CFW for KSGER/Quicko STM32 Soldering Stations

[Vuhda]

Thank you.  and in case I have 5 wire type and have thermistor, vibration sensor switch, please help me how to connect.
 1. Connect the resistor, vibration sensor, red, white, green, blue wires to the c245 controller circuit (picture below). 
2. How to connect the wires to the gx12-5 jack. 
Thank you again

[yelkvi]

Do it in this way:

[Vuhda]

Do it in this way:

Thank you very, very much.

[yelkvi]

David, thank you very much for the latest firmware.
Everything works very well!

[yelkvi]

David, I noticed this nuance:
If you put the soldering iron handle on the stand, the delay timer will turn on (for example, 10 seconds). But if the handle is removed from the stand before 10 seconds, the timer is not reset. And after 10 seconds, the soldering station goes into standby mode.
When you have free time, please fix it.

[DavidAlfa]

Fixed 

[yelkvi]

Thank you. I'll check tomorrow and post the result.

[yelkvi]

Now everything works correctly. Thank you very much!

[islonina]

I have a problem. My board is the queco green one from github. Moded for c245 handle (not cuting via) and added one wire from c19 to c7. Handle with gx-12 6pin and mesuring plug with SK tip i got: Heater-Gnd is 3ohms,  TC-Gnd ~0.7ohms, Heater-TC ~3ohms. So eather i have broken tip and short in the tip or i dont know. When i remove tip there is no short. Set voltage to 24V and chage in system to c245. When i try to calibrate the station reset itself. What im missing? The firmware v1.09.1 for 1306 oled(one without specifiing oled version like in the older relese)

[DavidAlfa]

C245 have ~2.6ohms, you can't drive them with most 72W PSUs, they will just overload.

You should measure something like this:
- Heater-GND : ~2.6 Ohms
- TC-GND: ~0.7 Ohms
- Heater-TC: Sum of both (~3.2Ohms)

Try setting the power limit to 80W and increasing the PWM multiplier to 10x, so it runs at 50Hz instead 5Hz, the PSU will forgive shorter pulses easier than heavy, long ones.

If still resetting, try setting ADC time to 100 or 50 mS, with the PWM multiplier still set to 10x, so it runs at 100/200Hz.

[islonina]

THX. That give me heating and all now no restart. Only calibration is an error at this power. at 250 during calibration i get max 165.3. So only solution is to replace PSU, thats sad, The psu i use 108W 24V 4.5A

[DavidAlfa]

TC response is not linear so the error will be huge if trying to calibrate with >50°C Target vs Real difference.

In that case,.the calibration fails and tells you to follow the manual procedure under Calibration/Adjust.

If getting the generic "Error, aborting!" message, the system is unstable, too noisy, triggering tip detection error.

Read the manual at Github! 

[islonina]

I know it give me an error. on t12 setup i have planty of tips calibrated. Thx for all the help will order better PSU. Thx again for help and firmware.

[DavidAlfa]

Try enabling the debug menu and watching the TIP values at ex. 300°C.
If it jumping like crazy (The RAW reading will, it's unfiltered), then it's probably the PSU.
My Quicko did pretty well with JBC, has the 108W PSU.
Can't help much more, consider making a video of the debug screen.

[yelkvi]

David, what are the optimal PID settings for 245?
You posted here earlier for T12. And these settings are very different from those in the firmware.

[islonina]

Ok after using manual calibration on 250 and 400 and setting: 250 value 1390 and 400 value 2320 and then auto calibartion for every tip all works fine. Tem is ok no jumping or anything weard. PWM multiplayer is still at 10 and power at 110W. No problem now. I try secound PSU from diffrient unit i have same behaviour. Temp is check with fg-100 so i guess im in clear. On diffrent question. On standby/sleep i hear light buzzing sound after heting started the buzzing deaspierd. No problem just curious. Debug screen 300 on fg100
https://youtube.com/shorts/n1aSUF5qnvU

[DavidAlfa]

Yes, that's the tip detection routine, before reading the TC it outputs a very short pulse.
Normally it's barely noticeable, but reducing ADC time will make it noisier.
If you added a 1M resistor between Heater and 3.3V (Or 10M across the MOSFET drain and source) then you can disable the "Active detection", noise will go away.

Also, now it's working you can try lowering the ADC time back to 200ms, then lowering the PWM multiplier, adjust so it doesn't reboot when fully loading the tip, this will reduce the noise.

I can't tell the optimal PID settings, it's depends heavily on the filtering, the ADC frequency, the calibration... It would be easy if we had only one controller type, but there'ro dozens of variations.

[islonina]

Ok so i will live active detection alone that useful safty feature. I lowered to 200ms and 2multiplayer. And i use t12 and c245 just changing system setting. Thx a lot again. Now i need to tweak PID slightly because it overshoot from 360 to 372. I can work with that after years od PID tuning drones. The only thing on c245 dosnt work perfect is reading room temp at idle. I zero in calibration on cold tip. The most important thing i can use jbc tips and t12 so i dont need 4th soldering station.

[DavidAlfa]

Don't worry about it, that overshoot doesn't really reach the tip, also most stations do this, they just hide it.
10°C is negligible anyways, 360 vs 370 °C don't make a difference at all.
Trying to get it like a 0.01% lab equipment is simply a waste of time.
If you're using chinese C245 tips, just don't, stay with T12, they're absolute crap, performing worse than T12.
Only genuine ones worth it.

[islonina]

I have jbc one and jabe. Both perform the same for my short test for now. Time will tell

[DavidAlfa]

David, I have a problem with the LCD display. When turned on, the display works normally. And when the soldering iron handle is on the stand and the soldering station is in Stand By mode, then artevacs appear on the screen, as in the photo. It may even be a completely black screen.
The only way out of this mode is to restart the soldering station.
Latest firmware: January 2023. Firmware from May 2022 - also.

Sorry, didn't read this before.
Hard to tell, those artifacts are usually caused by data transmission issues, like too long tracks, badly routed (Picking nearby interferences), bad connections, poor VDD filtering...
It's a common thing in DIY stuff.

You could try opening the .ioc file for your board and change the SPI prescaler to a higher value, so the SPI frequency is lower, compiling afterwards.

[yelkvi]

David, thanks for the answer. I once again disassembled and reassembled the soldering station. Flashed the latest firmware. And now the indicator is working fine. What was the cause of the artifacts - I do not know. The conductors that connect the display to the controller board are very short ~ 1 centimeter.

[DavidAlfa]

You could shield clock and data, twisting each wire with another one, grounded at both sides.

[yelkvi]

If the problem repeats, I'll try to make shielding.

[tinfever]

Pardon me if I should have found this in my searching and reading, but is there a go-to open source hardware design that runs this firmware?

It feels ripe for an option of "Go to JLCPCB, upload these files, and they'll build and assemble one for you, then just get X enclosure, install PCB front and back panel, plug in Y power supply, and you're done and have a known good design". (Obviously it's only this easy because of all the work already done reverse engineering and developing all the firmware though.)

*Starts rocking back and forth mumbling* Must not start a new project...must not start a new project...must not start a new project...