CFW for KSGER/Quicko STM32 Soldering Stations

[Tatanka1961]

@Gandalf_Sr
Hi, I think it's a great idea. I would certainly be interested.
As the controller that was originally used is not available anymore, there are only 2 options:
1. Modifiy PT Dreamer's code to work with the currently available controllers. This would be the simplest solution (at least for someone who knows programming) and provide for the possibility for additional features to be added.
2. re-design the pcb to work with the original firmware as you propose. An option would be to use a STM32F103 development board such as this: https://www.ebay.com/itm/STM32F103C8T6-Minimum-System-Development-Board-Module-Core-Learning-For-Arduino/113797404420?hash=item1a7eda9704:g:MkcAAOSw6CxdEMMg
They are cheap and it would considerably simplify the PCB design and assembly.

Also, I would opt for a header to connect the connector for the handle. This leaves it free to choose the type of connector you want.
I'm also in favor of keeping the existing form factor for the enclosure.

My main requirements are that I want to use a JBC T245 handle + tips; the handle is much smaller and I believe the tips are better and shorter. Furthermore, the tips can be easily removed/inserted without screwing.
Unfortunately, the chinese KSGER stations don't allow using JBC handle/tips.

I would be a candidate for beta testing either way.

[dam234]

I have this pcb version of ksger t12.




Does anyone have any firmware for it and can share it with me?

[alex79]

First of all, thank you PT Dreamer for custom firmware and Jomat for adapting it for v2.1S board. Great work.

I just bought 2.1S with 1.3 OLED and STM32F103C8 ... just try one on aliexpress and this come.

Even Im not developer, somehow I have managed to get the firmware compiled and flashed based on tutorial: https://www.ptdreamer.com/chinese-stm32-oled-soldering-controller-part2/

Just few things that I have to manage to make it work. Can help someone. I dont understand GIT so just how I have done the changes:

• Jomat - your code is looking for i2s driver and Delay driver in /home/jomat/t12/hello/helloiron/Drivers/ instead of ../Drivers/ which fail on error during compilation
  I have changed in file .cproject lines:
  <listOptionValue builtIn="false" value="/home/jomat/t12/hello/helloiron/Drivers/software_I2C"/>
  <listOptionValue builtIn="false" value="/home/jomat/t12/hello/helloiron/Drivers/1us_delay"/>
  to
  <listOptionValue builtIn="false" value="../Drivers/software_I2C"/>
  <listOptionValue builtIn="false" value="../Drivers/1us_delay"/>
  
• My screen is 1.3 OLED, so I needed to change:
  In file ssd1306.c change "//#define SH1106_FIX" to "#define SH1106_FIX"
  
• I have also added into code 2 latest PT Dreamer fixes, that are missing in Jomat code:
  fix standby timing  - https://github.com/PTDreamer/stm32_soldering_iron_controller/commit/c82850fb84bf1f76b8c03f231cffd560f6b82ccd
  fix tip overheating - https://github.com/PTDreamer/stm32_soldering_iron_controller/commit/55332eaa6803638e1a5b382a8f639544c576266b
  

Im also planning tio get rid of Debug screens, however its not big issue for me now.
For enyone interested also link to Jomat code (without changes mentioned above) https://github.com/jomat/stm32_soldering_iron_controller

[dreamcat4]

First of all, thank you PT Dreamer for custom firmware and Jomat for adapting it for v2.1S board. Great work.

Hey there I looked at the 'Insights' graph today...

https://github.com/PTDreamer/stm32_soldering_iron_controller/network

and it shows this other Github user LuckyThomas, who is many commits ahead of Jomat.

https://github.com/luckyTomas/stm32_soldering_iron_controller/tree/flawless_testing

Although it is not clear what exactly was added without manually walking through all of his commits log there. Since he did not update the README with any summary of whats those additional changes.

Furthermore:

There is another person found elsewhere who has been making some improvements towards supporting v3.0 or v3.1 pcb board. So I have asked them to kindly please come and share that with us. Since it is not on Github yet I also offered to put it up on Github, if they cannot themselves learn the git system etc.

Either way I am very happy to find these 2 new additional contributors and hopefully it will slowly improve the level of support for other board revisions. Perhaps the support is not fully complete but it is a great way to help others and fill in the gaps. So it is not so completely daunting the task.

We shall see / give some time. Personally I am still too busy to fire this thing up yet. Because I want to complete my fume extractor project first before going back to soldering tasks... however working on this is actually for me a priority task. Love to work with these JCB T245 tips. About time to give more attention to it.

 

[CapnBry]

Oh snap this is awesome news that somebody has been working on this. I think I still have my beeping device broken controller in a box somewhere so I'm going to have to check this out. My working device says:
HW Version [2.1S]
SW Version [2.10]

Is that what you had, alex79?

EDIT: luckyTomas's git includes the two fixes that alex79 posted, so that's a good sign.

[dreamcat4]

Haha you know I never actually got around to turn mine on. It never occured to me to see if the version / revision was actually being reported and displayed in the menu... better do that myself.

 

edit:

My unit is reporting reports v3.00 hardware, and v3.02 firmware versions.

[CapnBry]

I did find my old PCB and when I put 24V to it, it still goes BEEEEEEEEP so I'll try to flash it with the luckyTomas git source when I have a chance in the next few days. My current job has declared my project non-essential without pay and suddenly I'm overwhelmed with the number of hobby projects I'm picking back up, this included!

[CapnBry]

Hey what do you know, it beeps when it is supposed to now! I guess now I have to decide if I want to take apart my iron to put the old board with the new software on it.

[CapnBry]

I soldered an old handle to my newly flashed controller and it worked like a charm. I'm trying to decide if I flash my stock KSGER with this code or if I do all the work of removing the old board and desoldering the connector and moving it over to this board. Probably not that because I damaged this OLED a bit when replacing the STM32 chip when trying to fix my beeping.

Im also planning tio get rid of Debug screens, however its not big issue for me now.

If you mean just get rid of booting into the debug screen, then just change iron/Drivers/graphics/gui/splash.c

Code: [Select]

static int splash_processInput(screen_t * scr, RE_Rotation_t input, RE_State_t *state) {
if(HAL_GetTick() - splash_time > SPLASH_TIMEOUT)
return screen_debug3;
return -1;
}
Change return screen_debug3 to return screen_main. If you want to remove them from the settings menu, just comment them out in settings_screen_setup() in settings_screen.c (same directory).

Code: [Select]

comboAddItem(widget, "DEBUG3", screen_debug3);
comboAddItem(widget, "DEBUG", screen_debug);
comboAddItem(widget, "DEBUG2", screen_debug2);

For those wondering how to get to the settings screen, it is hold down the encoder button and turn the knob left or right. You can switch from one settings screen to the next with the same interaction, once you've selected a setting from the list.

[czajunia]

Long time lurker first time poster so Hello to everyone.

Slightly out of topic so probably not the best start but I just wanted to check something. Does anybody know if there are custom firmwares available for the attached board? It's written on the other side "T12 Controller Board V2.2" and the screen is "OLED V2.0 SH1106".

Many thanks!

[scythefwd]

Heres a pic of my new v3.0 controller from KSGER. Which arrived today. The table @ https://www.st.com/en/microcontrollers-microprocessors/stm32f103.html?querycriteria=productId=LN1565 says it has enough flash space on it.... so it aught to be compatible i guess



and typical (as per usual), we can see this board arrived with some flux residue left on it. No biggie really. Should come off

@dreamcat4 Am I viewing that PCB correctly and pins 3 and 4 share a common connection to ground?

[dreamcat4]

@dreamcat4 Am I viewing that PCB correctly and pins 3 and 4 share a common connection to ground?

Just checked it with a multimeter... that is correct, yes.

[floobydust]

That's a KSGER Ver 3.0 controller? It's a fail, or going to soon. Older version boards died due to the voltage regulator U5 overheating because it had bad a copper pour for no heatsinking, like what you have. It's just not enough to lower 24VDC to 3.3VDC and power the OLED. You'll see when you burn your finger on it.

[dreamcat4]

Yeah there was also some Russian comment somewhere else online saying that these are in fact garbage. The other issue is that somebody tried adapting the custom firmware and could not get it to work properly. Its still sitting here on my desk, after all this time. The piece of garbage that it is.

 

What we should be recommending for everybody is to only buy ones that are specifically sold as being v2.1s going forwards. This is the most recent known good model that is also supported in the firmware. Much fewer of these online listings actually specifically state or guarantee v2.1s board. So it is important to be sure before buying that you trust the seller, or it has existing reviews to confirm that etc.

The other implication is that it's usually best not to buy as some complete unit inside of a housing with the PSU etc. Since it's less clear what version is inside. Unless you already happen to know that the model you are buying contains a v2.1s (or similar) controller.

This isn't really so bad if you wanted it that way you can still buy the other parts separately as a complimentary kit. And need to open it up anyhow in order to upgrade the firmware, and solder in a different 6 pin DIN connector for JBC T245 handle etc.

 

[floobydust]

Any STM32 soldering station controller boards using LDO with no copper pour for heatsinking-  they just roast unless you add a metal tab for heatsink or lower Vin below 24V for the Vreg. Add a cap or diode across the heater to protect the mosfet, then they work OK.

[scythefwd]

@dreamcat4 Much thanks. 

[darkflame808]

Good afternoon, I've been studying the thread and have been interested in exploring the KSGER CFW scene.

I ordered what was advertised as a 2.1S model as suggested in the forum and this arrived.

It does say 2.1S when I boot it and it does has the 103 version of the ship but it looks nothing like any of the others.

Do you think it's a fair assumption that this might be yet another one in the many millions of versions of this product? LOL

I'm going to just put it back together until I can find some type of shining light that this is safe to proceed on. I am hopeful this photo will help others in pursuing a proper board to proceed wth.

[darkflame808]

Well for the sake of science, I went ahead and flashed to see what happened... Sure enough the beep of doom.

Ended up flashing back 2.11 in the mean time. Might explore other firmwares until an update comes out for this HW revision.

Thanks!

[dreamcat4]

Well for the sake of science, I went ahead and flashed to see what happened... Sure enough the beep of doom.

Ended up flashing back 2.11 in the mean time. Might explore other firmwares until an update comes out for this HW revision.

Hey there. I understand you have read up on the details of this already. However you did not mention here in your last message which specific version of the firmware you actually tried to flash. I only mention this in case you didn't notice that there are other forks on Github which claim better or improved support for the v2.1s... either way, please if you could just let us know which of those specific firmware your tied which was not working. Thank you.

 

[darkflame808]

Thank you for your response. In all honestly I didn't do due diligence. It was due to an error on my part. I ended up flashing PTDreamers version. When I read of the progresses people made to adapt to other boards, I thought they were all consolidated on to PTDreamers github. I will revisit the six pages and try all the different versions I can find.

If one thing positive came out of this, I learned how to go back to stock! now i'm not so afraid to jump off the cliff anymore. I will update once I've tested all the versions I can find.

[dreamcat4]

Cool man.

thought they were all consolidated on to PTDreamers github. I will revisit the six pages and try all the different versions I can find.

OK so a more direct method to find all of the current activity on Github is to click on the 'Insights' tab. And then click on the 'Network' page. This gets you to the Network graph view. From which you can see everything. All of the known activity (unless a user has chosen to upload their changes back up as a separate repo, which cannot be tracked using this specific method).

Here we go:

https://github.com/PTDreamer/stm32_soldering_iron_controller/network

Then follow the commit lines to identify which other fork(s) contain useful activity that was not already merged back into the primary repo.

As you can see here LuckyThomas seems to have made an extensive number of commits.

You can then review those commits one by one. However since there are so many it is sometimes easier to just diff only what has changed in the README.md file.

To do that we have to generate a diff. So append `/compare` to the repo url (of this fork). To get here:

https://github.com/PTDreamer/stm32_soldering_iron_controller/compare

Then we can see amongst all the changed files

https://github.com/PTDreamer/stm32_soldering_iron_controller/compare/master...luckyTomas:master?short_path=04c6e90#diff-04c6e90faac2675aa89e2176d2eec7d8

those changes only in the README.md:



At this point I now see that Thomas has actually bothered to include schematics etc. for the v2.1s, in this same repo:



And indeed... not the same PCB! You are right.



While similar in some places, yet also clearly its different in the middle area especially.

Dear me. They do want to make our lives harder for us, don't they? With all these different PCB revisions. And not being sure which one you will actually end up getting.

I suppose it makes sense now to try to check and trace where the pins on the right of the STM32 are connecting into that central little cluster. How this diverges from the Thomas v2.1s code fork.

Interestingly Thomas' PCB looks like it's also got a cluster of SMDs moved over (or relocated) from the back on to the front side of the PCB. As shown in his front photo. And some markings he seem to have annotated for some reason (sure why myself yet):



At this point (*so many having gotten the wrong PCB!!!) It would be useful for us to actually try harder to find out which specific online resellers exactly will source the correctly the right supported version of this 'v2.1s' board. The Thomas one.

From these pictures the main giveaway seems to be that it has the version number printer on the back side of the PCB. Wheras this other type has the 'v2.1s' only printed the front next to the Rotary encoder Dial, and nowhere on the back. Of course that itself could also be subject to change if somebody else comes along and tries to spin another so-called 'v2.1s' revision. So rather just use that as a first look way to filter out / narrow your search results. Then compare more fully the whole PCB in as much detail as is permitted by the seller images being provided by the listing.

Hope we can improve this situation over time. For example I think we can submit a PR to LuckyThomas' git repo. Just to add some extra lines in his README file to explain that there are multiple different PCB v2.1 revision floating about. Clearly his code was not tested on this other type.

BTW this is also my 2nd soldering controller and neither one of them are the right one.

 

[darkflame808]

I envision a “cfw generator that will walk you through the process”....

Toss up a image of the STM32 with all the PIN numbers.

The have pop up boxes showing components on the board and asking you.

[example - shows picture of OLED screen]
‘Starting from pin one [left side] trace the connection back to STM32 and enter the PIN number here: ______”
[repeat for pins 2-xxxx]

[example - shows picture of buzzer]
Starting on pin 1 ......
[example - shows picture of.  Xxxxxxx]

Rinse and repeat lol

After you fill out the questionnaire it will compile and generate hex wishful thinking!

[Gandalf_Sr]

I'm clicking away on those github links but never do I see a schematic?

If we can agree what the best schematic is, I am willing to redesign the PCB to iron out any issues like hot LDOs.  I started this a while back but then got sidetracked by work but I am idle right now.

[dreamcat4]

I'm clicking away on those github links but never do I see a schematic?

Here you go, up 1 level in the parent folder:

https://github.com/luckyTomas/stm32_soldering_iron_controller/tree/master/schematic/Ver2.1S

BTW i am also trying to email Thomas today just briefly for some few extra notes / infos. If he can help us find it online (from his order). And tell us if he developed and tested his version with either the JBC T245 handle or the Hakko T12 element.

[dreamcat4]

Well I tried to compile the most recent commit from the flawless_testing branch here

https://github.com/luckyTomas/stm32_soldering_iron_controller/tree/flawless_testing

unfortunately I could not get it to work. These ST toolchain are so convoluted it's a total nightmare to use them. Installed the stm32cubemx on github.. which worked. However then you have to press this 'generate code' button, open it up in another tool (Workbench for STM32). Which is actually Escipse IDE. From that point it complain about some orphaned toolchain or something to that effect. In the project properties --> C++ build settings. It's all utter nonsense the amount of unnecessary complication added. BTW I have never successfully been able to use Eclipse IDE in my entire life.

So that is really a dead end after googling the error message. Could not find a direct solution. Instead a bunch of slightly different posts about similar error messages for other specific toolchain dependancies.

Tried, then after a few hours of total frustration just gave up pretty quickly.