CFW for KSGER/Quicko STM32 Soldering Stations

[TodayIsFridayInCalifornia]

The screen is definitely the 4-pin variety (1.3"), and the board is identical to one shown in the V2 schematics folder. 

[DavidAlfa]

Strange... Could you post some pictures of the board?
Also from the programmer log
I guess you're using the file from release v1.02 ?

[TodayIsFridayInCalifornia]

I am using the 1.02 build, yes. CubeMX seemed to get stuck when I attempted to run building_script.bat and compiling it in the IDE itself didn't seem to work right (programming to the STM32 over the ST-Link finished in 1.1XX seconds!).

Here's a couple macro shots of the board:

[DavidAlfa]

Where's the stlink programmer message output?

[Jirrex]

Hi all,

Recently I bought a Quecoo T12-945 mini soldering station (I decided to go for the brand instead of the KSGER one, in general people seemed to be more positive about the Quecoo / Quicko ones.
I choose for the mini variant on purpose, it's very compact and can be used with a battery pack, which elimates leakage. From what I have read at several sources, an adapter can cause slight AC leakage which isn´t desirable for the most sensitive components.

Anyway, I wanted to get a station with STM32 chipset since that one is firmware upgradeable. However I can find very little information about my particular station. There's plenty to find about the KSGER stations though.               

I took a look at deividAlfa's GitHub repository, https://github.com/deividAlfa/stm32_soldering_iron_controller, which has plenty of schematics of KSGER stations to determine if a particular station is suitable for the custom firmware. When checking at Quicko. there's no images of PCB's which makes it easier to see whether a station is suitable for a firmware upgrade or not. There's only some references to different kind of STM32 chips, but that's about it.

Is there anyone here who could give me some pointers about how to proceed? I am supposed to disassemble the station and check which STM32 chip my station exactly has? Before getting a station with STM32 chipset I assumed all were firmware upgradeable, hopefully I'm not wrong .

Cheers

[DavidAlfa]

Newer quicko/quecoos are basically ksger v2... Look in that folder.
If the display has 4 pins, go straight with ksger v2 fw, post a picture of your board just in case.

[Jirrex]

Hello David,

Thanks for your reply. Because I have been in a hurry, I just took photo of the STM32 chip, but I assume you need more information?

Anyway, I will post the picture here and when I am back home later, I will make some photo of the PCB or anything else that's necessary.


Cheers

[wickated]

thats the board. just use ksger v2 fw.

[Jirrex]

Really glad to hear that its firmware can be upgraded.

Thanks a lot to you and a David for your fast responses!

[Jirrex]

I have been reading a bit about the procedure of upgrading to the custom firmware, which involves a bit more steps than I would have been expected at first.
 
Anyway, I was wondering if is there's any requirements when it comes to the STM32 ST-Link (V2) to do the actual firmware flashing. At Amazon a pretty affordable one is available:

https://www.amazon.com/Aideepen-ST-Link-Programming-Emulator-Downloader/dp/B01J7N3RE6/ref=sr_1_3?keywords=ICQUANZX+ST-Link+V2&qid=1666207844&qu=eyJxc2MiOiIxLjgwIiwicXNhIjoiMC4wMCIsInFzcCI6IjAuMDAifQ%3D%3D&sr=8-3

Are they good enough to do the job?

If I understood correctly, you first flash the custom firmware, and then you start the calibration process. Could you recommend me any (cheap) high temperature meter for the calibration? Amazon would be desirable because of their fast shipping in general.

Since the custom firmware uses more power, it possibility causes the regulator to overload and will shut down. This applies to any SMT32 based soldering station or it's just a few that suffers from this issue?

[DavidAlfa]

The regulator issue applies only to very specific older ksgers, it failed commonly even with original firmware too, was very much at the limit!
Yeah, any cheap stlink should work...

[ygi]

You could buy a dedicated iron calibration device but most of us simply use a multimeter with thermocouple attachment (they usually go up to 500°C). There are also fake Hakko FG-100 available for cheap (~25euro). A genuine one will cost you around 250euro.

[Jirrex]

Thanks for that suggestion, I do have a thermocouple attachment for one of my multimeters. Actually haven´t used it yet myself, you're supposed to add a bit of soldering wire to the tip and then contact the end of the attachment, or you're supposed to contact the attachment with the bare end of the soldering tip without soldering wire?

[DavidAlfa]

You don't need a crazy accuracy for soldering, +-10°C won't make a great difference, except for fools 

[dcore]

Hi,

Thank you for sharing this open source FW, in particular to the developers, but also to all the contributors.
Excellent documentation!

I have this V3 board: https://github.com/dreamcat4/t12-t245-controllers-docs/blob/master/controllers/stm32-t12-oled/v3.0-and-v3.1/r0/f101-back-hi-res-12.jpg
I flashed it with latest David build (KSGER_v3_OLED.bin V1.02) and after calibration is working very well with T12 tips.
I also like a lot all the customization options and the menu organization.

Next I would like to try C245 tips/handle, I've read here the connections and trace I need to cut (for another PCB): https://www.eevblog.com/forum/reviews/nt115-jbc-handle-for-ksger-t12-with-cfw-by-davidalfa/msg3871406/#msg3871406
Are the T245 connections in the above link ?

One question I have is what pin from my GX12-5 PCB connector should I use to pass the (yellow) TC+ wire?
I was thinking to isolate/separate GX12-5 PIN 4 from the PCB and use that for C245 TC+.
Then I would need to shunt pin 5 and 4 of T12/c210 handles to undo the PCB trace I cut.
I'm not sure if I got this right? So would need some help before I butcher the PCB.

I'm thinking to desolder the GX12, enlarge the PIN 4 PCB hole, so it doesn't touch the GX12.
Is there a better connection option?

[ygi]

Thanks for that suggestion, I do have a thermocouple attachment for one of my multimeters. Actually haven´t used it yet myself, you're supposed to add a bit of soldering wire to the tip and then contact the end of the attachment, or you're supposed to contact the attachment with the bare end of the soldering tip without soldering wire?

Always add solder for reliable reading. It's too finicky to do it dry because you then have to contact the perfect spot at the perfect angle and hold it there several seconds.

[grigore]

Can i use hakko 907 handle with this firmware?
the handle has  a1321 heater
https://hakkousa.com/amfile/file/download/file/2470/product/6798/

[DavidAlfa]

Not directly, but possible modifying the circuitry.
Instead of a 1-15mV thermocouple, that handle uses a PTC resistor of 40-60ohms.
You would need to add a pullup resistor to create a resistor divider, then modify the amplifier to reduce the gain by a lot, so the end signal tops up at 3-3.1V when 450°C are reached.
The curve of that sensor is unknown to me.
You'll have to figure out by yourself!

[dcore]

@DavidAlfa

Sorry to bother you David.
Thank you for all the development work and documentation.
Excellent firmware, much better than the sellers options and it's open source!
It's also an opportunity to learn a bit how this works, had no idea and liked a lot to read the documentation.

I wanted to also use a T245 handle.
I was thinking to isolate/separate GX12-5 PIN 4 from the PCB and use that for C245 TC+.
Can you tell if that is the correct way to proceed?
I'm just an hobbyist and need some guidance.

[DavidAlfa]

Hi,
The message you linked shows the correct jbc connection.
Yes, you could use one of the gnd wires (3/4) for that.
But you'll have a hard time desoldering that connector!
The easiest way would be to cut the pin, so it doesn't reach the board, then solder a wire to it.
Once you do that mod, you'll have to fix the T12 handle, joining both gnd wires together into the same pin, also T12+ with the pin you'll be using as tip temperature sensing (3).

Another approach, tidier and easier, would be this:
Since the T245 handle has no NTC, you'll have to disable it in the settings, so it uses the internal temperature sensor.
Now you can repurpose the NTC pin with few little modifications:

- Remove R10 so the CPU becomes isolated from the NTC pin.
- Cut the trace on R0 to isolate the amplifier from the tip power pin.
- Join the NTC and temp. sense pads.
- Now use the pin 2 for the tip temperature sensing, all should work fine.

The T12 handle would need to be fixed, disconnecting the NTC and joining T12+ with the pin 2.

Disabling the NTC doesn't make a lot of difference, perhabs some °C, but nothing crazy.

Check out your board schematics.

[dcore]

Thank you for the detailed reply.
I now have all the information to do the changes during the weekend.
Still deciding if I'm losing NTC (much easier and clean solution) or use one of the GND pins (if possible without desoldering the socket).

-EDIT-
Without any HW change, I switched the T12 NTC on/off in FW settings.
With NTC ON and after T12 cold tip calibration, I saw around 1ºC (max 2C) deviation from FG-100 clone in 245,255,320,370C.
With NTC OFF and after T12 cold tip calibration (I calibrated the cold tip with CPU at 40C) and I saw around 6ºC deviation, after fooling around with temp. to 400C, suddenly the difference was 12C to the FG-100 clone. Same T12 tip, it's old and well cooked.
These results are based only on a few tries, so no for sure conclusion can be taken with so few tries.

I was now looking for a way to use one GND pin without desoldering the GX12-5 connector and keep the NTC if possible. If to much trouble I would go for the other option, removing NTC and use that pin.

Because PCB is soft, I decided to try to grind around PIN4 to separate it from the PCB GND connections in both PCB sides.
I used a battery power, hand size much smaller dremel like pen (15cm length) and what looks like a carving bit that I have never used before.

It worked much better than I expected, in a few minutes (3 or 5, I didn't clock), also pausing to examine progress and brush dust, the thing carved a perfect pit around the pin and also ate a bit of the pin top. (I don't have any hollow drill bits or milling cutter bit to try.)

[dcore]

Yes, you could use one of the gnd wires (3/4) for that.

After I cut around pin 4 isolating it from PCB, I realize maybe pin 3 was a better option.
Pin 3 is closer to PCB JBC pin to make a connection and I would have kept T12+ and GND pins closer (easy to remember). Anyway any of the options work.

Now I'm facing an issue, with correct connections, C245 temperature read by the controller is wrong and much higher (double or more) than the FG-100 real tip temperature.
It only displays the correct TC+ = ambient temp. while sleeping and cold.
Will do some more tests in next days.

I read the problem is the PCB opamp or DC converter noise.
It makes sense and it's clear the opamp isn't passing the correct temp to the cpu.

I have 3 ksger controllers, all exactly the same PCB version I posted above.
I bough them with coupons a few years ago (when it was cheap) for 9 - 10 eur each.
To dismiss that unit had a problem, because I had repaired it in the past, I used a 2nd controller (not repaired). Now I only cut R0 and solder the correct T245 wires to the PCB and still have the same temperature issue.

Not in any hurry to solve this, so I ordered some ALI opamps, probably more of the same.
I'll search the thread about this T245 issue and make more tests in next days (when I have a bit more time).

[wickated]

looks like u dont need easy ways to do something 

[dcore]

I get it, the easy and clean solution, is indeed the logical solution, so it's the best solution and I agree with this. 
Still there's a bit of logic in my (drilling) "madness", I have 3 ksger controllers and 3 T12 handles and wanted to keep NTC.
(I bought these cheap in 2019/2020. I use 2x T12 D08, with both hands like tweezers when I do SMD and one controller broke so I got a spare.)

In a fast test, I saw 6C to 12C difference with NTC off in FW and want to test more.
Hopefully in the end I can also arrive at the conclusion that NTC is not important.

Still without NTC, every time we power on, the cpu is at ambient ~24C, then after 20mins or so, it'll be ~40C, those 16C startup difference don't affect tip temp? We either calibrate at 28C or at 40C, so I decided to calibrate at 40C.

I need to take some time to search the thread about my T245 temp. issue, but if you have any (easy) solution , I promise I won't drill another hole to implement it. 

[DavidAlfa]

Yes, you might lose 10ºC in teh worst case, is that so critical? Nope.
Did you calibrate the system? Otherwise you're losing your time, of course the readings will be wrong!