ksger stations from several years ago and I want to try the custom firmware, which is the best chip to buy?  I have the 48-pin stm32f103cbue installed in one, and in the other I have a 64-pin stm32f103rbt6.  A friend has another one with a 64-pin chip but this one comes with the stm32f103r8t6, which chip should I choose for the tests?  Are the differences between them not important?  or does the firmware differentiate between them and adjust automatically?

[ststefanov13]

I have 2 ksger stations from several years ago and I want to try the custom firmware, which is the best chip to buy?  I have the 48-pin stm32f103cbue installed in one, and in the other I have a 64-pin stm32f103rbt6.  A friend has another one with a 64-pin chip but this one comes with the stm32f103r8t6, which chip should I choose for the tests?  Are the differences between them not important?  or does the firmware differentiate between them and adjust automatically?

STM32F103C8T6 , STM32F103CBT6  or CKS32F103C8T6 , CKS32F103CBT6

[DavidAlfa]

Read the Readme!
It's perfectly explained, any will work, 48 or 64 pin, C8,CB,R8, RB...

[Hunghanh2009]

Hi david. This is my hardfault in FW v1.11.8. PC code: 08003468. It's a commercial version so you can ignore hardware errors because I did it myself. there is an LCD version I made myself that also has the above error. And let me ask, in the schematic folder there are many different schematics, which schematic is most suitable for the KSGER_v3_OLED and Quicko_STM32F103_OLED FW versions. thanks

[DavidAlfa]

People always forget to provide proper details:
- FW version 1.11.18, but which type?  V2? V3? Quicko Fxx?
- How to replicate the problem? (What to do to cause this)

Which schematic is most suitable for the KSGER_v3_OLED and Quicko_STM32F103_OLED FW versions.

All V3 are very similar. You will have to check them out in  DreamCat repo.
Quicko F103 / F072 is *only* for Quicko F103/F072, is a unique controller sold few years ago, available here.

[asupsp700]

I flashed my station with KSGER_v2_OLED_1.11.8 firmware downloaded from here https://github.com/deividAlfa/stm32_soldering_iron_controller/releases/tag/v1.11.8
The error appears immediately when the station is turned on...

[Hunghanh2009]

Sorry david. which is version v1.11.8 KSGER_v3_OLED

[DavidAlfa]

Aha. Well, I found out that I'm bad at maths. 128 - 4 - 6 =124 . Just joking, but this was a real bug.
Fixed the linker scripts, but I don't think this was the issue, as the firmare didn't reach that area.
I made a TON of changes between yesterday and today, seeking simplicity and tidiness.

Edit: This was probably the cause (Accessing undefined address).
Please try v1.11.10.

There were so many changes recently that I got a bit overloaded, thus the bugs, but it's getting it back under control now.
If any left, should be pretty quick and easy to fix them.

[DavidAlfa]

I literally spent an entire day rebuilding the firmware with massive changes.
There's much more control the settings handling, and the ram usage was greatly reduced, opening a lot of options for the future.

v1.12.0 is out, check the release notes.

[asupsp700]

I flashed from 1.11.5 to 1.11.10 after completely erasing the chip. Simply programming without erasing gave an error when turning on the station.
   The profile does not save on 1.11.10. Every time you turn it on, the T12 profile is loaded....

[asupsp700]

I literally spent an entire day rebuilding the firmware with massive changes.
There's much more control the settings handling, and the ram usage was greatly reduced, opening a lot of options for the future.

v1.12.0 is out, check the release notes.

Now I'll flash 1.12.0 . And I will test it!

[asupsp700]

I updated it to 1.12.0. My profile is not saved when the station is turned off... Attached video.
https://youtu.be/G2tdXowkQ5o
Added:
The editor of stings only works with the first one in the list.... I choose to edit 4 in the list and the first one is edited...
But I think all this will be resolved . I would like to wish everyone a Happy New Year 2024!

[oleg401169]

I have the same problem, I had to roll back to the previous version

[DavidAlfa]

Thanks for testing, I found the issue, I'm working on it.
This issue is not critical, so instead rolling back, please try other features and report if you find any other issues.
Don't worry, settings won't be erased in the next version, this is just a initialization issue, so you can set up the tips and settings like normal.
(At least the Hardfault is gone!)

[DavidAlfa]

Everything should be fixed by now : v1.12.1.

Uhh... I need to sleep, I literally spent the entire night hunting the issues
(And they want me taking beers in 4 hours LOL)


This firmware got pretty complex, about 14K lines of code!
Lots of things are placed in a critical order, and everything depends on everything, it's like playing Jenga.
So it's not easy to optimize or make heavy changes... Well, it's done now!

[asupsp700]

David, there is a problem when editing already created tips of stings. Even if I want to edit 3 on the list, I still end up on the first one.
Version 1.12.1Beta