KSGER 2.1S r3 controller with defective STM32 processor

[cosmin1]

Hi.
Started the discussion in other tread, but i think it's better to treat this problem separately.
I just received this KSGER controller from China, already defective.
Actually it still works, but the STM32F101 processor it heats to 100 Celsius right after powering on.
Is there any chance to copy the firmware to a new STM32F101 and replace this one?
Or is there any custom firmware compatible with STM32F101RBT6 so i could replace the processor?
Thank you.

This is the model: https://www.aliexpress.com/item/32790819535.html?spm=a2g0s.9042311.0.0.65e44c4dV0jLHZ

[MarkMLl]

What makes you think it's the microcontroller chip that's defective? It could equally be the power controller, or a firmware fault.

At that price it's barely worth fiddling, but if you were to get an ST-Link style programmer e.g. https://www.aliexpress.com/item/32849743354.html you might find that you were able to read the existing firmware and write it to a replacement chip.

That's also the type of programmer you need if you want to e.g. write an Arduino-style loader to a "Blue Pill" module, so the expense wouldn't be wasted.

MarkMLl

[wraper]

First check if it's powered by proper voltage.

[S. Petrukhin]

Check the short circuit here

[MarkMLl]

Very well spotted.

MarkMLl

[BlackICE]

Yes great catch and example of typical quality control, or lack thereof, for products from China. I don't work with STM controllers and to lazy to look up the datasheet, so does anyone know what pins are shorted and is it fatal to the chip?

[S. Petrukhin]

Yes great catch and example of typical quality control, or lack thereof, for products from China. I don't work with STM controllers and to lazy to look up the datasheet, so does anyone know what pins are shorted and is it fatal to the chip?

These are the PC10, PC11 pins. I don't know how they are used in the circuit diagram. If they are both used as a rail-to-rail output, they can heat the chip. Although, in my practice, a short circuit of the output in STM32 always led to the failure of this output. And I have never observed crystal heating, even at the maximum output frequency at the outputs, directly controlling the MOSFET gate.

[BlackICE]

The PCB looks like it is hand soldered and not cleaned up too well. I see a lot of flux still on some joints.

[bson]

Check the short circuit here

And that diode might need replacing...

[floobydust]

On previous version controllers 2.1S, STM32 three outputs are tied together to get more drive for the buzzer.

OP mentioned a bad connection to the resistors R3, R6  3.3 Vreg feedback  and the MCU likely got overvoltaged? He'd fixed it to get 3.3V but the MCU died I think.

Almost finished drawing a schematic for this controller board.
If anyone has one, to make a few measurements where traces go, let me know.

[S. Petrukhin]

On previous version controllers 2.1S, STM32 three outputs are tied together to get more drive for the buzzer.

OP mentioned a bad connection to the resistors R3, R6  3.3 Vreg feedback  and the MCU likely got overvoltaged? He'd fixed it to get 3.3V but the MCU died I think.

Almost finished drawing a schematic for this controller board.
If anyone has one, to make a few measurements where traces go, let me know.

https://drive.google.com/drive/folders/1qLmAF7NNqVy2zQ2gsIhTnrgWpaj_WG1P?usp=sharing

[floobydust]

Thanks for that link. I did not see 2.1S r3 sch there?
I have made this DRAFT schematic for this 2.1S r3 board that cosmin1 posted pictures of.
But some traces/vias I cannot see to follow them. If people can tell me the MCU pin or just markup the page, the I can finish it and will update this post.

[S. Petrukhin]

Thanks for that link. I did not see 2.1S r3 sch there?
I have made this DRAFT schematic for this 2.1S r3 board that cosmin1 posted pictures of.
But some traces/vias I cannot see to follow them. If people can tell me the MCU pin or just markup the page, the I can finish it and will update this post.

Does it make sense to torment yourself by studying a not very good scheme?
Take a look at the R20 - why are they sucking 300 uA out of the battery? And there are still some questionable solutions.

[floobydust]

I think the chinese sell all their mistakes, their bad boards even after they build 5,000 of them 
It took two years to fix the overheating 3.3V LDO 

I do not fix the KSGER design errors in the schematics.
The 2.1S controller r2, r3 has design mistakes - yes including the battery killing R20 and JBC traces (r2). I include it all in the schematics, and put a note mentioning the problem. I draw the schema for people to repair their boards.

We do not have an open-source controller that is compatible with the KSGER versions, that has all the hardware errors fixed.

[S. Petrukhin]

I think the chinese sell all their mistakes, their bad boards even after they build 5,000 of them 
It took two years to fix the overheating 3.3V LDO 

I do not fix the KSGER design errors in the schematics.
The 2.1S controller r2, r3 has design mistakes - yes including the battery killing R20 and JBC traces (r2). I include it all in the schematics, and put a note mentioning the problem. I draw the schema for people to repair their boards.

We do not have an open-source controller that is compatible with the KSGER versions, that has all the hardware errors fixed.

Yes, the Chinese are often surprised. It seems that they can create good things, but every time there is some error in them. They have science and high technology, but they don't achieve quality.

I have a Rigol MSO5074-a good device, a complex development, but why they did not put load resistors on the LVDS receiver side of the logic analyzer is not clear.  And at the KSGER thermofen so in general the wires are connected incorrectly.

Probably, rapid progress does not provide a reliable Foundation, which is developed by long experience.

[MarkMLl]

It seems that they can create good things, but every time there is some error in them. They have science and high technology, but they don't achieve quality.

But despite that, they have a space program.

I think that the only possible interpretation is that the PRC's government is prepared to be ruthlessly brutal in enforcing quality control when it thinks that something matters. It's unfortunate that protection of Western-owned intellectual property and branded designs don't get the same attention.

MarkMLl

[S. Petrukhin]

It seems that they can create good things, but every time there is some error in them. They have science and high technology, but they don't achieve quality.

But despite that, they have a space program.

I think that the only possible interpretation is that the PRC's government is prepared to be ruthlessly brutal in enforcing quality control when it thinks that something matters. It's unfortunate that protection of Western-owned intellectual property and branded designs don't get the same attention.

MarkMLl

I already spoke in this forum... As someone who has lived in the USSR for a long time, I understand the essence of socialism/communism.
In this system, there is no intellectual property, no profit, only recognition and rewards, not money.

All people work for the good of society, all inventions for the good of humanity, not for profit. All developments in the USSR were always open (except for the military), people shared their inventions with joy to make other people feel better. All schematic diagrams were open and always included in the documentation with a full description.

This is a system where there are no rich and poor. In China, this system has been slightly changed, and there are already many rich people there. I do not know how they organize it and how it is embedded in the Communist idea.

[MarkMLl]

I already spoke in this forum... As someone who has lived in the USSR for a long time, I understand the essence of socialism/communism.
In this system, there is no intellectual property, no profit, only recognition and rewards, not money.

All people work for the good of society, all inventions for the good of humanity, not for profit. All developments in the USSR were always open (except for the military), people shared their inventions with joy to make other people feel better. All schematic diagrams were open and always included in the documentation with a full description.

This is a system where there are no rich and poor. In China, this system has been slightly changed, and there are already many rich people there. I do not know how they organize it and how it is embedded in the Communist idea.

Having all citizens working towards the common good is of course orthordox Leninism... and in fact there are few of any political persuasion who would be brave enough to speak against the idea, even if they have vastly divergent ideas on how to get there.

But Hell hath no fury like a Chinese engineer who has found his product ripped off by one of his compatriots, irrespective of whether the design is really his or has been appropriated from a Western company for whom he has done contract work.

However I feel that this distracts from my point that China has what appears to be a fairly successful space program, which implies that when they have their attention appropriately focussed they /can/ get the quality assurance right.

In actual fact I recall an interesting story that goes with that (please don't anybody ask for a reference... via Wp and a chain of various other stuff IIRC). In the 90s(?) there was a collaborative project between the PRC and the USA that involved launching a payload using a Long March rocket. The launch failed, and somebody in NASA told the Chinese /why/ it had failed. When he got home he found himself in serious trouble: not so much for telling the Chinese what had happened, but for revealing that it was possible to deduce such a thing from the telemetry which had apparently not occurred to them.

MarkMLl

[S. Petrukhin]

I already spoke in this forum... As someone who has lived in the USSR for a long time, I understand the essence of socialism/communism.
In this system, there is no intellectual property, no profit, only recognition and rewards, not money.

All people work for the good of society, all inventions for the good of humanity, not for profit. All developments in the USSR were always open (except for the military), people shared their inventions with joy to make other people feel better. All schematic diagrams were open and always included in the documentation with a full description.

This is a system where there are no rich and poor. In China, this system has been slightly changed, and there are already many rich people there. I do not know how they organize it and how it is embedded in the Communist idea.

Having all citizens working towards the common good is of course orthordox Leninism... and in fact there are few of any political persuasion who would be brave enough to speak against the idea, even if they have vastly divergent ideas on how to get there.

But Hell hath no fury like a Chinese engineer who has found his product ripped off by one of his compatriots, irrespective of whether the design is really his or has been appropriated from a Western company for whom he has done contract work.

However I feel that this distracts from my point that China has what appears to be a fairly successful space program, which implies that when they have their attention appropriately focussed they /can/ get the quality assurance right.

In actual fact I recall an interesting story that goes with that (please don't anybody ask for a reference... via Wp and a chain of various other stuff IIRC). In the 90s(?) there was a collaborative project between the PRC and the USA that involved launching a payload using a Long March rocket. The launch failed, and somebody in NASA told the Chinese /why/ it had failed. When he got home he found himself in serious trouble: not so much for telling the Chinese what had happened, but for revealing that it was possible to deduce such a thing from the telemetry which had apparently not occurred to them.

MarkMLl

Yes, they have a space program with good success, and there are many other achievements. The Chinese are doing great, they are developing very intensively, they are hardworking. But I think they have an excessive formality in the organization and some irresponsibility. They do not hesitate to make outright trash, do not strive for pride in their product. But this is already starting to change, more and more they have really high-quality products. It was the same in the USSR - we had modern and serious weapons, we had great planes, but washing machines and almost everything household was disgusting. And our planned economy was terrible because any low-quality product was still consumed, there was no incentive to make it better.

[DavidAlfa]

Yes, the Chinese are often surprised. It seems that they can create good things, but every time there is some error in them. They have science and high technology, but they don't achieve quality.

I have a Rigol MSO5074-a good device, a complex development, but why they did not put load resistors on the LVDS receiver side of the logic analyzer is not clear.  And at the KSGER thermofen so in general the wires are connected incorrectly.

Probably, rapid progress does not provide a reliable Foundation, which is developed by long experience.

Don't get it wrong. Chinese are like all other countries, you get what you pay for.
Buying a $30 soldering station, that  includes aluminium casing, oled screen, handle, iron tips... Don't expect a miracle!
The fact that it works is already surprising  .
There're different target markets. The high quality, expensive ones... and the aliexpress market.
Tons of cheap stuff,  but don't compare them to what you usually would buy for 500% it's price.
Since Aliexpress targets that market, the engineers try to low the cost to the minimum.
When comparing options, let's say you find a soldering station for $28 and another for $30.
From the average user, maybe the ~70% just want to play with arduino and such, but aren't electronics themselves, so will buy the cheapest.
So the numbers make the rules. More quality for $5 more? Or $5 less and sell a lot more that the rivals?
People like us might do a more in-depth research trying to identify the working parts and difference the good vs the crap ones.
Unless they make pictures of the internals, who knows? Nowadays they also do that trick, rise the price $10 to look like it's much better, then you receive the same crap.
So that's the problem. It's not easy to buy with confidence.

But in the quality market, you can get a Jabe UD-1200 clone for $180, half the JBC price, includimg some iron tips (JBCs will costs $30+ each one).
But a you see, the price is not in the same league!


I must admit that they could do much better with some research. Maybe spending $2 more they could fix a lot of things.
But, just looking at the firmware... They don't even use the hardware peripherals. They drive the SPI devices by software and such things.
That means, they don't want to lose any time in development. Take already made code that works (and will work on any device because it's software), and build the firmware with copy & paste.
That leads to what we often see... horrendous performance from a relative powerful MCU, with a screen that have tearing lines and 5HZ refresh rate, laggy controls...

[MarkMLl]

do not strive for pride in their product.

Perhaps the problem is that their sole metric is that people buy it, rather than that people are pleased with it.

Don't get me wrong: I have enormous respect for the way that China is progressing... much like many people in the West had enormous respect for the way the USSR was progressing until they realised what a monster Stalin was.

And I remember when Japan, and before that Hong Kong, unloaded trash on Western markets...

MarkMLl

[S. Petrukhin]

Yes, the Chinese are often surprised. It seems that they can create good things, but every time there is some error in them. They have science and high technology, but they don't achieve quality.

I have a Rigol MSO5074-a good device, a complex development, but why they did not put load resistors on the LVDS receiver side of the logic analyzer is not clear.  And at the KSGER thermofen so in general the wires are connected incorrectly.

Probably, rapid progress does not provide a reliable Foundation, which is developed by long experience.

Don't get it wrong. Chinese are like all other countries, you get what you pay for.
Buying a $30 soldering station, that  includes aluminium casing, oled screen, handle, iron tips... Don't expect a miracle!
The fact that it works is already surprising  .
There're different target markets. The high quality, expensive ones... and the aliexpress market.
Tons of cheap stuff,  but don't compare them to what you usually would buy for 500% it's price.
Since Aliexpress targets that market, the engineers try to low the cost to the minimum.
When comparing options, let's say you find a soldering station for $28 and another for $30.
From the average user, maybe the ~70% just want to play with arduino and such, but aren't electronics themselves, so will buy the cheapest.
So the numbers make the rules. More quality for $5 more? Or $5 less and sell a lot more that the rivals?
People like us might do a more in-depth research trying to identify the working parts and difference the good vs the crap ones.
Unless they make pictures of the internals, who knows? Nowadays they also do that trick, rise the price $10 to look like it's much better, then you receive the same crap.
So that's the problem. It's not easy to buy with confidence.

But in the quality market, you can get a Jabe UD-1200 clone for $180, half the JBC price, includimg some iron tips (JBCs will costs $30+ each one).
But a you see, the price is not in the same league!


I must admit that they could do much better with some research. Maybe spending $2 more they could fix a lot of things.
But, just looking at the firmware... They don't even use the hardware peripherals. They drive the SPI devices by software and such things.
That means, they don't want to lose any time in development. Take already made code that works (and will work on any device because it's software), and build the firmware with copy & paste.
That leads to what we often see... horrendous performance from a relative powerful MCU, with a screen that have tearing lines and 5HZ refresh rate, laggy controls...

I studied at the Institute the theory of how to occupy the segment of expensive goods. Almost all expensive products do not give a tangible advantage, which can be estimated by such an increase in price. They do not give properties, they give a sense of satisfaction, status, show-offs, demonstrate success. Of course, there should be a sufficient level of quality, appearance and packaging.

After all, in the modern world, we all go on about the hucksters: rather, rather, work harder, earn more, buy yourself at a high price, show how cool others are, climb up. What's the use of it? If people don't buy expensive thing, if a modest man, the stars in the sky will not fall, the Moon not falls and the Sun goes not out.  Just don't think that I will soon become a monk and renounce all worldly things.  I'm just as susceptible. But I like Henry Ford's advice better: sell a lot for cheap and then you will become rich, and not a little for expensive.

Chinese guys are very smart. They do not dig in their own sandbox, they greedily absorb all the world's experience, science, and techniques. Most recently, it was a low-tech agricultural country. They were not stopped by the great competition of strong players. I have great respect for China's success and wish them development and happiness. However, it will be good if all people are happy.

[S. Petrukhin]

And I remember when Japan, and before that Hong Kong, unloaded trash on Western markets...

When we changed the system in the USSR, we were taught that freedom of choice is the best thing, that it is better when the buyer decides what products to be on the market.

Please tell me, you have been living in a market economy for a very long time - has the Western buyer formed a desire to buy the best quality? Do you have immunity, brought up over the years, from fraud in the market? Is it possible to attract a Western buyer with a low price?

[MarkMLl]

Please tell me, you have been living in a market economy for a very long time - has the Western buyer formed a desire to buy the best quality? Do you have immunity, brought up over the years, from fraud in the market? Is it possible to attract a Western buyer with a low price?

(Chuckle) I am possibly not the best person to ask.

By and large, a substantial proportion of consumers are swayed by "bling": you might have noticed the extent to which products are emblazoned with the name of the manufacturer even when the brand is known to be trash. I'm prepared to go out on a limb and suggest that that's /particularly/ the case with people who are (a) younger and (b) less prosperous and educated... in short the very people who would benefit from spending their limited resources on robust clothes, decent tools and so on.

There is very little individual immunity to fraud in the market. Instead in Europe there are consumer protection regulations which have to be met before something can be placed on the market (which is where the ubiquitous ce mark comes into play), while in the USA there's much less up-front regulation and disgruntled consumers tend to either sue or send death threats depending on their political inclination.

MarkMLl

[DavidAlfa]

Totally agree with the brand reputation issue.
They smash you with advertising until you definitely associate their name with first-class category.
Not my case, but sure I know a lot of people how do.
I'm much more objective, a glossy finish or high social reputation won't cheat me.
For example, here in Spain the top reputation car makers are german.
They never break down, never have issues, are absolutely perfect.
That's how media sells them and most people do.
And it breaks, they won't tell anyone, what a shame if they know!
If you research a bit more, you'll find the usual problems with automatic boxes, clutches, cylinder heads, injectors,  turbos, electronics... like most brands.
But some people are so arrogant that will say that having to open the engine apart and replace half of the internals is considered "maintenance" in their german.
Now, if other brand breaks a $20 rubber hose, they will say it's crap...

For example, JBC, they are well made. But hell, a linear transformer and a nice made rubber handle doesn't worth $600!

[cosmin1]

Sorry, i forgot about this topic. Problem is solved long ago. New STM, programmed and work like a charm.
Thanks floobydust for all support and help.
I attached here the finished schematic for this black board, if anyone will need it in future.

In the meantime, if that black board was dead, i ordered other version. Blue PCB 'Ve2.1S' r2 KSGER Version (LWFP-64).
Problems here too. Controller worked just fine until i connected the 3V battery.
Suddenly it does nothing. Just the green led from the back lights continuously. The ST-link utility does not see it either, so MCU is not working.
i'm really lucky with these boards... 

[floobydust]

I would check the battery cable has the right polarity. Maybe it is backwards? (-) is to outside of pcb edge.
Did you order these from the KSGER store, or some shady dark alley? Sometimes I think they ship different quality boards.

This Ve2.12S board has the mystery op-amp "621K" that the Russians have problems with, low accuracy on temperature. What does it report for H/W and S/W version?
Thanks for the Ve2.12S pictures, another schematic to draw lol.

[S. Petrukhin]

Problems here too. Controller worked just fine until i connected the 3V battery.
Suddenly it does nothing. Just the green led from the back lights continuosly. The ST-link utility does not see it either, so MCU is not working.
i'm really lucky with these boards... 

After disconnecting the battery, the controller does not work?

Maybe you have a large static charge on you and damage the MCU with it?
Or when connecting the board bent and damaged the soldering.

The STM will not respond to the programmer when properly connected (preferably with an nReset wire) if it is dead or has no power supply. Have you checked the power supply?

[cosmin1]

I would check the battery cable has the right polarity. Maybe it is backwards? (-) is to outside of pcb edge.

I checked. Polarity is good.

Did you order these from the KSGER store, or some shady dark alley? Sometimes I think they ship different quality boards.

They are from KSGER stores (there are 2-3 stores on Ali).

This Ve2.12S board has the mystery op-amp "621K" that the Russians have problems with, low accuracy on temperature. What does it report for H/W and S/W version?

As i remember, it's just the usual H/W 2.0, S/W 2.12, MCU STM32F103 (now can't check it anymore). I had it connected in the past with ST-link without problems.
Can the op-amp be replaced with the 8551? I ordered a few pieces.

Thanks for the Ve2.12S pictures, another schematic to draw lol.

We'll do it, friend! I have more accurate pictures, will send you over if you want. I think is a bit different (or not) from your Ve2.12S.

After disconnecting the battery, the controller does not work?

No. Only the LED lights continuously after 2s from connecting the power and no tip connected.

Maybe you have a large static charge on you and damage the MCU with it?
Or when connecting the board bent and damaged the soldering.

I don't think i had any static charge, because in same time i played with the black board and he's ok. Anyway, i didn't touch the MCU.

Have you checked the power supply?

Yes, i have 3,3V on MCU.
But interesting is, i have only 1,4V in battery connector (without battery connected). On the good board i have 3,3V in it.

[cosmin1]

Solved this too!

Resoldered the quartz oscillator (didn't seem soldered right) and didn't start. But... at least i had again communication with ST-link.
Tested all available fw and only one work: T12_v.2.1S_OLED1.3.
SW version is reported as 2.11.

[S. Petrukhin]

Did you connect the nReset wire to the ST-Link, select a hardware Reset in the settings?
STM32 always starts after Reset on the built-in generator and an external faulty quartz does not prevent the programmer from seeing the MCU. But to do this, the programmer must be able to perform a Reset over the wire.

[cosmin1]

No, i did not.
Just resoldered the quartz and already had the communication with ST-Link.

[SunnyDay]

In the meantime, if that black board was dead, i ordered other version. Blue PCB 'Ve2.1S' r2 KSGER Version (LWFP-64).

It's interesting that your version from the KSGER store has the STM32F101RBT6, but the assembled one I ordered this month from the "KSGER Store" on Amazon advertised as V2.1S came with the identical blue PCB marked Ve2.12s on the silkscreen, but fitted with an STM32F101R8T6 chip, and Q3 and D3 installed to control a vacuum pump. The info screen says
"Soldering Iron
 HW Version 2.1S
 SW Version 2.10"

And I noticed that on some pictures of KSGER irons that I've seen posted, there is a KSGER logo on the screen. Mine doesn't have a KSGER logo on the bootup screen. Does yours have a KSGER logo on the boot screen?

[cosmin1]

My Ve2.12s blue controller has the STM32F103, not F101 MCU.

The KSGER logo i have never seen. How is it look like?

[SunnyDay]

My Ve2.12s blue controller has the STM32F103, not F101 MCU.

Yes, you're right. I didn't look carefully enough at your photo . So that's another difference in these "KSGER" Ve2.12S versions.

Here is the picture from a Russian forum with the KSGER logo:

[cosmin1]

Nice.
I have no such a logo. It all depends on firmware. The chinese did not copy all the details...
Maybe DavidAlfa could help us out, adding boot logo to certain firmwares.

[DavidAlfa]

RB and R8 are the same thing, the R8 has part of flash not tested or defective, therebefore sold as 64KB

[cosmin1]

What do you say David, can you add a boot logo to firmware i use?

[SunnyDay]

I know that the only difference between the R8 and the RB within the same series is the amount of memory, but what I don't understand is if the F101 has enough features to run the iron, why do they put the F103 in some of their controllers?

The other thing I wonder about is how they fit the the stock firmware into F101R8 controllers like mine, because the stock firmware files I've looked at have a binary size of about ~75KB. So are they putting 75K of firmware in an MCU with only 64K of tested memory, and risking malfunction, or are they taking something out of their firmware to make it fit the smaller 64K memory space available?

[DavidAlfa]

Easy, if the flash is defective, it will fail the verification.
They use whatever is cheaper in that moment, simple as that.
Adding a logo would be a huge and hard work, patching a compiled fw needs a knowledge in assembler programming that I don't have..

[SunnyDay]

Easy, if the flash is defective, it will fail the verification.

The problem with that approach is that although the flash memory may pass verification under the specific conditions of temperature and voltage that they test it at in their factory, that doesn't necessarily mean that it will work over its full rated temperature and voltage range that it could be exposed to after hours of operation inside a sealed case without ventilation. It could still be partially defective, and might fail under other conditions. Maybe that accounts for the reports of peoples' controllers freezing, and irons spontaneously glowing red hot. Borderline or defective flash memory is very sensitive to environmental conditions, and chips that can't be read at all at one temperature may read out perfectly at another higher or lower temperature.

I couldn't imagine any of the reputable soldering iron manufacturers stuffing 75K of firmware into a chip that was marked and sold by its manufacturer as a 64K device, even if it happened to pass verification.

[DavidAlfa]

Sure, that behavior is not profesional. But for the DIYer it's a cool hack.
These glowing irons are pretty much software bugs,  they didn't consider certain possibilities and it stucks somewhere.
I've had few of these, mainly caused by hard faults or the main error handler.
So I set some debugging output there, and disable the pwm when It happens.
Didn't had any more iron "bulbs" haha.

[cosmin1]

So these boards can really control JBC tips?
My JBC 245 tip+handle arrived and i would give it a try.

Merry Xmas everyone!