CFW for KSGER/Quicko STM32 Soldering Stations

[dark_hawk]

It's very sad that you're leaving the project, you breathed new life for those units.
Cosmetically, the firmware now for me is in a perfect spot, a little more tweaking in the PID section of things and you can leave it in a great spot if you decide you had enough and want to leave the project.
The 10-08 is better than the last one for sure, still spikey, but you have to lower Kd to 10 instead of 3-4 to get rid of the spikes.
I have to agree with the other posters here, the old PID algorithm was more stable and gave smoother curves and less black magic. With minimal editing of PID you can make a really bad tip to behave.

Edit: There is a small bug with the standby feature. It takes 1.5 minutes over the specified value. I tried it with 1,2,5 minutes. And it always adds 1-1.5 minutes over.

[DavidAlfa]

The menus, settings and everything are there for something. Not just to be fancy.
I have to loose an afternoon trying 100s of pid settings, then making new builds just for you to try. All I can do is guessing, I don't all have the tips and controllers in the world.

Let me explain what should be happening and it's not: You, the community, should be trying, finding and discussing settings for your hardware.
A blinded monkey can build the firmware. There's a video for that, it takes 3 minutes.
If you suspect of the pid, guess what? There're only 3 lines in pid.c that says "old" and "new". Disable the new pid, try the old one and show your findings.
But is easier to just wait, isn't? I'm tired, people seems to translate open source as "it's free, and there's someone doing for me everything I ask".

I completely deleted the new pid code. I hope it does better, because it's not going to come back, ever.
At first, it didn't any better, in fact it was way worse, I had to adjust the PID settings to avoid wild spikes.

dark_hawk, I don't know what are you talking about. It takes exactly the set timeout. Remember encoder activity also wakes the station (and/or resets the timeout) if enabled.

[Cuboy]

The menus, settings and everything are there for something. Not just to be fancy.
I have to loose an afternoon trying 100s of pid settings, then making new builds just for you to try. All I can do is guessing, I don't all have the tips and controllers in the world.

Let me explain what should be happening and it's not: You, the community, should be trying, finding and discussing settings for your hardware.
A blinded monkey can build the firmware. There's a video for that, it takes 3 minutes.
If you suspect of the pid, guess what? There're only 3 lines in pid.c that says "old" and "new". Disable the new pid, try the old one and show your findings.
But is easier to just wait, isn't?

I completely deleted the new pid code. I hope it does better, because it's not going to come back, ever.
At first, it didn't any better, in fact it was way worse, I had to adjust the PID settings to avoid wild spikes.

dark_hawk, I don't know what are you talking about. It takes exactly the set timeout. Remember encoder actitivity also wakes the station, if enabled.

For me it works properly, some tips depending on the profile go crazy and the board asks for a reset, that is because they are new and unstable.

Did testing with 5 tips already, after some baking and then some calibration they work properly.

The new PID looks fine, I'm using the 10-08-2021 build. As I see, the old PID is a more "simplistic" version of the new PID, which I guess you will get less accuracy. I Will test that later.

You're working with hardware that has a lot of different versions, which is pretty difficult in the sense that everyone has its own errors. For me the defaults work properly using the new PID;

systemSettings.Profile.tip

• .PID.Kp          = 7000;           // val = /1.000.000

systemSettings.Profile.tip

• .PID.Ki          = 6500;           // val = /1.000.000

systemSettings.Profile.tip

• .PID.Kd          = 2000;           // val = /1.000.000

systemSettings.Profile.tip

• .PID.maxI        = 75;             // val = /100

Haven't got any errors apart from new tips being unstable (which is normal for new tips). When asking for power(lets say 450c), there is an oscillation of +-2 degrees, which I don't care about.

[DavidAlfa]

The 2021-08-11 has the previous pid algorithm, also similar pid settings to older versions.

[wickated]

its almost fun to look at diffrent user complaints compared to my own rumbles about small spikes on genuine hakko tips. 

still ticketing small bug with stand sleep to permanent stanby on lift.

everything else is good for me, maybe old pid didnt overshoot on liftoff? i dont even remember 

[mastershake]

well i know i appreciate all the time and work you put in for sure. i feel really bad i wasnt able to test as much as i wanted to due to the eye. but know i do appreciate it very much.

[dark_hawk]

I hate to say it but It's still spikey, you still need to lower the Kd under 10 to get rid of the spikes, so I don't think the new PID is the culprit here, more with the filter and reset limit.
Reason for reporting is that tip and the others are your run of the mill Chinese tips, nothing special and no problems, so if I'm having the issues I'm sure a lot do too.
I don't expect you to fix it, but I thought I just report it. I still appreciate your effort and your time, but not all of us monkeys can edit and compile code, some of us - including myself - are hobbyist with limited knowledge in these areas.
For me the solution is easy, go back to the last known working - for me - iteration of the firmware.

[tatel]

I hate to say it but It's still spikey, you still need to lower the Kd under 10 to get rid of the spikes, so I don't think the new PID is the culprit here, more with the filter and reset limit.
Reason for reporting is that tip and the others are your run of the mill Chinese tips, nothing special and no problems, so if I'm having the issues I'm sure a lot do too.
I don't expect you to fix it, but I thought I just report it. I still appreciate your effort and your time, but not all of us monkeys can edit and compile code, some of us - including myself - are hobbyist with limited knowledge in these areas.
For me the solution is easy, go back to the last known working - for me - iteration of the firmware.

Yep, you are right. New PID wasn't the culprit. I have to say that last firmware, after just calibrating with default PID values, never gives me more than a 5ºC oscillation. Good!

Now let me explain what I was speaking about. Here's how my KU tip behaves, after just calibrating with default PID values, on FW 2021-08-2a

[tatel]

Sorry, tried to put pictures aside and it seems it's no good... so here there are...

[tatel]

It takes just one attachment?

[tatel]

Yep

[mastershake]

so which one specifically are you speaking of firmware wise? there was a bunch he released in the past couple weeks. which version was the better for you guys? i was trying to test for a while till they got on me for not getting smoke in the eye. i saved both the new one with the old pid and the last one before that i believe which had the new pid. id love to mess with the firmware but at the moment seeing clearly to read is my issue. its tough still.

[tatel]

This is 2021-08-02a... I don't know abouth others but I'm sure they are there. Please stay health. Best wishes.

[tatel]

Sorry I couldn't resist... this is how things looke like in 2021-08-10

[wickated]

just increase delay 200/20 like it was for t12

[DavidAlfa]

Not having that issue wickated, in stand mode the station goes into run state on lift, in any low power mode.

[alex-z]

I didn't find anything about the CEIAH IC.
Seems  to be an independent voltage regulator for the analog? Or at least it sems to power only the op-amp and the stm32 VDDA.

It's definitely an EEPROM in sot23-5 package (U2 chip)

[tatel]

Let me explain what should be happening and it's not: You, the community, should be trying, finding and discussing settings for your hardware.
A blinded monkey can build the firmware. There's a video for that, it takes 3 minutes.
If you suspect of the pid, guess what? There're only 3 lines in pid.c that says "old" and "new". Disable the new pid, try the old one and show your findings.
But is easier to just wait, isn't? I'm tired, people seems to translate open source as "it's free, and there's someone doing for me everything I ask".

I completely deleted the new pid code. I hope it does better, because it's not going to come back, ever.
At first, it didn't any better, in fact it was way worse, I had to adjust the PID settings to avoid wild spikes.

dark_hawk, I don't know what are you talking about. It takes exactly the set timeout. Remember encoder activity also wakes the station (and/or resets the timeout) if enabled.

What should be happenning and isn't is that it would be better to have a serious developer who listens what the testers have to say. Instead you closed your ears and throwed yet another temper tantrum.

Now we all know that a week ago we already got the 30-bucks-almost-JBC-station. Now we all know that you, the developer, not just didn't even noticed it, but you were complaining we users do nothing while managing our reports with pure comptent. Refusing to think that non-programmers could notice a 5x greater oscillation and know what's better. Saying that it was cheap chinese hardware's fault. Well, it's now clear that it was instead a turd in your code, it isn't?

I'm not here to feed your ego. Please spare the teenager drama and explain where the turd is and how it got sneaked into your code.

Alternatively, put your cojones were your mouth is and go away as you said here many many times. I already got even more than I (and you) were hoping for and I just cloned your repository. Not that I'm wishing to build it. But I can work with 2010-08-02a until I eventuallly have the time and the need for it to do better. That would be bad, but after your answer to wickated, I'm thinking it couldn't be worse that having to support our dear big great toddler.

I don't really think wickated was speaking literally, do you?. Rajoy style won't work.

And please apologize to dark_hawk as a real man would do. I think you owe some apologizing to -at the very least- him.

[DavidAlfa]

Sorry, but if you read carefully, it wasn't against dark_hawk, it was more like a global relief.
In fact you have a been a lot worse with your dramatic 2°C spikes.
Go and install your Quicko/Ksger fw that hides the real temperature, if that makes you happier... Even jbc does that.

Serious developer? Maybe that means: huh I have 2 degree oscillation, please spend a whole day trying something for me! Or is it someone that keeps programming night and day to keep everyone happy?

It has worked great for me since months ago, I was fine stopping there, but check my ego: I carried on, and at least 90% of my time has been for the community!
Almost every requested feature has been added to the firmware, some requiring days of work to fully develop and bugfix. Specially f** bugfix.
So clearly you don't know me at all.
Instead blaming my ego, join the project and start developing...you'll see that pid, tips, pwm and filters are anything but easy to tune.
Maybe it works for you, but then wait, because ot will cause havoc on other stations!
I didn't knew I wasn't allowed to complain! I have a life!
One thing is to support the project and a very different one is people wanting everything done.
The next firmware will remove all options, as it seems some people only know to flash and run, instead using the damn options to tune things up
Otherwise, don't get into the custom firmware thing if you want someone doing all the work.

[Cuboy]

Let me explain what should be happening and it's not: You, the community, should be trying, finding and discussing settings for your hardware.
A blinded monkey can build the firmware. There's a video for that, it takes 3 minutes.
If you suspect of the pid, guess what? There're only 3 lines in pid.c that says "old" and "new". Disable the new pid, try the old one and show your findings.
But is easier to just wait, isn't? I'm tired, people seems to translate open source as "it's free, and there's someone doing for me everything I ask".

I completely deleted the new pid code. I hope it does better, because it's not going to come back, ever.
At first, it didn't any better, in fact it was way worse, I had to adjust the PID settings to avoid wild spikes.

dark_hawk, I don't know what are you talking about. It takes exactly the set timeout. Remember encoder activity also wakes the station (and/or resets the timeout) if enabled.

What should be happenning and isn't is that it would be better to have a serious developer who listens what the testers have to say. Instead you closed your ears and throwed yet another temper tantrum.

Now we all know that a week ago we already got the 30-bucks-almost-JBC-station. Now we all know that you, the developer, not just didn't even noticed it, but you were complaining we users do nothing while managing our reports with pure comptent. Refusing to think that non-programmers could notice a 5x greater oscillation and know what's better. Saying that it was cheap chinese hardware's fault. Well, it's now clear that it was instead a turd in your code, it isn't?

I'm not here to feed your ego. Please spare the teenager drama and explain where the turd is and how it got sneaked into your code.

Alternatively, put your cojones were your mouth is and go away as you said here many many times. I already got even more than I (and you) were hoping for and I just cloned your repository. Not that I'm wishing to build it. But I can work with 2010-08-02a until I eventuallly have the time and the need for it to do better. That would be bad, but after your answer to wickated, I'm thinking it couldn't be worse that having to support our dear big great toddler.

I don't really think wickated was speaking literally, do you?. Rajoy style won't work.

And please apologize to dark_hawk as a real man would do. I think you owe some apologizing to -at the very least- him.

So much drama when just trying to develop firmware, goddammit.

Please just let the drama behind and let's try to work through it. We are here to try to help develop firmware, testing...
Errors can happen, we're all humans.
Adding drama to the equation will not help more, so it's better to leave it out of the equation.

[Cuboy]

Sorry, but if you read carefully, it wasn't against dark_hawk, it was more like a global relief.
In fact you have a been a lot worse with your dramatic 2°C spikes.
Go and install your Quicko/Ksger fw that hides the real temperature, if that makes you happier... Even jbc does that.

Serious developer? Maybe that means: huh I have 2 degree oscillation, please spend a whole day trying something for me! Or is it someone that keeps programming night and day to keep everyone happy?

It has worked great for me since months ago, I was fine stopping there, but check my ego: I carried on, and at least 90% of my time has been for the community!
Almost every requested feature has been added to the firmware, some requiring days of work to fully develop and bugfix. Specially f** bugfix.
So clearly you don't know me at all.
Instead blaming my ego, join the project and start developing...you'll see that pid, tips, pwm and filters are anything but easy to tune.
Maybe it works for you, but then wait, because ot will cause havoc on other stations!
I didn't knew I wasn't allowed to complain! I have a life!
One thing is to support the project and a very different one is people wanting everything done.
The next firmware will remove all options, as it seems some people only know to flash and run, instead using the damn options to tune things up
Otherwise, don't get into the custom firmware thing if you want someone doing all the work.

Will say the same as last reply David, we're being drama queens and we have a spanish flag on our profile. Leave the drama behind, it will hurt more than it will help!

On a serious note here, how do you compile-build for different versions of the firmware once you add a piece of code?

[wickated]

dafuq did i read.

David isnt a developer, he is a contributor. he doesnt sell his firmware, his fw is hobby fork of ptdreamers old one. fw is open source, go to rep, copy it and give it to stm developers who definetely wiil make it work with some china shit flawlessly.

since this noone have a permission to rant on David`s errors or end of development or so on. just use what u have, u can even easily rollback on older fw for stability instead of whining. actually David spend like 2 weeks in a row rewriting code for free before burnout.

p.s. also u all can buy high quality components for hw upgrades or even genuine tips. using faulty fake opamp and shiity solder joints and wires and blaiming fw not supporting this crap is just plain stupid.

i have everything working great, but i did tons of work to make my hw work flawlessly, why cant all of you?

[StephenR0]

I don't like drama.  I joined this project to contribute positively where I can.  I haven't changed my mind.  That is what I will continue to do.  That's my vote.

[mastershake]

i wish i would have kept some of the builds along the way. i would have no idea how far to go back now. some of them worked darn near perfectly and i only updated to get the newer features he was adding even some of the very early builds were great and pretty stable. i may have to grab all the old ones and keep them for testing now.

again i want to say thank you and i would love for you (if its possible) for you to release whatever you feel was the version YOU feel was the best one as kind of the final one. features aside i just want stability myself.

i wish i could take it over i have the time but im so new to stm32 and without being able to read very well im not sure that would be easy to do.

no one should be yelling at him or jumping on him there were hours spent on this and i doubt there were many donations. i was and still willing to send parts to him im not sure others were. just know some of us VERY much appreciate your time you spent on this for us and are sad to see you go and it end this way if it will. maybe offer it to those who want updates for donations or a one time donation something like a patreon or something like that i would be up for that and im sure others would also. at least to get some compensation for all your time.

[ygi]

On a serious note here, how do you compile-build for different versions of the firmware once you add a piece of code?

I'm not quite sure what your question is. If you have watched the video on github ( https://github.com/deividAlfa/stm32_soldering_iron_controller#building ), then you know how to build the firmware from source. The process doesn't change after you've modified the code: save your modified files, press the "hammer" button, done. Now, if you want to know where to change the version string to keep track of your various builds, it's in Core/Inc/settings.h at line 37