CFW for KSGER/Quicko STM32 Soldering Stations

[mikes]

Not trying to make more work for you (well, I guess I am ). Just thinking about how to make it better (it's great compared to the original). It's only a suggestion, the decision is your's - especially since there doesn't seem to be anyone else contributing code. Even if it was "cool and easy", thanks for adding it, because it's more than what we had before.

[mastershake]

i saw this project where he has a graph for the pid setting on his custom station. i played around with pid today and found a few kind of sweet spots. i got super busy at the shop today though so i didnt have as much time as i wanted to mess around with it to much. i do have it really pretty stable with the regular firmware. tomm ill try the one you did specific for ksger with the pid adjustments i would have today as i said i just ran out of time i do apologize. i will mess with that tomm and report back

[mikes]

i saw this project where he has a graph for the pid setting on his custom station.

Can you provide a link to the code? Hardware? Is it open source?

A link to a video sucks. Just saying.

[mastershake]

this is his custom built station i think with a hot air attached as well as the iron. based on a blue pill board. here is a link to the info https://www.hackster.io/sfrwmaker/stm32-based-soldering-and-rework-station-e3955b and the code is here: https://github.com/sfrwmaker/F1-T12-858D#readme

[DavidAlfa]

Remember that RAM usage is around 9.2KB actually.
Each plot will eat 128bytes (128pixel).
It'll come the moment when a decision will have to be done. Forget 10KB stm32s or limit features.
Replacing the stm with a 103 (20KB) is easy but not for everyone.
Nobody tried the pid test build?

[mastershake]

ill be testing it today. i should be at the office soon.

[mastershake]

so i tried the ksger pid firmware. temps were way lower then set temp was out of the start. i did make sure to erase then also reset to be sure. so i tried calibrating the tip and i could calibrate at 200 but it would never get to 300 even if it went above 300 for a second or two on the display it would not give me the option to calibrate there or then of course 400. if i used the pid settings from the normal firmware then it worked fine calibration wise but after calibration (tested three times with fake and genuine k tips) temps are fine at 200-300 but as the temp setting goes up the actual measured temps are much higher with the same pid settings as your normal firmware. at 400 i was seeing around 424 with the hakko themocoupler. and one other slight thing when set to 300 and up the screen would read below 300 or 350 etc. so if i set it to 350 it would hover around 345-348 but was measuring 356-358 same before calibration except the measured temps were much lower then it was set for. it seems i had better temps and control with the normal firmware from you vs this one.

[DavidAlfa]

I was thinking about the tip changing method. What are your thoughts about this.
Change EDIT TIPS menu entry to TIPS
Single click will select the tip and return
Long click will edit the tip.
"Add new" will be still there.

[mikes]

I was thinking about the tip changing method. What are your thoughts about this.
Change EDIT TIPS menu entry to TIPS
Single click will select the tip and return
Long click will edit the tip.
"Add new" will be still there.

Sounds good to me - don't change tips often, and the "push and turn" method doesn't save that much time vs. a TIPS menu. It takes longer to physically change the tip. Move the new TIP menu to the top.

But, if you're going to do that, let me suggest also moving the PID coefficients, Heater R., Power, and "No iron" to the TIPS configuration, since they are stored with individual tips. I don't think PID TIME is, though - should it be? If that's added to tips, the PID menu could then go away.

[DavidAlfa]

Yeah but as PID is often modified, it's much more convenient in the main menu.
And PID mixed with the tips would be weird...

[jesusvallejo]

as debug is no longer by default available,  could there be an option in calibration to enter directly the values?

[DavidAlfa]

Did you see what I answered few posts back about enabling the debug menu?

[mikes]

as debug is no longer by default available,  could there be an option in calibration to enter directly the values?

If you don't have a tip thermometer, how would you know the ADC values to enter? And if you do have one, just go through the proper calibration. And if you calibrated before, but don't have a thermometer to do it again - you should have recorded the actual temperature at each step, then simply entered that again for a new calibration (which would only be needed after the controller configuration was cleared).

Or, just fork the code, then modify and compile it yourself - make it whatever you want it to be.

[jesusvallejo]

Im not asking for someone to implement it, i know i can code and compile my self thanks.As there is plenty people tweaking pid etc, it would be usefull after a hard reset to be able to just enter the values(or a reset in pid menu), not every body has the knowledge or the means to compile, i do, and always try both provided and self compiled binaries. i was just proposing, mb david finds it not worthy or even silly, he'll say why he thinks so and thats all, no need to tell me i can code it my self. 

[DavidAlfa]

But he is right, the calibration will be pretty much the same entering the last temps
The tip and the board will be the same.
Anyways, a full calibration takes 20 seconds!
Also, the reset in pid menu, why? Change the 3 values and done haha.

[jesusvallejo]

yeah , im not saying he isnt right, or that mi idea is good  , i just did not like the 2 cnts

[DavidAlfa]

I wouldn't take it personal 
I know it might seem a bit harsh, but it's the reality when people wants to change everything to their liking.
I took PTDreamer's base and continue it, because it had a lot of potential, but for my liking it lacked a lot of things, had bugs and I wanted better hw support.
What would you think if I started to harass him? Hey I want your firmware to work with my station. I don't like this. That is weird. And so on. He'd probably ignore me or answer some well deserved ugly words. Things don't work like that, specially when most people asking don't contribute in any way (not specifically against anyone).

Better that you never know what passed through my mind (again, not against anyone in special) when I had a bad day and the silly requests started
I understand everyone want to give ideas, but in most cases they should have been analyzed better in first place.
Is it really needed?
Is there actual alternative to fix it/do the same?
Does it worth the effort vs functionality?
"It would be nice..." is not a good reason.
Otherwise, it also would be nice that you guys buy me one of these expensive JBC stations. And introduce to me some hot single ladies. And a nice car.
Moreover, I can give up on the car and the JBC  .
For now the development is stopped until I receive the tips/displays and check the PID stuff. I might fix any bug that  might appear, but nothing more.
But I already said  back then,  it's pretty much done and over. Currently I have no more plans for it. Just achieving it to work nicely and without bugs.

[DavidAlfa]

The oleds arrived. The white was OK, the blue had a worse traveling experience .
I opened a claim, the packaging was very poor, they were dancing inside the envelope with nothing protecting them.
It's a miracle that both weren't crushed.

This is the white result:



The white temperature is warmer than the camera sees.
Also it's a little dimmer than the blue, but definitely better for my eyes, more comfortable to look at.

The first picture is much closer to the real brightness and temperature.

[sergo351]

Hi, I have a KSGER Ve2.1S OLED1.3 board, an STM32F102C8T6 controller (48 pinout), I can use this firmware, or I need to change to an STM32F103CBT6 controller, I looked at the datasheet for stm32F102 / 32F103, the pinout is the same, the board revision is Ve2.1S How does it affect or can only the controller be changed? I use T12 (KSGER) for soldering, but the quality is bad, I bought the Original Hakko T12-B I think the Firmware will work with it Good), Photo of the board   Full review of the Ve2.1S controller here https://mysku.ru/blog/aliexpress/64124.html

[DavidAlfa]

That board is not supported actually...
Since I don't know if you can restore the original firmware, it's better to replace the stm32 instead.
I'm only doing f101 builds, since they will also work on 102 and 103.
Not difficult, it only needs profiling the board for the correct pinout.
But I'm no longer working on it.
I can try to help, but I won't spend a lot of time debugging problems.

[sergo351]

I have two such boards, assembled in the case and a spare, I have already updated the firmware 2.10> 2.12, on the second board my native 2.10, the firmware is on the network, I can solder well)), but I do not understand programming ((, I will buy STM32F103C8T6 for replacement, I wonder what will come of it)

[DavidAlfa]

I only told you to replace it because I didn't want you to lose the original fw, so in the case the CFW was not good for you, you could revert it.
But if the original fw is available, you don't need replacing!

Edit: There's no need of new profile.
Although the pin count is different (48 vs 64 pin), the pin names are identical, it will work in the exact same way.
So you can flash the 64-pin 101 2.1 version.

[DavidAlfa]

The T12 tips arrived.
I only calibrated and made a quick pid adjustment.
When calibrating, remember to wait until the tip temperature is stable.
Although the TC reads 200°C, the tip heat is being transferred, in my case it kept rising for about 10 seconds untils stopped at about 230°C. And that is the value.
The tip needed about 10 minutes to stop doing weird things.
There were noises, like moisture escaping the heat, now it's pretty fine.
I use EMA 2 and pid 140-0-0. In AVG it clearly takes some spikes, EMA completely filters them.
The real temp doesn't overshoot, like the reading does, as I said the tip still needs several seconds to fully transfer the heat.
The fact that the temp never reaches the setpoint is mainly because the PID nature works that way, it needs an small error to keep going. I will probably do like normal stations, fake the temperature if it's in +-10°C range.
I don't see any of that weird issues going to 700°C like crazy.
If it's a KSGER issue, sorry, I don't have the hardware. Someone will need to attach an oscilloscope and see what's going on.

[totalnoob]

@DavidAlfa - Just out of curiosity, is that a real Hakko tip or clone in your last video?

[DavidAlfa]

Hey, that was fast!
Of course they're not. I will use the JBC ones.
I just bought a pair of cheap T12s for testing (BC3 and D24).
If you read some pages back you will see some real hakko tests, they didn't do better in the readings, although the heat transfer probably will be much higher.

Edit: With the stock pid values it worked perfect, much better.
I did so much testing back then... I don't understand why it fails for so many users.
Temp regulation is great. You can perfectly see tip transferring the heat.
Once the TC reaches the setpoint, then the power slowly goes down, as the tip stops absorbing heat.