KSGER T12 STM32 V3.1S soldering station

[Gandalf_Sr]

I will start another thread but the burning question is this, what version of software will my design use?  If no single coder such as PTdreamer will commit to supporting my hardware design then I will put a different processor on the PCB (probably a 32 bit PSoC) and adapt/write the code myself.

[EDIT] I also want the firmware to be 'good'; I have only ever used the KSGER firmware (if they wrote it) and it's actually OK but the clock is useless mainly because it keeps terrible time although that's probably because of the cheap 32,768 crystal they used. Am I right in saying the KSGER firmware isn't publicly available? If not, then what's the closest or 'best' firmware to aim for?

[dreamcat4]

I will start another thread but the burning question is this, what version of software will my design use? 

This is the version:

https://github.com/dreamcat4/stm32_soldering_iron_controller

I do plan to add some small fixes, and improve the readme file. To make clear simple instructions about what boards its been tested on. And how to compile etc. That is for the v2.1S hardware.

In terms of knowing who will publicly support it. That is really very much a case of this codebase being "the most well supported" of any of them right now (most broadly speaking). Of any of these projects so far. If you make your own entirely new design based on a different MCU then that will have the side affect of fragmenting things even further to have yet another new  project with a smaller user base that is actually interested in it.

So unless *you* are going to support the firmware, on the new platform... then your stated goal will not result in being a better option in the respect of software support. Because you are now requiring an additional burden that they make themselves a PCB and populate it etc.

There are ways to mitigate that. For example is this other guy (on Github) who's project is designed in mind with SMD assembly on JLC PCB. So a project like that would be less hassles for somebody else to order the hardware. Or if you can get the PCB up on ebay etc. The main question people will be asking is what makes your project better than the 5-10 other alternatives out there.

So which features are you personally interested in implementing? Do those line up with what everybody else is looking for? More consideration on that aspect would help to make your project stand out from the other existing projects. And make it more likely to be the one that people choose instead.

Honestly you also need to come up with a better reason than saying that the 'best supported project of all of them' is not well enough supported software wise. And that should (I am guessing) be with hardware features which cannot be implemented as mods to these STM32 OLED boards.

[Gandalf_Sr]

Thanks,

I am not trying to be a dick, I just don't want to do a design that has no usable firmware available.  It seems to me that having a PCB available that eev bloggers could build their own might get away from the $50 junk versions that seem so ubiquitous.  Is the github link you gave for firmware that has similar functionality to the the KSGER version, menus, saving Cal for different tips etc?

What I was saying before was that, if there was no 'good' firmware available, and I was going to write my own, I may as well do it for a processor that I had experience with.

The idea of changes actually wasn't to try to change the circuit but put better components such as an accurate Xtal or make the existing LED visible from the front. 

[dreamcat4]

Is the github link you gave for firmware that has similar functionality to the the KSGER version, menus, saving Cal for different tips etc?

What I was saying before was that, if there was no 'good' firmware available,

We shall see how good it is.... I have not gotten far enough to try it myself yet... because I am still working on the README file. And uploading that to github. Here it is so far:

https://github.com/dreamcat4/stm32_soldering_iron_controller

These BRAND NEW instructions (as of today). Will get you all hooked up over SWD. And into a situation where you can flash the firmware. Which is heh - the point where I managed to get up to today. And have all fully documenteded above there ^^. But no further.

However before flashing it I really want to backup the existing OFW image that comes on the device. If i can. Otherwise if I can figure out how to partially hash the onboard firmware against the numerous different official KSGER firmware images. To get a match for which KSGER firmware is compatible for this particular device.

By including those extra steps in the README. That then enables more other people with different PCB revisions to safely try out this 'v2.1s' branch of the open source PT Dreamer firmware. Since they can then know how to flash back to the OFW if it not work with their specific PCB version. Rather than left with a device with a blank screen, not usable afterwards.

I think this is worth spending the extra time on. Before I flash my own device here. Even though I don't particularly want to go back to the official KSGER firmware myself. Since it does not support JBC tips, does it? Well that's no good!

In fact that is the other main thing I have not gotten around to (yet). Before starting this up. I need to modify the PCB (as per PTDreamer's guidelines). So that the wiring on the controller PCB is correct to work with my JBC tip. Which is also absolutely needed for ensuring the cartridge is grounded properly, etc. And not be a risk to damage.

That I have not started either yet, and will not get around to until the existing OFW firmware is safely identified here (so that it can be restored).

 

[bitwelder]

I'm trying to design in parts that we can actually buy from suppliers such as Digikey and Mouser.

How about making it suitable for LCSC / JLCPCB, so to have a cheap option to get PCB plus SMT assembly, at least for the jelly-beans parts (or similar far-east PCB+SMT manufacturers) ?

[Gandalf_Sr]

I'm trying to design in parts that we can actually buy from suppliers such as Digikey and Mouser.

How about making it suitable for LCSC / JLCPCB, so to have a cheap option to get PCB plus SMT assembly, at least for the jelly-beans parts (or similar far-east PCB+SMT manufacturers) ?

Well I've had most of my recent PCB designs made by JLCPCB (bare PCBs) and the 'jelly bean' parts can always be assembled by them but even that is missing my point about avoiding the issues of the design becoming a lowest cost race to the bottom.  If that's the point then why not just buy what exists today?

Instead I envisioned a bare PCB designed to take quality parts like accurate crystals and reliable rotary encoders and that could be assembled by hobbyists.  As for the code, I am close to just designing my own hardware and adapting/writing the code myself.

[amix]

Gandalf, look at the post counts. This thread is KSGER marketing BS.

If it is, then it is orchestrated very well.

EDIT: Just as I saw my message appearing, I realized, that I also only have a post count of five, lol. But that's, because I do not hang around here often, mostly only reading. You can check the registration date in my profile, I guess.

EDIT2: I also found this on Hackster (Soldering Iron Controller for Hakko T12 Tips on STM32) with this firmware: hakko_t12_stm32

[totalnoob]

Looking at the Aliexpress store at KSGER and it appears they have a disclaimer for the 3.1S units.  There looks to be an issue with what the unit says the temperature of the tip is vs what people are measuring it to be.  I don't see that same disclaimer with the 2.1S units.  Can anyone confirm if the older 2.1S units were fairly accurate in reading tip temperature?  Anyone with a 3.1S notice an issue with tip temperature vs. what the LCD is saying it is or are you getting accurate readings?

Here is the exact wording lifted from the ad on Aliexpress:

Regarding the problem that the temperature displayed on the soldering station is not accurate with the actual temperature:
Please use the soldering iron tip temperature tester (please do not use a multimeter) to test the temperature. When the soldering station is working at low temperature (less than 280 degrees), the temperature displayed on the soldering station and the actual temperature of the soldering iron tip are only for reference , which is not accurate because They are affected by the surrounding environment. The surrounding environment will carry away heat, causing the soldering iron tip to lose some heat. If using a HB tip thermometer, pls make sure the soldering iron tip is suspended at the center of the temperature sensing line. If you are very strict with temperature, please do not buy this soldering station.

[benadamson]

Has anybody worked out if there are any benefits to the versions after 2.01, or is it just more money for no good reason? There seem to be bugs in the newer versions. Not sure if other undocumented bugs are present in the older ones.

[totalnoob]

The newer reviews I have seen online seem to indicate the parasitic drain was fixed in the later 2.1s models.  Outside of that problem, most of the reviews for the 2.1s's have been positive, I am not sure what other "bugs" there are in the 2.1s models, except for the hardware safety issues (i.e. the heat sink potentially providing a path between the household voltage side vs. the electronic voltage side, grounding issues, etc.) but they affect all models so far.  I have not seen any reviews of the 3.1s yet, which is why I find it puzzling that on Aliexpress, KSGER specifically calls out that the tip temperature reading is off.  Since I have found no reviews calling this out, I can only assume they have had quite a few returns over this issue and since the reviews of the 2.1s's indicate the temperature sensor readings are fairly accurate, I am thinking its an issue with the 3.1s firmware.

[dreamcat4]

In addition compatibility for JBC tips although that potentially could be modded on an older PCB. New versions have wapped linear 3.3v regulator for a switching one. Which is going to introduse switching noise on the rails for the ADC. Which is bad for the temperaature measurement. They also cheaped out on the op amp (perhaps also the crystal). In particular the cost reduced part of the op amp is also going to mes up the temperature sensing circuit before it gets into the ADC. So there you have it.

If you want any better answers then I suggest you go ask KSGER themselves for why they keep on change such things. If they care at all for their future reputation? Perhaps not.

[totalnoob]

Thank you, Dreamcat.  So to clear things up, were these changes (i.e. switching regulator and cheaper op amp) made only to the 3.1S version or did they make these changes in the 2.1S as far as the units you have seen (since it is possible they have incorporated these changes into the latest 2.1S offerings, it has been documented they made a mid-model change, and kept the same model designation, to eliminate the R10 issue in the 2.1S's)? 

Thanks.

[dreamcat4]

Unfortunately dude, the KSGER version numbering schemes are not unambiguous. So when we all talk about 'version X.y {S}' it's not completely clear which one we actually mean when we say that. Which really makes these things more difficult to talk about.

Even KSGER themselves has taken a same PCB revision and labelled it both '2.1' and '3.0' simultaneously at the same time. Which just goes to show how totally worthless this version numbering scheme actually is.

But what you can do is look at my research pages here on github. It is a work in progress. I am still halfway through gathering source materials. Mostly translating from other online sources (that is writting in Russian or Chinese). YMMV.

I think the Blue 'updated version' shown here is arguable 'the best'. Because it adds in the corner 2 pads marked for "JBC or T12". Wheras the original green one did not have explicit point for connecting the JBC. Regardless of that connecting a JBC is perhaps a more complicated matter.

https://github.com/dreamcat4/t12-t245-controllers-docs/tree/master/controllers/stm32-t12-oled#v21s-original-version-green-pcb

If you want to browse online into the many other subfolders you will the find further useful research links and information. Including other alternative board designs. And discussion pages which themselves contains a lot of really useful information and explanatory diagrams. Some of those sources have been translated but not been fully analyzed / documented. Because they are far too big. While others are currently missing because they have not been translated, put back together in a reasonably loadable / browsable version file yet.

BTW perhaps some pages view better on Firefox.

[totalnoob]

Thank you.  I know that at least the numbering system KSGER used is not totally to be trusted, as I said, from what I can make out looking at the various Youtube reviews, it appears that the early "2.1S" versions had the R10 that drained the battery while later "2.1S" versions had a different control board layout that did not have that R10 or used a "R10" labeled resistor for a different purpose.  What muddies things is when some reviews came out in the same timeframe but some received units with the R10 that drains the battery and others without the R10 that drains the battery, I just chalked that up to KSGER sending out a mixture of the older batch along with the newer batch.  I didn't think they made as many changes during the same "model" run as you indicated.  That's good to know, but adds to the confusion.  I will look at what you have gathered.

Thanks again.

[floobydust]

It's just that they treat a version number as a model number.
Corrections/changes are made and a board is a bit different- but want it considered the same model.
Also people are copying the design, adding mistakes or making it cheaper, but keeping the root version number.

To make it even worse, there are three version numbers - the controller hardware, the firmware, the soldering station.

Russians are saying V3.1S is KSGER taking control of the firmware, and it's buggy, so they recommend rolling back to V2.1S firmware?

[Acecool]

I bought the 3.1 and I didn't get the new case... they used the old smaller case without the silver front... I was hoping to slide it into a rack and use that to brace it.. but oh well..

The thing says error then heats up, then error, then heat sup...

The mini GX12-5 connector has no screw so the black plastic part rotates freely.. This isn't good especially if there are wires loose. The QuickGo or whatever station has a 4 pin out, the KSGER has 5... It seems like 2 of the pins are ground pins on the KSGER...

I bought mine on banggood and it was $56 USD shipped with a cheap plastic iron which may be the issue I'm having with the unit... I have gotten no support from them in trying to resolve the issue and I've sent them the video.



The case is not grounded, but some of the metal pieces do appear to have a ground.

My unit rattled inside - so I added a few dabs of hot-glue. I still need to add grounding wire. I'm not a fan of the 3.1 variant where the 2.1 variant had plugs from the switch to the board, etc... and a plastic case... the 3.1 went back to direct soldering the spade connectors of the switch and power socket back into the board.

3.1 uses a tiny battery welded directly to the front board instead of a larger battery on a long lead. This doesn't bang around which is nice.



3.1 Grounds: The 3 pin wire grounding pin does go to the tip of the handheld iron. It also goes to 2 of the 5 pins on the GX12-5 pin connector. The threads are NOT grounded on this connector as they didn't add a grounding wire. The metal underneath the rotary encoder is also NOT grounded. The case is NOT grounded. However, in most parts - if you got the BLACK anodized case, it is hard to find a spot which you can actually get current to go through - so that is good... but there are bare metal parts where the screws go in, and scratches or breaks in the anodized material near the corners - front face / rear face, mating points because of cutting, filing, scratches or something else. So it is possible to find exposed areas you can get current to flow through and get 0.00 - 0.02 Ohm resistance through those areas.


Overall - I'd modify mine if it doesn't need to be replaced to add grounds to the inside of the case - actual wires to any exposed metal, the lids / case, etc... and I'd probably also change the GX12 connector for something someone suggested to me for the DE-DER 5000 unit... audio XLR or something connectors as they are high quality, quick release, grounding pin, easy to solder to and has shields.... This would mean that any replacement iron I got would need to be resoldered... or I'll make an adapter for testing...

I will likely also make it my stand-alone rework station by building a box around my fan controller for my hot air gun which is on my rework station now ( and I removed the permanent connection for a GX16-8 connector ) and change it to a 4 or 5 pin connector, inline controller for fan speed, and have the station control heat... Then maybe add pump support with inlet and outlet so I can plug in a desoldering gun with a small motor and also use a hot air micro-pencil too.... then get a few of them so I can use the features all at once.

Then, instead of having 3 stations on my desk, I'll have 1 capable of all 3... or 3 mini stations which are the size of 0.5 of just the desoldering one. I've seen some posts on desoldering gun motors which are absolutely tiny and work just as well... they need the inlet and outlet so that'd work for a hot air pencil with the heat coming from the pencil, and air from the outlet of the pump...

Also, since it is STM32 based, I have already worked with that chipset on another project and recoded a few things so I have the programmer... and I think there are pins available to hook up... so I'll just add the color-coded headers as I did to the USB to logitech adapter and look at the software and see about making it more in-line with what I want...

The press-in and rotate seems a bit meh for changing tips or calibration... and long-press is too long for me. quick tap to change temp is ok.

[Acecool]

Does anyone have the 3.1 hex file? The STLink V2 utility told me to disable read protection to read and I did and I clicked ok, in the options and it decided to read immediately without me telling it to read and it actually wrote and wiped the entire chip. Such an annoyance right now to have the chip wiped when I didn't tell it to wipe the chip, and when I was in the settings menu, and not in the programming utility....


STM32F103RBT6

[totalnoob]

In the interest of keeping all information about the V3.1S in the same thread, I am going to give a preliminary review since I just received it and have not actually used it beyond turning it on and plugging in the handle and a tip.  However, I will comment on the handle I ordered with it.  I went "whole hog" and bought the SS handle.  I can't comment on whether it heats up and becomes uncomfortable when using it after a long period of time.  I can only comment on the tip fit in the handle.  There is no positive "click" or feedback to let me know the tip is in enough and the tip is very easily pulled out.  As a matter of fact, the fit seems loose enough that I am concerned it may just fall out, although that did not happen in my very limited testing (but I did not actually use it).  Not necessarily a bad thing to be able to take the tip out easily, but I was concerned it was loose and that it being loose is why I kept getting an error message as the tip was heating up in my first test and I was not sure if that was because the tip was not seated properly or if it was because it was a new tip the controller was not reading it properly (KSGER's Aliexpress store mentions something about this).  All of a sudden, the error stopped flashing and the temperature reading stopped fluctuating. There was no indication that the tip was loose as I made erratic movements to see if the error came back, it did not.  I will likely order one of the carbon fiber aluminum handles because they tend to get decent reviews, to see if they have a better indication that the tip is seated correctly.  I cannot recommend the SS handle at this time because of not being able to determine if the tip is seated properly.

I do reserve the right to add a reply later to provide updates on what I find I like or dislike. 

[dreamcat4]

Dude,

 

best handle is a JBC handle. The unbranded T245 version (that comes with other JBC clone stations etc.) is completely authentic, but not so expensive. Less than $50 off aliexpress or elsewhere.

For T12 handle I would look for something.... cannot remember now where i saw it. But it was probably some Russian dudes. So if you go into my research folder, and click on those "Accessible Version" (translated) pages. And scroll thru the comments sections in each page:

https://github.com/dreamcat4/t12-t245-controllers-docs/tree/master/research/mysku

You will probably find it being reffered to (eventually, after much scrolling, many comments).

Or you could ask the Chinese guy who makes those new 'Woka' stations. He would probably know too. I have been meaning to email him on his WeChat email address at some point. Never actually got around to it. Oh well!

[totalnoob]

Thanks, I'll have to look up the Woka station.  In this mini-review, I was referring to the T12 handle, obviously.  I'd like to get a fairly decent one, I am sure the JBC is the best handle, but I am really only interested in the ability to convert these stations for use with JBC handles, academically.   And from what I have seen/read is that the clone JBC handles seem to be pretty good quality even if not completely up to JBC quality.  If the clone "JBC" tips had as large of a selection of different styles and were relatively as cheap as the T12 Clone tips (even if they were a couple of bucks more per tip), then I'd probably be more interested in the conversion process as a real project, rather than something to keep in mind as a "someday" project.  I'll check out your folders again, about handles, these were specifically about T12 handles, right?  Were they reviews your archived?     

[totalnoob]

Following up on my mini review, mostly of the handle.  I wanted to confirm that when I played around with it tonight, the tip I had used before worked right, right from the start.  The display did not "dance" between "error" and displaying temperature, so I suspect it only happens to new tips.  This is somewhat confirmed that I swapped tips hot for the second tip that came withe unit and that tip did cause the display to "dance" until it finally settled, just like the first tip.  Other than the handle not having the positive feedback to let me know the tip was seated, the handle works well.  Pushing the tip in until it stops seems to work.  There's no disconnection while waving it or when I tested the tip temperatures, about the closest to soldering I have done with this set, yet. 

A mini-review of the soldering station itself, since I had some time to look through the menus and test out the interface some.  I will admit, I need to d/l the manual and read through it, the interface is NOT user friendly.  In my shop, since I do a lot of woodworking projects, painting, general mechanics stuff, I prefer to have as few buttons on things as possible to press because my hands are frequently dirty with glue, oil, paint, etc. so the fewer buttons on things the better.  However, I think this unit could use a couple of buttons to take some relief off the push/rotary dial.  I mean pushing in and turning left or right while holding the knob in is kind of a pain.  Having a "Back" button to back out to the main menu would be great rather than holding the knob pressed in for several seconds to accomplish the same thing.  As pointed out in another thread, it appears under the Tips menu that the firmware supports JBC tips.  If that is the case, and the tip numbers that I suspect are JBC tip numbers are in fact JBC numbers, then all those tips are just jumbled together, so you have a fairly long list to scroll through with no order, although most T12 tips seem to be grouped together, there are the suspected JBC tips between the groups of T12 tips.  KSGER should have added submenu's and grouped the T12 tips in one menu and the JBC tips together in another menu.  And a third menu option for keeping  the user defined tips settings in.  Anyway, other than those nit picks, the unit seem to work well.  As mentioned earlier, I did not actually solder anything, but checking the tip temperatures on a Hakko FG100 clone, seemed to agree with the first tip's temperature reading on the KSGER (the K tip), but the second tip was off by about 30 degrees C (DL32).  The active tip selected, however, was the default,B1, once I figured out how to change the active tip to DL32, the temperature reading was pretty much spot on (about 4 degrees C different).  I did not re-check the K tip with the active tip set to K. 

[totalnoob]

Stumbled across this video of a person who has apparently tested the various KSGER T12 handles, although he only shows two in his video.  I specifically avoided the black aluminum handle because of John Salt's (another YouTuber) experience with his, but this reviewer apparently found it to be a good handle, once he fixed the assembly issues with the wiring/strain relief.

I post it here in this thread since handles were specifically discussed and for future reference in case.  Although perhaps handles might deserve their own thread. 

[dreamcat4]

The carbon fibre handle, from ksger store on aliexpress. Is supposed to be the best one. It was recommended to me by somebody else. It lets you solder for longer durations, not metal, it will not conduct the heat through into your hand. Unlike the other metal ones.

.... Unfortunately the wiring seems pretty suspect on all of them. I'm sad to say.

[bozza]

In the menu there is a desoldering option.
Does that mean that it is somehow possible to attach a desoldering gun?

[floobydust]

KSGER firmware supports a desoldering gun and pump, but not all of their controller boards have the extra hardware provisions.
I don't believe the OLED-V3.0 or OLED-V3.1 (black) boards have it. The "S" suffix is for desoldering control I think.

Controller Ve2.12S has that provision, as well as a dual-OLED option. Underneath the OLED display there are pads for two SOIC-8 mosfets like SI4800, two diodes like FR107.
Controller Ve2.1S has the mosfets/diodes/connector pads on the component side of the board.
The desoldering gun trigger switch or maybe it's a foot pedal, connects to the shake switch input.
One output is for the 24VDC vacuum pump motor, and the other for a vacuum solenoid valve.

You can also make a power mosfet board and take off the pump/valve control signals from the MCU at PB6, PB7 if you run fine wires from there.