Author Topic: CFW for KSGER/Quicko STM32 Soldering Stations (Read 655462 times)

wickated · « **Reply #1075 on:** June 10, 2021, 05:36:50 pm »

Quote from: DavidAlfa on June 10, 2021, 04:51:24 pm

How? We're talking about putting a diode in the source to kill negative spikes when the fet goes off.

diode on source is already onboard. but diodes are used to shut gate.

DavidAlfa · « **Reply #1076 on:** June 10, 2021, 05:43:07 pm »

The diode in reverse polarity to the source and 1M resistor between the transistor src and drain (to ease iron detection) are the only mods I'd recommend.
Some boards already have them, others don't.
I don't have all the schematics in my head

Tugo · « **Reply #1077 on:** June 11, 2021, 05:27:33 am »

Thank you all of you for any kind of reply. I like to "fight" with different opinions because sometime we can all have something out of it

To make it all clear: I have Ksger v2.1s and based on the dreamcat4 and his EXCELENT github it is revision 1: https://github.com/dreamcat4/t12-t245-controllers-docs/tree/master/controllers/stm32-t12-oled/v2.0-and-v2.1-original/r1
From looking schematic in that folder and my PCB (which is the same as on the pictures on that link) I can say a couple of elements are different (I only found R5 to be 249k and not 207k, R3 is 2k and not 20k, transistor driving MOS is not bipolar and is actually also MOS, R13 (10M) is missing on the board and voltage regulator is different) but schematic looks pretty much the same.

Looking into the 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 think I need to change position of C8 (GND side to source of MOS - as on floobydust PDF) and add a RS1A. In case that is not ok, please let me know

The only thing I don't know exactly where to put is 1M ohm resistor.
There are a couple of options:
- in parallel to C8 (between S and D)
- pin 3 of op-amp and 3.3V
- on the R12D before going into op-amp and 24V
Last 2 options... I don't know which is better but they are for future use in case I decide to buy and also use JBC tips. I know I will have to cut the trace like on the floobydust PDF but that is a minor thing.
David, which of the 3 options should be the best based on your knowledge and testing?
Again, thank you all!

DavidAlfa · « **Reply #1078 on:** June 11, 2021, 07:41:41 am »

In my case I can't do that because it's in the back of the PCB, I'd have to unsolder the GX12 connector and OLED.
So I did between the mosfet D-S pins, it works just as well.
Just put a 1M...10M resistor on R10 and you're done!

Tugo · « **Reply #1079 on:** June 11, 2021, 01:23:12 pm »

I desoldered OLED, connector where you connect GX12 and I took battery out because of plastic since I desoldered STM32 off with hot air.
If I understand it correctly: for T12 it is ok to solder 1M between S and D. When you have JBC you don't have a trace between MOS and opamp. And STM32 is reading output from opamp. Am I wrong?
So I will solder it to R12 and 24v (drain of MOS) which will be the same as on your case because I will use just T12 at the moment.
Thank you. I will keep you informed when I will put everything together and test it. After that is T12 test with a lot of tips. That will be for my friend to have a large base

DavidAlfa · « **Reply #1080 on:** June 11, 2021, 06:33:26 pm »

Why? You have the exact place in the board to solder R10. It goes directly to the amp, so it's the best.
Edit: I forgot, I no longer have the D-S resistor. I had changed it to 3.3V and direct input to the amp! Found alternative places in the pcb.
I didn't want to unsolder anything, because after lots of testings developing the firmware, I've done it few times already and the pads are way delicate now, some falling apart.

Tugo · « **Reply #1081 on:** June 12, 2021, 06:43:48 am »

R10 is 4x 4k7 and is under oled when it is on. I don't know what did you mean by R10.
I thought you made a mistake and thought R12 which is R12d in floobdust pdf and is going into opamp.
So 3.3V and plus input of opamp it is

After that oled, gx12, power cables, power supply, fw upgrade and first testing... wish me luck. Good luck to me

But first thing first... breakfast

DavidAlfa · « **Reply #1082 on:** June 12, 2021, 09:09:13 am »

Oops yes, I mixed R13 and 10M making R10

wickated · « **Reply #1083 on:** June 12, 2021, 09:34:47 am »

Quote from: DavidAlfa on June 12, 2021, 09:09:13 am

Oops yes, I mixed R13 and 10M making R10

"ksger pcb design" way

DavidAlfa · « **Reply #1084 on:** June 12, 2021, 10:11:46 am »

That's a serious statement, mind your words!
You can't start calling everyone "ksger designer" !

Tugo · « **Reply #1085 on:** June 12, 2021, 02:27:17 pm »

Dear Ksger designer

,
Help needed...
I always have NO IRON displayed (if I change no iron from 4090 to 4100 I get 999C instead of NO IRON).
SW: 2021-06-02 (latest compiled)
HW: Ksger v2.x
Temperature sensor is working, shake sensor is working.
When there is no IRON inserted I see that LED is blinking brighter then after I insert the tip.

Before I remove 1 M or C8 or diode from S to GND... do you have any suggestion what should I do? HEEEEEEEEEEEEEELP

EDIT: I removed 1M resistor and looks like it works. I will update you later...

DavidAlfa · « **Reply #1086 on:** June 12, 2021, 03:41:05 pm »

Ensure that R13 was 1M. It's easy to make mistakes while reading smd codes.
1M should do nothing unless the tip is removed. Post some pictures!
Something you did created a strong path into the amplifier.
Did you put the diode in reverse polarity, right? Anode to GND!

Tugo · « **Reply #1087 on:** June 12, 2021, 06:29:01 pm »

R13 is 1M. How I know? I also measured it before soldering it on ;-) Actually it is 1.034M if I remember it

As I already typed before: after removing 1M it started working.
Diode in reverse... well, of course ;-)

Attached picture is from before it started working. If I don't have 1M resistor soldering iron detect NO IRON perfectly. I don't see any problems.
But final solution is puting 1M resistor between 24V and R12d (4k7 resistor before going to opamp - pin 3) - on the attached picture marking with yellow.

I left everything on default except ADC values for 250, 350 and 450 C and I found a small "problem". Temperature is not very steady. It goes up and down around the set temperature but power is also not steady it also goes up and down. See the attached 2nd picture. I have to find our PID settings you left for us in one of the later posts and test again.
Anything else I should do?

EDIT: different tips and different PIDs don't change that power/temperature is gioing always up and down. I would guess that this should be almost flat power when you are near set temperature. Maybe I am wrong?

DavidAlfa · « **Reply #1088 on:** June 12, 2021, 08:01:40 pm »

Don't connect it directly to the input, should be before the 4k7 resistor.
Check my Quicko schematics, R999 is the detection fix.
https://github.com/deividAlfa/stm32_soldering_iron_controller/tree/master/BOARDS/Quicko/Schematic

Let's say the 4k7 resistor is connected to GND.
You put the 1M resistor connected to 3.3V, making a voltage divider.
What voltage you get at the amp input?
The calcs give aprox 14.4mV.
14.4x250 gain= 3.6V, so you completely saturate the amp.
Even with 10M, you will offset the temperature by about 70°C (Remember it's just ~21°uV /°C).
Before the 4.7K resistor the load will be 8ohms when connected, adding only 21uV offset.
For 24V, use 10M instead, as 1M makes an offset of 191uV, about 9°C.
You have some oscillation, not huge, seems to be 4-5°C.
First read the GitHub documentation, then tell your situation.
There are some causes that can make that.

StephenR0 · « **Reply #1089 on:** June 12, 2021, 09:24:44 pm »

DavidAlfa has asked for calibration reports. Here is my attempt at a calibration report template with the values for my first calibration report. I've put it in a code block for text alignment purposes. Basically, the template maps to the calibration steps as I now understand them. It starts with adjusting the ADC values in real time measuring the temperature and getting it as close to the step value as possible. Then the calibration is performed. And last, the resulting ADC values are read from the Edit Tips menu. DavidAlfa, does this capture the information that you need for a calibration report? Do you want to make any changes in the format?

Code: [Select]

Model:  Quicko STM32/CKS T12
Tip:    T12-BC2

Calibration > Adjust > 
  Cal. Step 250°C
    ADC Value    1100
    Temperature   253°C
  Cal. Step 350°C
    ADC Value    1600
    Temperature   354°C
  Cal. Step 450°C
    ADC Value    2100
    Temperature   456°C

Calibration > Start

Edit Tips > (T12)
  Cal250         1087
  Cal350         1587
  Cal450         2079

DavidAlfa · « **Reply #1090 on:** June 12, 2021, 10:23:24 pm »

Yes, exactly. The more reports like this, the better.
If I see that same boards calibrate closely, within ~5% margin, then I could put them by default, so calibration would no longer be a must.
Getting 350 or 370°C... It's not a dramatic difference, the tip won't get cherry red.
The problem comes when it's 50...100°C difference.
Either not heating enough or burning everything.
Thanks!

StephenR0 · « **Reply #1091 on:** June 12, 2021, 10:29:35 pm »

Wonderful. I was hoping that I got it right. The key for me was your new FAQ section where you mentioned that the ADC adjust happened in real time. That had not even occurred to me.

Thanks.

Let the calibration reports begin!

DavidAlfa · « **Reply #1092 on:** June 13, 2021, 01:25:05 am »

You know, when you made everything, you assume a lot of things. I realised the firmware had become kinda complex.
Reading all the documentation, things I thought were there, were not. So I tried to add more details.
Anything not clear there, please comment. Anyone must be able to use it based on the documentation, without needing to ask.

Tugo · « **Reply #1093 on:** June 13, 2021, 09:18:41 am »

Since I don't have a problem without a 1M resistor I removed it completely.
I always get NO IRON when I don't have any tip inserted even with that 1M resistor so this is fine for now.

I read your manual and you should put your information that you put here regarding uV/C and mV near opamp in your manual. It is a good point when something is not ok.
After that I went into settings and changed a lot of different parameters one by one checked how it reacted.
Yesterday I found out that my problem is related to lower temperatures and not to highest.
Then I bumped to ADC delay and PWM time. Lowering PWM time is basically how soldering iron reacts to temperature change but is limited with power sending into a tip. ADC delay is basically the same but lowering will give you more time to power inserting into tip but... here is where I found solution for my problem... On lower temperatures I had to set ADC delay to 50 ms otherwise I don't have stable output. Maybe around 45 ms is also ok, but I will play with that later. At the moment I have all settings set to default and ADC delay to 50 ms.

Here is calibration report for Ksger v2.1s with K tip - fake one (this is the tip I don't use at all and I wasn't afraid even if it got damaged

):
Adjust:
1200 - 245 C
1900 - 345 C
3200 - 450 C

Cal 250 - 1459
Cal 350 - 2215
Cal 450 - 3213

I must also point that I am using a laptop power brick 20V/4.5A (actually 20.6 V) and it is not going to be my main power supply (I am waiting for power connector to put everything together).
Is it going to change any settings? We shall see

Also I must tell you I really like how fast it is turning on and response when changing anything

I will keep you posted when I'll have some other testing done!

DavidAlfa · « **Reply #1094 on:** June 13, 2021, 11:34:53 am »

Huh, again the ksger gremlins

.
For a moment disable I term, by setting Imax to 0.
Lower Kp to 45 and Kd to 10.
That should make it pretty lazy, but it serve to ensure it's not the Delay. Leave the PWM to 200mS, it's more than enough.
And check what happens with the Delay.
If you have oscilloscope, checking the voltages would be good too.
Not wanting to cause too much envy, but I've been using it for 2 days almost non -stop. It works really nice!

RobHon · « **Reply #1095 on:** June 15, 2021, 03:19:52 am »

DavidAlfa, Thank you for supporting this project.

I'm here with a Quicko, exactly the same as StephenR0 shows us in the Reply #1065.

I'm not able to find a STM32F103 to buy (shortage due to COVID), so I use a CH32F103C8T6. I decided to switch the uCs because I couldn't figure out how to unlock the CK32F that comes with the Station.

The instructions you provided on GitHub and here make it easy, even for me that had never played with ST uC before).

During my tests I see some unexpected resets... even an error message was showed (I cant remember now)... I will investigate it.

Regarding the calibration processes the last set of messages here make it very clear, see below my calibration. A Fluke 289 + Fluke Type K and a FG-100 clone was used to measure the tip temperature.

Code: [Select]

Model:  Quicko STM32/CH32 T12
Tip:    T12-D24

Calibration > Adjust >
  Cal. Step 250°C
    ADC Value    1350
    Temperature   254,5°C
  Cal. Step 350°C
    ADC Value    1750
    Temperature   351,2°C
  Cal. Step 450°C
    ADC Value    2200
    Temperature   447,1°C

Calibration > Start

Edit Tips > (T12)
  Cal250         1450
  Cal350         1955
  Cal450         2350

DavidAlfa · « **Reply #1096 on:** June 15, 2021, 04:32:10 am »

Very nice and clear report.
Great that you easily understood everything by reading the available docs

There's a noticeable difference from the Stephen results.
I guess it's the potentiometer coming in slightly different positions.
Note down any error you get! Maybe it's a CKS incompatibility? I don't know, never tested the fw in clones.
Is it working well now?

Tugo · « **Reply #1097 on:** June 15, 2021, 04:56:02 am »

David, about Gremlins... did anyone else had similar problem like me?

I attached a couple of pictures. They are in pair and you can see what is going on at MOS source (yellow) and at opamp out (blue).
One pair is with your suggestions (Kp 45, Kd 10, Imax 0, PWM at 200 ms, delay at 20 ms) and the other... I forgot what exactly it was set to. I believe I set delay to 50 ms but I had Kp set to 80, Ki to 45, Imax to 0.60... I really cannot remember what exactly it was because I was trying a lot of things.

One more thing: after checking schematic from Quicko I noticed Quicko have a lot of small capacitors round the opamp. That could also eliminate some spikes that are seen on the output of opamp.
Also Ksger have a LED from (+)opamp to gnd and Quicko has normal diode + capacitor.
I need to find a lot of small capacitors and put them round the opamp.
What do you think?

DavidAlfa · « **Reply #1098 on:** June 15, 2021, 07:17:27 am »

It seems to be a lot of noise there. I guess CH1 probe was set at 10x, while CH2 was 1x.
Put the CH2 V/div (Out) much lower to better see the noise, set the trigger to CH1 fallig edge.
Then change CH2 to the amp VCC.

The out signal is about 5mv/ °C, there shouldn't be more noise than few °C, which gets filtered in the software.
I will explain a bit how the filtering works:

The ADC bit precision (LSB) is ~0.8mV.
It makes 18 samples, discards the highest and the lowest values, and makes the average of the remaining 16.
Then the EMA filtering is done. It uses an acumulator, applying a % of the old value and a % of the new value.
This works very well at filtering noise, but can add a lot of delay if the signal changes quickly.
To overcome that, I limited the filter range for ADC differences of to 200 between new and old values.
Between 200 and 300, 1 to 100% of the new signal is applied, so over 300 the filter is completely resetted using the last value.
200ADC is 161mV, ~32°C, that's the max noise the filter will filter. If you get huge spikes, then the filter will be resetted.

I usually get 5-8°C noise. Check my last pictures
Your's seem pretty bad. I smell poor VCC filtering.

Tugo · « **Reply #1099 on:** June 15, 2021, 11:38:54 am »

Thank you for your explanation and all suggestions. I read it and I will read it later today again when I will do the testing (after I get home).

Both channels are set to 10x, but I forgot to put probe for CH2 to 10x

That is why you see higher voltage per division on channel 2. I know... just call me stupid

I don't know if I will have enough time before Friday but I will do a couple more tests and mods that I thought of. But I will set everything to default and do all the tests you suggested and also my own on default settings just to see differences between them.

I will keep in touch and THANK YOU!


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: CFW for KSGER/Quicko STM32 Soldering Stations (Read 655462 times)

Share me