CFW for KSGER/Quicko STM32 Soldering Stations

[LILO]

Possible. But it doesn't eat too much, microAmps according to datasheet. So, I opted to leave it there. Doesn't worth an effort.
And probably one day we'll find a use for it ))

Another interesting idea could be to replace the bat with supercap, and program stm to charge it. Not essential at all, just thoughts aloud ))

[clytle374]

Another interesting idea could be to replace the bat with supercap, and program stm to charge it. Not essential at all, just thoughts aloud ))

I ran the math for a couple days trying to work that out. Self discharge is what gets you. I do love the idea. I built a bench power outlet that uses super caps to keep the display up for a couple hours when the output is shutoff.   But wanted to design something that works with this firmware.   

[spleenharvester]

Revised version of the T12-958 V2 schematic attached, fixed lots and lots of errors. I need to get the capacitor values but I think it is otherwise complete

[DavidAlfa]

It's the FW working correctly then?
Post some calibration values as reference.
The menu should go down when rotating the encoder clockwise.

As for Vbat, I have a small 50mAh battery from a broken bluetooth earbud, connected to 5V through two series diodes to drop the voltage and a 100R limiting resistor, stays at 3.6-3.7V and the settings haven't been lost after a month off.

[spleenharvester]

It's the FW working correctly then?
Post some calibration values as reference.
The menu should go down when rotating the encoder clockwise.

As for Vbat, I have a small 50mAh battery from a broken bluetooth earbud, connected to 5V through two series diodes to drop the voltage and a 100R limiting resistor, stays at 3.6-3.7V and the settings haven't been lost after a month off.

I'm still waiting on my replacement T12-958v2 as I tore the display cable on the first one so haven't tested it yet, but will update when it arrives

Edit - Attached finished schematic with capacitor values, I consider this done now. Hope this helps someone.

[LILO]

Busy weekend. Didn't have time to calibrate it yet. Will update once I do.

Do I understand correctly that NTC temp is the reference point, i.e. CFW assumes that NTC temp is equal to the thermocouple cold end temp and uses it to calculate actual tip temp?

[LILO]

NTC
Pull up 10k
Res 9k (measured at 25C)
Beta 3500

Quicko CF4 tip
Cal 250 1500
Cal 400 2400

[DavidAlfa]

Thanks.
So that's pretty much the same after all.
The amplifier gain is lower (x201 instead x250) so it makes sense the calibration values is lower.

This gain allows up to 410ºC with genuine C245 tips, where in other KSGERs it will saturate at 330ºC.
Thus no modification is required for normal temperatures (Unless you want 450ºC)

Please upload the original FW when you get it!
Alternatively you can extract it with some attack method like stm32f1-picopwner.

This board is now supported from v1.13.11.

[LILO]

Original fw is on previous page
Q958V2.zip

The dump is verified, I flashed it back after playing with CFW and it works.
And actually there was no read protection. Usual 'st-flash read...' did the job.

[LILO]

[Edit - Attached finished schematic with capacitor values, I consider this done now. Hope this helps someone.

Some corrections, as I saw it.

Buzzer, emitter goes to ground.

For RTC
VCC2 goes to +3.3 it's main supply

[spleenharvester]

[Edit - Attached finished schematic with capacitor values, I consider this done now. Hope this helps someone.

Some corrections, as I saw it.

Buzzer, emitter goes to ground.

Is BOOT0 grounded out then? The emitter trace actually leads to BOOT0, which itself seems to have continuity to ground? I thought maybe this trace was responsible for the 'beep of death' when the firmware is incorrect.

For RTC
VCC2 goes to +3.3 it's main supply

Right you are, will fix this in the next release!

[LILO]

It's trivial switch, load in collector, emitter grounded, base is control signal. So, BOOT0 is also grounded. And AFAIK it should be, unless you gonna do something unusual during boot process.

[LILO]

Update on NTC settings.

Pull up 10k
Res 10k
Beta 3950

Checked at several points from 25 to 80 C.
Looks close enough.

[spleenharvester]

Updated schematic of T12-958v2 attached. Changed BOOT0 to GND, fixed the RTC and added a missing 4K7 resistor on +3V3 in the ADC area. Let me know of any more issues. If anyone else with a T12-958v2 could check the ADC section that would be great, I have checked and re-checked and it seems to be correct but some on Badcaps can't make sense of the circuit (it seems to trip at 25A which seems high?).

[DavidAlfa]

In my calcs it tripped at 65A?

[spleenharvester]

In my calcs it tripped at 65A?

Do you think the schematic is correct then? Over on Badcaps they thought it should trip at closer to 2A and hence I might have gotten something wrong on it, but I've checked repeatedly and can't find what I would have missed

[DavidAlfa]

If you have a 2.5V reference through a 20k/3K divider, the output is 326mV.
You'll need 65A to have this voltage drop over 0.005R (5 milliOhms).

[ststefanov13]

I checked the LILO thread about the Beta coefficient of the NTC. I checked the Beta coefficient of the NTC on the few soldering handles I have . After that I bought 100 pcs NTC 10k 1% of which the seller claims that their beta is 3950. I checked 15 of them and it's really that much, please Deivid fix the default value in the next version. But PULL UP resistor is 4.7k.

[DavidAlfa]

Thanks!
How large is the temperature variation in 3450 vs 3950 coefficients?

[oleg401169]

Hello, David! I already contacted you with the problem of connecting the display to the st 7565 (in my case, the display did not want to turn on without connecting the cs pin) and you slightly modified firmware 1.13.10 (for ksger 3.0) to connect the cs pin to pin pb 12. Now I wanted to update the firmware to version 1.13.11 and again encountered the same problem. I would like to ask you - what and where do I need to change in the source code so that I can compile the firmware myself and not bother you again?

[rauloliv]

Hello all,

I decided replace my old hakko 936 with a T12 budget station, and ordered one from china, supposedly a STM32 one.

Of course i got a mm32 
MM32SPIN27PF

Here are some photos for those interested:






Unlike other photos I found, my board has a solder shorted D1 !?

Any recommended  improvements to be made on it?

Thanks a lot

[DavidAlfa]

Without the diode it'll be more noisy due the pwm, but also the diode output as a very small capacitor so not a good design anyways.
If the ADC readings are noisy, I would put the diode and a 100uF cap after it.

[GDN]

Dear DavidAlfa
Please, if possible, add a buzzer sound for the JBC soldering iron when you put the pen on the stand and there is a delay. There is sound only after the delay time has expired.

[spleenharvester]

Dear friends

I've attached an updated schematic again for the T12-958v2. This fixes the errors in the temp sense circuit, and replaces the ferrite bead symbol with an NTC (I assume this is the cold junction thermistor).

I've also bought a Quecoo T12-A Milwaukee cordless station - the one we suspected may have the same board as the T12-958v2. I'll report back on what I find.

[spleenharvester]

Hi all

Received my Quicko T12-A Milwaukee battery T12 iron today. I haven't checked the capacitor values but can confirm the board layout and all other components appear to be identical to the T12-958v2. The only difference is that the pad with the chinese writing is occupied by a wire joining to the metal stand affixed to the side, so it does seem to be some sort of stand sensor that works by connecting that pad to the metal of the iron. The system info reports the same as the T12-958v2, HW version 3.20 and software version 3.2S. It's operated by a latching power switch on the side of the unit.


I haven't tried the T12-958v2 CFW on it (haven't got time at the mo) but I can't see any reason why it wouldn't work. Pics attached.


Cheers