CFW for KSGER/Quicko STM32 Soldering Stations

[bozza]

The purpose of having only one tip by default is to calibrate it, then the new tips will copy that calibration and be more or less close.
I'd add more tips, but it's stupid for a simple reason: You'll have a lot of completely uncalibrated tips, every single one requiring manual calibration.
It's much easier to simply add or copy an exiting calibrated tip and write 4 letters.
Currently, enabling/disabling the tips cannot be easily done. ToDo, maybe in the future.

Does this mean it is okay to run all T12 tips with just one profile? I still don't understand why it is only possible to make 20 profiles? I thought all tips needed a different PID config! Can anyone explain this?

[DavidAlfa]

Do you really have more than 20 different tips? If so, that's current limitation.
This firmware supports 3 profiles with 20 tips each, so that's 60 tips.

Starting a profile from an already calibrated tip usually works pretty well, most tips perform nicely with the same PID settings.

For the picky people wanting it absolutely perfect, not even professional stations do this, they just hide the overshooting and oscillation, and it's fine!

[bozza]

I do have more than 20 different T12 tips, yes. Regularly I might only use 5 though.

Currently I am running the original firmware on my v3.1s ksger. I don't know yet if it has the overheating issue concerning the display.
I will take some photos of the PCB later to post it here.

To me all these PID profiles and configs are a complete alien concept. I have no idea what any of it means.
As a novice I am too concerned about frying a tip by using the wrong PID configs on a different tip.

[DavidAlfa]

I'd appreciate you read the firmware readme and operation manual before asking, most of these questions are described there.

Normally the user only needs to calibrate the tip temperature, not other advanced settings.

Burning a tip? That will hardly happen, more than 2 seconds >100°C from setpoint or >500°C absolute will trigger the safety mechanism and lock the station with a warning.

Of course, if your tip is badly calibrated and it reads 200°C while it's actually at 400°C, it will be unable to work correctly.

The typical consequences of wrong PID settings will be some oscillation, or tripping the safety lock in extreme cases, but current settings works pretty well for most controllers and tips.

Calibrate one tip, then test the temperature accuracy at 300°C with other tips, usually you'll get small (5-10°C) differences, not to worry about, otherwise better make a new profile and calibrate that one.
I have a generic profile (T12) for most, then specific ones for those being too different (JL02, whatever)

Keep in mind there're dozens of controller versions and variations, also, depending on the brand, the tips change a lot, so it's impossible for me to perfectly tweak every posibility/combination.

I'm not working as a development engineer at Hakko/KSGER with access to lots of tools, controllers and tips, neither I make any income from it to justify spending all my free time in it (I worked on it for months during the lockdown), except very few small beer/coffee donations I get from time to time

Edit:
About the regulator overloading, you don't need to worry about.
If it happens, it'll just reset randomly, damaging the hardware shouldn't be a concern, proceed with any workaround to fix it.

[bozza]

Don't get me wrong. I am not asking you for features or to change anything! You are a legend in this community and you made an excellent firmware!  Reading all your interesting posts and being able to follow you on your journey is a privilege in itself!

I was under the impression the PID values were some magical numbers which could destroy the tips or the soldering station if they were wrong.
My philosophy, especially when it comes to firmware on embedded devices, is don't fix it if it isn't broken. Your firmware however might be an exception to this rule and it is tempting to install nonetheless!

I just took some pictures and I was pleasantly surprised to find the updated PSU version in my unit even though I ordered at the beginning of 2021. The production run of this unit was from January of 2021.
The PCB is the black 3.0 OLED. Does this PCB suffer from the overheating voltage regulator? Do all SPI OLED models suffer from this issue? Would you personally consider this to be a "good" PCB?

Is there anything else interesting about this particular unit?

Thanks again David for all your contributions!!!

Edit: Interestingly system info claims the hardware to be version 3.1 but the PCB clearly says 3.0.

[DavidAlfa]

Ah, I think nobody got it wrong here! Just exposed why it isn't perfect
That board should work fine, maybe a bit noisy because it's running a switching converter for 3.3V, while the best would be converting 24->5V and then use a low noise LDO for 3.3V.
You might get some noisy readings due that, but overall it should perform nicely.

AFAIK, there's a backup for your controller here (3.1Snew1.hex), so you should be safe.

Me, a legend? C'mon, this is just a firmware for cheap soldering controllers, not a NASA Rover on Mars .
Perhabs it reduces the cancer risk of being exposed to some terrible chinese firmware.

[bozza]

Hahaha you never know.
Today you have conquered the soldering station firmware - and tomorrow Mars!

I will try out your firmware. There was another project where they were using a raspberry pi 0 to program the stm32. I am not planning to get an st-link any time soon so I think I will attempt to flash it with a raspberry pi.

[DavidAlfa]

If your soldering skills are good, you can lift BOOT0 pin (It's connected to gnd by default), and connect it to gnd again, but through a 10K resistor.

Then solder a small wire from the pin to a switch/jumper, and the switch to VDD.

Now you if you turn on the stm32 while pushing that switch, it'll boot in bootloader mode and you could program it with a simple USB-serial converter as explained here.

[bozza]

Has no one here tried to connect a raspberry pi to the SWD/JTAG interface instead?

I found three references of this capability using the openocd software

http://zansprojects.blogspot.com/2019/01/stm32-remote-debugging-with-raspberry.html?m=1

https://stm32duinoforum.com/forum/viewtopic_f_37_t_940.html

Someone on the raspberry pi forum tried it and has had some partial success. He wasn't able to flash anything but he could read and erase the chip: https://forums.raspberrypi.com/viewtopic.php?t=316090

Instead of soldering to the uart I would rather get the St link instead .

The ksger is my primary soldering station. My only other soldering iron is some cheap weller without temperature controls and a big non removable tip. It is good enough to solder the header pins onto the SWD interface and that's about it.

[totalnoob]

I do have more than 20 different T12 tips, yes. Regularly I might only use 5 though.

Currently I am running the original firmware on my v3.1s ksger. I don't know yet if it has the overheating issue concerning the display.
I will take some photos of the PCB later to post it here.

To me all these PID profiles and configs are a complete alien concept. I have no idea what any of it means.
As a novice I am too concerned about frying a tip by using the wrong PID configs on a different tip.

Even though I follow this thread, I am still using the OEM firmware on my KSGER "v3.1" I bought about 2 years ago. The fact that mine is 2 yo may be a factor because I think the OEM firmware works OK, it does not have the best UI, but I have not run into any issues, I have read about people who have bought theirs after mine and they seem to have issues with the firmware on their soldering stations (leads me to believe that either the boards were cheapened even after mine or the firmware was changed after whatever version came with mine or I just got lucky). That said, I can't say that this works across the board, I only have 4 tips, but when I tested it and switched tips to see what the temperature difference was between different tips using the same profile, it was really minimal (maybe 10 degrees). I have one of those FG100 clones and the temperature it read agreed with what was onscreen, switched to the correct tip profile and the temperature corrected so that it was right on the money. My guess is that with the OEM firmware you will be safe if you have the wrong profile active for the tip you are using, at least as long as the tip is similar in size to the last one you used (i.e. you may run into overheating issues if you use a larger T12 tip then switch to a small one with little mass).

I believe Dave's firmware has some safeties built-in so that the "runaway tip heating" does not happen under his firmware (correct me if I am wrong, Dave).

[DavidAlfa]

As I already said, yes it does, but it can only detect when it's firmware's fault, i.e wrong PID settings or some bug causing a temperature runaway (Outputting power when temp>setpoint).
Faults like shorted mosfet cannot be detected, the signal is the same as when the tip is removed, reads very high, there's no way of detecting if there's current flowing.
Must say I've never burned a tip since it was implemented, so it's doing the job nicely.

With the variations between brands and batches, it's simply impossible.
Buy a BC3, calibrate it. Buy another BC3 3 months later, it won't perfectly match the temperature.
Will be 5, 10°C off, but not perfect.

That's why it's hard to believe it perfectly matched the temperature by just swapping the tip profile, unless you previously calibrated it beforehand for that exact tip.

[totalnoob]

That's why it's hard to believe it perfectly matched the temperature by just swapping the tip profile, unless you previously calibrated it beforehand for that exact tip.

If this was directed at me, then I don't know. I never re-calibrated the unit for the tips, I only changed the active tips to get the correct temperature. I only have the 4 tips that came with the station I purchased, so perhaps, since the tips and the station were all from KSGER's "official" store on Aliexpress (funny since they had three apparently different "official" stores at the time, have not looked lately), they have/had tighter control over the tips so that the calibration lines up with what is in their stations and tips from another supplier would not be as close?  Perhaps they pre-calibrated the stations based on the batch of tips produced/being sent out at the time (and if I were to buy a new tip, today, of the exact same as one I already have, and tried it with the original calibration, the calibration would be off)? Perhaps they pre-calibrated the units with the tips they sent, although this would not line up with my experience using the tips the first time, since my tips had the typical non-reading for about 15 minutes (or whatever the time was) until the tip had "burned in" and the station was able to detect the tip and temperature.

[bozza]

There were some posts that did a teardown of some of the t12 imitation tips. Different manufacturers use different amounts of copper inside the tip. Some use more, while others use less. The hypothesis of ksger calibrating the soldering station to their particular tips is very plausible.

[KocsisV]

I do have more than 20 different T12 tips, yes. Regularly I might only use 5 though.

It is possible to change the number of profiles to 1 then increase the number of tips to 60 if that works for you: edit the settings.h and change ProfileSize to 1 and TipSize to 60 and recompile. The total number of tips should not be over 60 because there is simply not enough area reserved to storing it.
I have not tested this, it might not work properly, but only one way to find out

Temp cal is pretty important in my opinion, I have a few tips where 400C is 1950 adc count, while others need 2200. This would be around ~50 C difference if I use the tip with the more sensitive thermocouple with the wrong calibration.

[DavidAlfa]

I think it should work...

[ygi]

Unless this limit has to do with the 10KB RAM devices, might as well increase the size of settings partition to 3 or 4KB in addition to TipSize instead of reducing ProfileSize. I mean, what are the chances the firmware is going to need all that free space for future development?

[DavidAlfa]

There's one problem, though.
The flash must be temporary stored in the RAM, prior erasing and writing the new settings, so although the flash has enough space, I'm not sure about the RAM.
Increasing to 60 tips per profile boost the memory requirement from 1.62KB to 4KB. Each tip is about 60 bytes (Not exact).

Edit: Nope, 60 or even 50 is too much 10KB ram devices, but 40 does fit. The STM323F103 could store something like 150..200...
Is that enough? How many *** T12 tips exist?

Commited this and made new builds, report any issues...

The code could be modified to split profile settings and tips in different sections.
Then, always load the tips directly from the flash, instead storing them in ram.
Temporary read the tip when entering tip settings screen, modify it, and save it when exiting.
F072 has 2KB sectors, while F10x has 1K sectors, another enhancements would be to perform the operation sector by sector.
All this this needs time, work, testing...  I have other priorities right now!

[RaySoft]

Hi!

https://ibb.co/ySCPrSV
https://ibb.co/n7Km9W1

I didn't find any pictures of my board in the github repo.
It's a KSGER HW v2.1S (fw:2.10)

I guess it's the same pinout for the ST-LINK in the upper left corner as the rest of them?
And is that a STM32F103 C8T6 or CBT6? (I guess it's 8?)

Edit: Also I just noticed that the pin holes for the STM32 is too narrow for the 4 stlink wires to reside next to eachother:-/

[DavidAlfa]

Check compatibility section:
4-pin display = v2.x, use that.

It's a C8, but doesn't really matter.

Check programming section.
From left to right:
VCC, SWDCLK, SWDIO, GND.

[islcom]

Hi,
Struggling with a jbc clone handle on Ksger board V2.
The c245 handle verified working on a Aixiun T3A soldering station.
Ksger board display says 320c but is only around 200c
If I heat the c245 tip the opamp output ranges 0-800mv for 30c-400c.
Using the latest firmware and board verified working with t12 tips.
Have tried two boards, same result.
Am I missing something here? should I reread the documentation?
I have linked in a video.
Any help most appreciated.

https://youtube.com/shorts/ZkH15hy_oqY

Thanks

[ygi]

Hi,
Struggling with a jbc clone handle on Ksger board V2.
The c245 handle verified working on a Aixiun T3A soldering station.
Ksger board display says 320c but is only around 200c
If I heat the c245 tip the opamp output ranges 0-800mv for 30c-400c.
Using the latest firmware and board verified working with t12 tips.
Have tried two boards, same result.
Am I missing something here? should I reread the documentation?
I have linked in a video.
Any help most appreciated.

https://youtube.com/shorts/ZkH15hy_oqY

Thanks

A few things after watching the video:

1) Power supply voltage's sagging significantly under load. IDK if it's just a quick and dirty test setup or not but if you want a realistic idea of performance you need PSU, cables and connections that can handle the current drawn.
2) Let me be Captain Obvious there but is your tip even calibrated?
3) If after fixing point 1, you still get those massive overshoot and oscillations then PID settings should be adjusted.

[islcom]

Ok thanks.
I get similar results on a 24v 10amp supply.
I assumed it would work with the default settings.
I'll calibrate the tip and check the pid numbers.

Cheers

[DavidAlfa]

KocsisV, I've been trying (again) passing variables to the Makefile.
It should be possible to declare them in the Environment (Under Properties... C/C++ Build).

Ex. Declare "BOARD_TYPE", with content "KSGER_V1_5", etc, for each build profile.

Then in the Makefile:

clean :
ifeq ($(BOARD_TYPE), KSGER_V1_5)
    clean_KSGER_V1_5
else ifeq ($(BOARD_TYPE), KSGER_V2)
    clean_KSGER_V2
else ifeq ($(BOARD_TYPE), KSGER_V3)
    clean_KSGER_V3
else ifeq ($(BOARD_TYPE), Quicko_STM32F103)
    clean_Quicko_STM32F103
else ifeq ($(BOARD_TYPE), Quicko_STM32F072)
    clean_Quicko_STM32F072
else
    @echo "clean: BOARD_TYPE not defined"
endif


all :
ifeq ($(BOARD_TYPE), KSGER_V1_5)
    ./boards/KSGER_V1_5/src/generated
else ifeq ($(BOARD_TYPE), KSGER_V2)
    ./boards/KSGER_V2/src/generated
else ifeq ($(BOARD_TYPE), KSGER_V3)
    ./boards/KSGER_V3/src/generated
else ifeq ($(BOARD_TYPE), Quicko_STM32F103)
    ./boards/Quicko_STM32F103/src/generated
else ifeq ($(BOARD_TYPE), Quicko_STM32F072)
    ./boards/Quicko_STM32F072/src/generated   
else   
    @echo "all: BOARD_TYPE not defined"
endif

It should fix the unecessary code generation, but so far, everything I tried has failed, it's not getting the variable.
I'm always getting "BOARD_TYPE not defined".

I've seen a lot of examples in Eclipse, so I guess it's not a Makefile version problem?

Any ideas? I feel like we're getting close to make it work!

Ah, I forgot: Please use spaces instead tabs. Eclipse settings, Editor... I'm pretty sure you know all that!
I had very nasty viewing issues with some browsers or editors, they used different spacing and the code looked like crap.
Just to keep the formatting nice. I use 2 spaces for everything, viewing was perfect in any plattform!

[totalnoob]

Hi,
Struggling with a jbc clone handle on Ksger board V2.
The c245 handle verified working on a Aixiun T3A soldering station.
Ksger board display says 320c but is only around 200c
If I heat the c245 tip the opamp output ranges 0-800mv for 30c-400c.
Using the latest firmware and board verified working with t12 tips.
Have tried two boards, same result.
Am I missing something here? should I reread the documentation?
I have linked in a video.
Any help most appreciated.

https://youtube.com/shorts/ZkH15hy_oqY

Thanks

A few things after watching the video:

1) Power supply voltage's sagging significantly under load. IDK if it's just a quick and dirty test setup or not but if you want a realistic idea of performance you need PSU, cables and connections that can handle the current drawn.
2) Let me be Captain Obvious there but is your tip even calibrated?
3) If after fixing point 1, you still get those massive overshoot and oscillations then PID settings should be adjusted.

Also:

4) You made the alteration to the KSGER Board? I believe there is a trace that needs to be cut and perhaps a resistor removed?

[yelkvi]

Hi,
Struggling with a jbc clone handle on Ksger board V2.
The c245 handle verified working on a Aixiun T3A soldering station.
Ksger board display says 320c but is only around 200c
If I heat the c245 tip the opamp output ranges 0-800mv for 30c-400c.
Using the latest firmware and board verified working with t12 tips.
Have tried two boards, same result.
Am I missing something here? should I reread the documentation?
I have linked in a video.
Any help most appreciated.

https://youtube.com/shorts/ZkH15hy_oqY

Thanks

1. Calibrate not only the T12 stingers, but also the C245 - each profile separately. I have very different values on T12 and C245.
2. In the photo, your power supply is only 17 volts - first make a power supply.

P.S. Try it with a power supply from Aixun. I did that. I also have an Aixun T3A - with its power supply, such a controller with firmware from David works well.