CFW for KSGER/Quicko STM32 Soldering Stations

[DavidAlfa]

My newest tip, opened today only for testing, barely overshoots. However it seems the sensor bond loosens with the time.
The firmware version from the video shown in Github, which looks perfect, now overshoots and oscillates a lot using the same tips I used to make it, it's just that the tips have worked for few weeks-months and no longer behave the same.
My older tips are now very nervous and overshoot by just looking at them.
I don't really care, it's just for 1 second and then the thing completely flattens out .

Also, it seems moisture affects the tips, after being unused for few days, or just few hours but having cleaned it in the wet sponge, at first they will oscillate a lot.
After few seconds they suddenly stop and start working perfectly, like if there was a switch somewhere.
After cooling the tip and plugging it back it still behaves correctly. But tomorrow might happen again.
Things to take in count when blaming the fw

[StephenR0]

So, with today's firmware, after a full reset, I get a steady oscillation about 5 above and 5 below the set point.  But if I set the filter down to 80%, things settle down quite a bit.  It jumps around, but in a much smaller range, not the steady oscillation as before.  I tried 70% and 60%, but there was no improvement.  The measured temperature stayed in about a 3 to 4 degree range.  Then I tried my wild tip and things were more, well, wild.    So, I tried what has settled it before.  I set Kd to 50 (what I believe to be a very large number).  This tightened up the cycles quite a bit.  The pattern of these cycles is that double top and double bottom that we've seen before with my hardware.  But the range was about 4 degrees, even for my wild tip.  Now, the measured temperature cycles at most 2 sometimes 3 degrees.  These settings seemed to work for all of my tips.  I would say this is quite usable.

Can you help me understand the effect of this rather radical value of Kd?  I'm not really sure what I'm adjusting.  And since adjusting the new filter seems to have an effect, do you have any other filter parameter suggestions that I could try?

[DavidAlfa]

If previous reading and current one changes a lot, Kd acts like a brake or gas pedal on fast changes, like a damper.
But too much will cause it to react at every little noise.

Instead reducing the filter, try reducing the threshold a bit, so the filter self-regulates.
You can also increase the step to 5% for faster response.


Wickated, today I saw the bug you said, I didn't was in stand mode, so it must be a global issue.
When plugging the tip, it went from sleep to standby, could be waken using the button but it kept happening every time I removed the tip.
However after connecting the debugger I wasn't able reproduce the issue again.
I don't see anything strange, but these small little bugs are usually hard to catch at eyesight.
Try to find out what's triggering it. Could be happening after running for a long time, or after entering certain menu...

[StephenR0]

So, I put the filter back up to 95% and started to experiment with threshold and step.  If I understand the setting it's "step down" and you must have meant -5%.  I reduced threshold to 30, then 20 and still had some cycles.  Then I read your explanation of the filter parameters again.  If I understand correctly, we should first see what our noise range is and set threshold larger than that.  So, I went into the debug screen and observed raw to cycle in a range of 50, although the extremes weren't that common.  So, I set threshold to 50.  I'm going to experiment more, but this doesn't look too bad.  But I still have Kd set to 50.  I'm going to have to play with that.

[DavidAlfa]

Yes, not an easy thing when you're having noise.
50 is a lot of noise, however the oscillation shouldn't be considered as noise, it's another thing.
Noise is... 1890...1882....1895, random, fast changes between readings.
Try to find a temperature where it doesn't oscillate and measure the noise there.
 
The current filtering is by far the easiest to adjust from all, yet more useful and capable.
If you don't understand something just ask.
I don't have degree in technical english grammar and vocabulary, so it would't be anything new if I wrote something wrong

[Cuboy]

Does anyone know a socket model number for the STM32 F103 RCT6 - 64 pin, 4x 16 at about 0.25mm spacing...

I've been wanting to do this project for a bit but have a few others going on... so I am thinking of purchasing a few of these chips so that I can desolder the one on my replacement KSGER station, solder the socket on and cut a hole in the top, or side, of the case so it sticks out so that I can swap chips whenever...



Edit: I have board marked OLED-V3.0, in software it reports 3.0 / 3.1 or something along those lines.

I did get pretty lucky with my replacement KSGER - the stm32 chip I have actually has more memory than the original. I do want to see if I can find a way to read the data off the chip, though so I can restore it... but until then I think a socket would be the best bet so I can continue to use it....

If I can't find a inexpensive socket ( some I've seen are around $100 which is crazy ) I could just do a breakout board and then solder the chip to another and just plug the pcb into the breakout and do it that way... It would mean a much larger area for all the pins unless I can see if there is an old cpu socket and cpu I can rip apart and use... that would actually look kind of nifty... But I'd rather have something smaller.


Edit2:

From the previous message by Cuboy quoting a message from DavidAlfa - If you need to know capacitor values, etc... let me know... especially if you have the same board as me... I have a DER DE-5000... I'd probably have to remove the battery - honestly I'd actually be ok with added a header for it... But, it may work without having to undo that... Let me know what readings and what frequency and the position...


Edit 3:

Question: Does anyone know the OLED Model number off my particular KSGER unit? ie: is it color, or just single color? If it is color, I wonder why the colors aren't being used. It would make it much more readable.

We've got pretty much the same version, mine includes the power circuit with it. Do you have schematics for it?

Now the temp graph is stable and i have zero oscillations, which is pretty awesome. If you have the same version as me test it.

[DavidAlfa]

For the schematics, check out Dreamcat4's repo. I think this one is your model (Black pcb?)

[Strand17]

I am having two issues connecting/configuring a KSGER 'v2.1S' Original Version (Green PCB) controller for a JBC T245 handle.

1. The controller is saying "NTC reading high" or "NTC reading low" depending on if the "wake" signal is floated or grounded. The controller stays in "sleep" in both cases. The "wake" input pin is connected pin 17 on the microcontroller, just like in the schematic "iron.pdf". Could the pin setup of the MCU be swapped somehow?

2. The level on the non-inverting input to the temperature sensor OP is around 0.2 V during the measurement pulse, which means the output of the OP saturates. Using an oscilloscope, I cannot see much change in the OP input level when heating the tip with a hot air gun. Note that R5 in the schematic is 247 kOhm on my board, just like on Tugo's.

I have cut the trace between the heater and R12, installed R13 (not originally on my board) and replaced the LED D1 with a Schottky diode. The C245 TC is connected between R13 and R12. The firmware is 2021-08-19 from /BOARDS/KSGER/[v2]/STM32SolderingStation.bin.

A similar board without modification runs fine with T12 tips and firmware from 2021-07-07.

Any suggestions what is wrong?

Thanks in advance!

[DavidAlfa]

In Ksger v2.x firmware:
Pin 16  Wake
Pin 17  NTC
Pin 12  Tip temperature

That schematic might incorrect, all the rest is the same as others Ksger V2.x.
It's a bit strange that they only changed that, requiring firmware modification.

Check the 2.1S schematic, the stm32 pinout is the same, only these labels are swapped.

Try reversing them. Also post pictures to see what's going on.
You won't see much at the input, the T12 sensor will output 8mV at 400°C.
Check the output, it should output ~600mV per 100°C.
The T12 output doesn't work well below 90-100°C.

[wickated]

Great! I was about to make my station land somewhere in between Germany and Russia!

u mean Poland ? 

ill probably spend all day tomorrow building new station with this cool looking 4kg tiny cake

-how many board modification u do?
-yes
(probably it would be faster to just make plain pcb from the scratch)

[illiac4]

Acosmetic one. Not important so you can ignore it, if you find it stupid.

Under system/contracts maybe a better option would be to add just a scale from 1-10 instead of showing 255,230,205,......

[Strand17]

Hello DavidAlfa,

Thanks for the quick reply.

The iron.pdf schematic matches my board with respect to WAKE (pin 17) , TEMPAMB (pin 16) and TEMPIRON (pin 12). I will try reversing WAKE and TEMPAMB tonight and post pictures.

Roughly what signal level should I expect on the OP input during sleep and operation with a C245 tip? With the current input level to the OP, the TEMPIRON (tip temp) saturates at 3.3 V during the short pulses in sleep. I may have messed up something when modifying the board or connecting the C245...

[wickated]

Hello DavidAlfa,

Thanks for the quick reply.

The iron.pdf schematic matches my board with respect to WAKE (pin 17) , TEMPAMB (pin 16) and TEMPIRON (pin 12). I will try reversing WAKE and TEMPAMB tonight and post pictures.

Roughly what signal level should I expect on the OP input during sleep and operation with a C245 tip? With the current input level to the OP, the TEMPIRON (tip temp) saturates at 3.3 V during the short pulses in sleep. I may have messed up something when modifying the board or connecting the C245...

board require modification to support 245. u have to cut 24v line and add 1Mom from 3.3v line to opamp input

on tip-handle connector - outer casing (green) - ground, center rim(blue)- power, end pin(red)-tc

[Tugo]

Don't connect anything before checking what you did.
I believe the problem is your wiring.

You wrote: The T245 TC is connected between R13 and R12.
TC should go between GROUND (which is also ground for the heater) and on the cut trace line going to R12.

How many pins do you have going into the soldering handle? 5? And you don't have 2x ground going into handle, correct? In the original handle you have G and E going to handle and both are ground. One of them (E?) will be needed for temperature reading (TC).
Please make a schematic and/or take pictures and I know it is going to work.

You can also make an old handle with T12 that way it will work with the same board 

Maybe that should also be good to do before T245.


EDIT: I just saw what could be wrong... I hope I am correct! If you check schematic here: https://github.com/dreamcat4/t12-t245-controllers-docs/blob/master/controllers/stm32-t12-oled/v2.1s-and-Ve2.1S/KSGER%20STM32%20Ver2_1S%20schematic%20rev%202_floobydust.pdf I don't think everything is clear or I am thinking it wrong. You have JBC 6 pin on the bottom right. That is not what you should look when connecting to the 4+2 pin header for JBC! It is not pin to pin! T12+ must be connected to HTR+, G is connected to COM and JBC is connected to TC+. Shake and NTC must be connected as in the T12 schematic (E as earth should be connected to ground - G because E and G are the same spot on the PCB!). Please check this first and report back. Also wickated told you to use 1M for R13. I don't know if it is ok to use 10M (in the schematic you gave it is used 10M), but David should answer you this. He is the master 

[DavidAlfa]

Under system/contracts maybe a better option would be to add just a scale from 1-10 instead of showing 255,230,205,......

Will look at that.


It's normal that it saturates during the pulses, the ADC is read when the output is low.
10M should work ok, it goes straight to the amp input.
No master in anything!

[Tugo]

David, come on
You know you are. Just say it

David, again, thank you for everything you did to this FW (how great it become)! I really mean it. I know a lot of other guys also contribute it but I don't think today a lot of different PCB versions would have this FW if you hadn't come and start doing it. 

[DavidAlfa]

Done the brightness stepping. Yeah also I never liked 0-255. I'm not that grumpy, but I have to put some limits from time to time or people will rush with silly/useless/undoable suggestions
Wickated, it's been a long way. I feel sorry for the Polish people, probably triggering their ww2 air raid sirens every 2 days "KSGER coming!! Everybody get into the bunker!!".

So sad that only AndyC checked the code and found it  .
After long-clicking "Auto dim" setting while in edit mode, check what happens when entering sleep mode.

[AndyC]

---
Could the pin setup of the MCU be swapped somehow?
---

I have had to swap NTC/WAKE pins on two of the green 2.1 boards. The KSGER documentation is incorrect, they just fixed the wiring in the supplied T12 handles.

Isn't this a hacking forum? Why I'm not seeing anyone trying adding more caps, showing scope waveforms...?

Right, found an old screen capture from when I was troubleshooting the JCD station:

Watching it live one could clearly see that the first ~30ms after PWM off, the opamp output was not stable.
Did mod it a bit, but now it sits in the 'project box'.

Regarding modding, I had an SSD1309 display on one controller, but it stopped working somewhere between v1.10 and recent versions. I couldn't find the cause, any ideas?

These ksger don't put a cap in the VDDA pin, neither close to teh amp pins, which is absolute cost-saving crap.
Start by adding 1-10uF ceramic caps there, with 100nF in parallel to improve high frequency response.
Then replace the led with a 1n4148 in direct polarization to gnd, to clamp the voltage to ~0.7V.
That diode is very fast (<4nS) and should help with the spikes.

I would start trying this, hopefully will kill those nasty spikes:

Looks like something I should try.

My 'Mini' KSGER 3.1 'flat-lines' on the 210819 version. 

[DavidAlfa]

Is that ssd1309 i2c or spi?
The current i2c code is extremely hacked for best performance.
I thought it was going to fail in a lot of stations, but a few months have passed since then and to my suprise nobody has complained.

You should have reported the issue when you found it. Now.. who knows which was the last working version!
Although there are people actually using 2.42" Oled without issues.

[AndyC]

Is that sss1309 i2c or spi?
The current i2c code is extremely hacked for best performance.
I thought it was going to fail in a lot if stations but seems it's stable, nobody complained.

You should have reported the issue when you found it. Now.. who knows what was the last working version!
Although there are people using 2.42" Oled without issues.

It's SPI, I haven't used that controller for a long time, so I just found out. I'll test some builds in between to see when/what changed.

[DavidAlfa]

You could start by simply opening the .ioc file and increasing the spi prescaler to something much higher, to discard a frequency problem.

[AndyC]

You could start by simply opening the .ioc file and increasing the spi prescaler to something much higher, to discard a frequency problem.

Ah! That was an easy fix! SPI Prescaler 2->4 fixed it.

BTW, saw a toroidal transformer above. How well do these controllers do with a linear power supply? Better or worse than the stock KSGER SMPS?

[DavidAlfa]

Will be mostly the same, with lower high frequency noise, but doesn't justify the high cost.
If the board still has the 3.3v buck converter, it won't improve anything.
Adding proper VDD filtering will do just as fine, see mine


Are you connecting the display through wires? High frequency will mess things up.
If so, shield the signals, ex. use twisted pairs for clock and data (each pair is Signal+gnd), it will help a bit.
Also, play with pin slew rate, too fast might cause resonances, ringing...
Check the spi configuration, gpio, where it says "output speed" (low/medium/high...)
Try lowering that. It's like adding a series resistor, lowering the driving strenght.

[AndyC]

Will be mostly the same, just less high frequency noise, but doesn't justify the high cost.
Adding proper VDD filtering will do just as fine, see mine

Sometimes the cost can be quite low But some caps will be the next thing to try.

Are you connecting the display through wires? High frequency will mess things up.

Yes, since the pinout didn't match, about 100mm wire as it is now.

[Strand17]

David - my mistake regarding the temperature measurement OP. It is working fine and the output responds to heating the tip with the air gun. For some reason I was thinking the measurement was made during the pulses, even though I know that is not the case...

I have swapped the pins for NTC and WAKE, now the NTC is on pin 17 and WAKE on pin 16. The controller stays in sleep independent of the status of WAKE, but I guess that is since the NTC reading is seen an error?

When connecting a C245, what to do with the NTC input? Now it is left not connected, so depending on if pull-up or pull-down is selected, the controller will say NTC reading high or low. Can the NTC measurement be disabled in the software or shall I use two resistors to put the input it at a fixed level?

I will try to make a schematic and take some pictures, but I am not sure I can tonight and tomorrow I leave town for two weeks, so please be patient.