AxxSolder - JBC soldering controller

[AxxAxx]

Since it started a discussion about my project, AxxSolder in another forum thread: https://www.eevblog.com/forum/reviews/any-opinions-on-the-aixun-t3a/msg5115714/#msg5115714
I thought it would be best to break out and create a separate thread here on EEVBlog Forum.

So, AxxSolder is my attempt to create a soldering station (period  )

I want it to be simple and capable.
With so many soldering iron projects out there it is difficult to come with something new, but with a background in Mechanical design and Mechatronics, I thought I could pull together something that doesn't only feels good to use but also performs well for a hobbyist or for me to have on my desk.
Being used to working with the JBC range of irons, it is very difficult to go "back" or even choose something else (no offence Metcal and HAKKO users, you are also "nice"). Having the range of tips and irons that are available is truly amazing and gives you confidence over the competition.

The goal is not that AxxSolder should support every soldering iron in the world, nor be the cheapest. But it should perform well, be open source and perform equally or better than competitors. I have for example a bunch of ideas I can see as improvement possibilities from the 930 Euro JBC DDE-2C soldering controller. The boot up time of ~6 seconds, the slow update rate on the screen, it is slow to change the temperature by pressing +/-, the temperature rolls over from 400 deg C to 250 deg C, annoying. It can be done better.

So, let's keep this thread as a place for technical discussion and see what can be achieved. The last discussions were about the best way of grounding/earthing a soldering iron that is powered by either battery, floating or negative grounded SMPS..

The project page can be found here: https://github.com/AxxAxx/AxxSolder

[thm_w]

Incredible, good 3D print, very good documentation, high side driver, ESD protection, 1% dead time thermocouple measurement.

I assume power limiting for T210 is in there somewhere? I didn't look closely the code.

[AxxAxx]

Since there are many sources online on how the JBC cartridges are internally connected and they tend to not all agree I thought it would be good to clear this out once and for all, at least for C210 and C245.
A good thing about designing your own soldering station is that you may end up with broken (overheated) cartridges from the early development

The cartridges I have are C210-029 and C245-945. I grinded away half of it and took some microscope pictures of the cross sections. I hope below images speak for themselves. Why there is a void in my C245-945 is another question, could it be from me overheating it? maybe..


C210-029 below

C245-945 below

[UPDATE]
From external measurements it is not obvious how the internal configuration of the C245 looks.
One can easily be tricked by measuring the generated voltage between BLUE and GREEN while heating the cartrigde tip by an external heat source. One will then read a voltage of close to zero volts and probably draw the conclusion that there can not be any thermocouple element between BLUE and GREEN..
Upon closer inspection of my cross sections it is obvious that there are in fact 3 points where different components are joined together within the cartridge. And as we know, a thermocouple is just two dissimilar materials joined together, thus there are 3 thermocouple elements (shown as TC1, TC2 and TC3 in above image) within the tip. To prove that there are 3 thermocouple elements within the C245 tip I did some probing under a microscope while heating the cartridge tip with hot air to a temperature of aprox. 200-300 degree C. The voltage measurements were done with my FLUKE 87V in mV mode. I realize that even the junctions between my gold plated probes and the cartridge becomes a TC which should be compensated for, but for this measurement I neglect this effect.

It can be seen that the 3 thermocouple elements are present. From the measured voltage one can aslo derive their corresponding relative seeback coefficient which shows that C1≈C2+C3.
This in turn shows that the total seeback voltage one gets by measuring BLUE to GREEN is cancelled out by the three thermocouple elements in series thus yielding a total measured voltage of ~0 V.

This was also presented by user DerDaniel https://www.eevblog.com/forum/projects/jbc-handle-cartridge-data/msg3580774



For the C210 cartridge it can be shown in similar way by probing the cross section while heating the tip that the tip contains two thermocouple elements. But as these are in series they do always appear to be only one TC from external measurements.

[AxxAxx]

Incredible, good 3D print, very good documentation, high side driver, ESD protection, 1% dead time thermocouple measurement.

I assume power limiting for T210 is in there somewhere? I didn't look closely the code.

Thanks!
Yes the software limits the power depending on which handle is detected:
At the moment it is done in this code section:

Code: [Select]

/* If the handle_sense is high -> T210 handle is detected */
if(sensor_values.handle_sense > 0.5){
handle = T210;
max_power_watt = 60; //60W
Kp = 20;
Ki = 60;
Kd = 0.5;
}
/* If the handle_sense is low -> T245 Handle */
else{
handle = T245;
max_power_watt = 120; //120W
Kp = 30;
Ki = 60;
Kd = 1;
}
PID_SetTunings(&TPID, Kp, Ki, Kd); // Update PID parameters based on handle typeOn the ESD protection topic I think it could be done better, there is an active "issue" on the github page, meaning I'll probably make an update on that when I get time.

[thm_w]

Yeah 60W is around what the clone stations seem to use. I don't know what JBC themselves use, unless you had measured that to confirm or got that number from somewhere else.

Interesting to see the construction difference between the 210 and 245, and the void or blowout.. whatever that is. I wonder if thats filler material at the end.
I don't think anyone has posted a 245 clone tip cutaway yet, just some where the tip had completely fell off, they might have huge voids as well.

[c0d3z3r0]

Picking up the dicsussion regarding proper earthing of the AxxSoldere here again:

Ah, well, the problem is not mains earth but the fact that when PE + GND are connected, you get a current from the heaters positive through the TC to the grounded tip

Precisely. When using a dc supply, the best setup seems to be
  -connect the tip to earth
  -the thermocouple will then be referenced to earth, so connect the control board supply negative also to earth
  -use a floating and isolated smps for the heater and turn on and off its mosfet with an optocoupler

One can use an off-the-shelf isolated dc-dc converter to get the 24V down to 5V, for example something like PQQ6W-Q24-S5-S (https://www.mouser.com/ProductDetail/CUI-Inc/PQQ6W-Q24-S5-S?qs=zW32dvEIR3sbKuWv4BvdLw%3D%3D) to run the controller board.

After some research I too believe that using an isolated DC/DC converter for the MCU side seems to be the best option. However, all the measurement and detection needs galvanic isolation as well, so you don't introduce that ground connection again:
- both temperature sensing and current sensing need to be done with an isolated op-amp
  - some I2C isolated temperature sensing IC?
  - e.g. TI AMC1301 (or the cheaper AMC1101?) or Renesas RV1S9353A for current, but both are expensive. Maybe use a current sensing IC like ACS730?
- stand sense through optocoupler
- heater control through optocoupler
- handle sense (id pin) through
  1) optocoupler if only checking for T245 vs. T210 or
  2) through a) isolated op-amp or b) linear optocoupler as a future option to support other handle types (probably not the plan anyways, is it?)

So, the AxxSolder could then have a 3 pin supply connector VCC,GND,EARTH and it's up to the user what supply to use:
- Toroidal transformer, rectified: connect VCC, GND + optional EARTH
- Isolated SMPS (SELV / non-earthed) + optional EARTH
- Isolated SMPS (PELV / earthed secondary) + optional EARTH (or just bridge GND/EARTH if you're 100% sure you have a PELV, not SELV supply)

Another thing to note: without coupling primary and secondary side of the DC/DC converter, it will float (at the potential of the primary side, right?). I'm unsure if this creates any new problems.

[drksy]

That's not quite what I meant. I wanted the SMPS ouput to be floating, and then only ground the 5V supply for the controller, so that the heater return current isn't connected to earth:



This way, the opamps and all the electronics are referenced to mains earth which makes temperature sensing or stand sensing hassle free. You just need an optocoupler to drive the mosfet for the heater.

[c0d3z3r0]

Well, yes, but that would only make it work with a floating SMPS, not with a grounded one. My extension of your idea make it work with *any* SMPS.
Also, I'm unsure if your setup with floating SMPS would be compliant with regulations in various regions (e.g. Germany, I doubt it).

[drksy]

Well, yes, but that would only make it work with a floating SMPS, not with a grounded one. My extension of your idea make it work with *any* SMPS.

True, at the cost of adding significant complexity.

Also, I'm unsure if your setup with floating SMPS would be compliant with regulations in various regions (e.g. Germany, I doubt it).

The heater power supply could be capacitively coupled to earth, like for example in RPS-200-24. Then that cap will also be across the thermocouple. Would that cause problems? Or could just be a transformer-rectifier-capacitor combo or batteries. I've seen many laptop power supplies or generic brick supplies for printer or such that have no earth connection anywhere.

In the Meanwell LRS lineup, the output is floating, yet Y-rated filter caps are used on the primary side to ground. For example see LRS-200 I assume the ground symbol with circle lableled "FG" stands for frame ground, meaning it's the metal chassis. And if that's bolted to earth, then the mains input has caps to earth, presumably for filtering/emi reasons. However, the output is not anywhere tied to earth, so I don't think that's strictly a safety requirement. I've modified my diagram to make this more clear.

[drksy]

Another suggestion to make AxxSolder even more accessible to the DIY community. I'm thinking what if someone wants to use a different display like the Nextion or even a 16x2 that they already have laying around, so having an I2C and/or SPI header on the PCB would be super useful. Having looked at your code, one would have to just modify `update_OLED` in main.c.

[AxxAxx]

Yeah 60W is around what the clone stations seem to use. I don't know what JBC themselves use, unless you had measured that to confirm or got that number from somewhere else.

It is indeed not so clear what power should be used for different cartridges, 60W/120W comes from various sources online and my own experience.
I did however some more research in the topic and found from JBC:
- CDN Station (C115 Cartridges) Peak power (Tool) 25W https://www.jbctools.com/cdn-high-precision-soldering-station-product-2274.html
- CDS Station (C210 Cartridges) Peak power (Tool) 40W https://www.jbctools.com/cds-precision-soldering-station-product-1580.html
- CDB Station (C245 Cartridges) Peak power (Tool) 130W https://www.jbctools.com/cdb-soldering-station-product-1605.html
- CP Station (AM120 tweezer => two C210 cartridges) Peak power (Tool) 80W https://www.jbctools.com/cp-precision-rework-station-product-1618.html

So if this is correct I could probably bump up the max power of C245 cartridges a bit from the current max power of 120W for C245 cartridges..

Interesting to see the construction difference between the 210 and 245, and the void or blowout.. whatever that is. I wonder if thats filler material at the end.
I don't think anyone has posted a 245 clone tip cutaway yet, just some where the tip had completely fell off, they might have huge voids as well.

The material you can see next to the void in the image is just a thin section of cupper, it could be "peeled away" and the void is also under it.

[AxxAxx]

True, at the cost of adding significant complexity.

It seems like c0d3z3r0's idea above would be a robust way to handle all different power solutions. To a cost/complexity indeed.
For the heater control there are also Optically Isolated Gate Drivers designed for this purpose that one could consider.

I think if the selection of power supply should be up to the user the grounding should work on whichever SMPS is used. It is not always clear for the user if your SMPS is floating or grounded...

- handle sense (id pin) through
  1) optocoupler if only checking for T245 vs. T210 or
  2) through a) isolated op-amp or b) linear optocoupler as a future option to support other handle types (probably not the plan anyways, is it?)

I have also in my mind that it would be nice to be able to drive JBC tweezers. Then you need a second Sense pin to detect tweezers. (And a second heater/temperature sense circuit)
It is always nice to add features but sometimes simplicity is king. Don't really know where that limit is for AxxSolder yet though..

[AxxAxx]

Another suggestion to make AxxSolder even more accessible to the DIY community. I'm thinking what if someone wants to use a different display like the Nextion or even a 16x2 that they already have laying around, so having an I2C and/or SPI header on the PCB would be super useful. Having looked at your code, one would have to just modify `update_OLED` in main.c.

Good idea!
Adding a I2C header is simple, and SPI is already there.
Then it could be up to anyone that want another display to branch out and crerate a driver for that particular display.
Adding a I2C header also makes AxxSolder future-proof in case the selected OLED should go out of stock.

The current status is that I have added some USB functionality. I want a programming interface over USB so that you do not need to have a SWD programmer to flash sofware onto AxxSolder. Luckily STM32G431 has native support for USB botloader and this should be easily implemented. I have also added a USB-C - PD negotiation circuit (STUSB4500 https://www.st.com/en/interfaces-and-transceivers/stusb4500.html) with communication via I2C to the MCU. I'm not sure if this is a good idea but I had the circuit design ready from another project and I'll give it a try.  I mean 100W is quite respectful and I think that 100W capable USB-C supplies will be more and more common in the future.


The PCBs are ordered an I'll go on a 4 week vacation as of tomorrow. When I get back and have tested the USB solutions I think the grounding questions are next in line 

[miket6000]

Very nice project, well done.

Just a word of caution about using the BAT54s as ESD protection, specifically on the handle. The leakage current on Schottky diodes is huge, at 25°C and 3V3 it's nominally 50nA and can be up to 10x worse, as it gets hotter it gets worse again. This leakage current will flow through your 470k resistor and can create large voltage offsets affecting the temperature measurement accuracy.

I'd recommend using a low leakage ESD diode instead, something like a MAX14541 has a maximum leakage of 1nA (typical is less than 10pA). There are certainly many other options as well.

Cheers

[drksy]

It is always nice to add features but sometimes simplicity is king. Don't really know where that limit is for AxxSolder yet though..

That's always the challenge. If isolation can be done elegantly for grounded heater supplies, I'm all for it. But it sounded like it requires obscure isolated opamps and dozens of added components and surrounding passives. I hope you can figure out a good solution, thought I don't think it's confusing or too much to just ask users to avoid supplies whose negatives are shorted to mains earth, as this is the case with majority of power supplies.

Really like the look of that new fuse holder though. That car fuse had to go for sure. I wonder if you can merge the two USB ports into one? Can one port provide power delivery as well as be used to upgrade the firmware? That vertical micro-usb looks like an eye-sore. Maybe a USB switch chip with a select jumper? USB3740B could be a candidate. Cheap and readily available.

[Pooya]

This is my personal opinion, The box has to be redesigned. The power port does not have any lock that can stand completely in the enclosure and must be fixed with some glue. All screws holes and some walls are so thin and can break so easily, you can use heatset insert for screws, The Encolsure_JBC-AD-SF does not have bottom plate, XT60e-m should be a good option for power port.

[AxxAxx]

I wonder if you can merge the two USB ports into one? Can one port provide power delivery as well as be used to upgrade the firmware? That vertical micro-usb looks like an eye-sore. Maybe a USB switch chip with a select jumper? USB3740B could be a candidate. Cheap and readily available.

Yes it doesn't seem optimal to have two separate USB Ports. The design above will work as my test board and have Both USBs connected to the MCU. On this board there are solder jumpers to turn on/off USB-C data.. It's in a development branch so noone except me should make this PCB version.
The reason I did not go with only the USB-C is that I do not know how the firmware upload will work with different USB supplies. If you have a proper USB-C PD capable computer then it should be fine. But if you only have a 5V capable USB-C cable and want to flash new software, as the USB-C is connected to VDD to power everything I'm not sure how the DC/DC converter will handle that as it is outside the specs of min in voltage. By having a separate micro USB I can always know I get 5V from there and feed that through a diode and to the 5V bus on the board.

I am also not sure how the USB-C PD circuit will handle the case when you have 24V from a separate supply and plug in a USB-C cable.. Have to investigate that,which I can do on this board.

It is also a bit unconvinient to program the station version if the usb is horizontal on the board.. I think a redesign of both enclosures so that you have a USB-C port available for programming and power would be best.

[AxxAxx]

This is my personal opinion, The box has to be redesigned. The power port does not have any lock that can stand completely in the enclosure and must be fixed with some glue. All screws holes and some walls are so thin and can break so easily, you can use heatset insert for screws, The Encolsure_JBC-AD-SF does not have bottom plate, XT60e-m should be a good option for power port.

Yes totally agree with you. If AxxSolder turns out to be popular and more people are starting to make one a redesign should be done. The selection of PP45 Connector comes from that I have used a brunch in other projects and I have batteries with them. It has been on my list to redesign the mounting of that connector. A panel mounted XT60 could be an option. However what I like about PP45 is that they snap together with a positive feedback. I personally prefer a PP45, I could design a panel mount for that. There could maybe be a enclosure version for Xt60e too. At work I would probably use something from Molex, TE or similar industry brand but thoese crimping terminal are not nice if you dont have the correct crimp tools.
Regarding the thin sections, I assume you mean around the display, there they are very thin in order to fit the display flat to the top surface. I printed my cases on a Markforged printer with their Onyx filament. I assume a more hobby-level printers will struggle more with some features.. I have create an Issue on github for this. Feel free to comment on that issue if you have more ideas regarding this topic. It will be done  after my 4 week vacation.

[Pooya]

I did some changes to portable version with thicker face plate, XT60, heatset insert and rocker switch, removed the 4 screw hole (OLED mount) from faceplate and removed the metal piece for sleep mode because I want to attach it to my chinese stand, It is compatible with JBC handle with input and output, useless for me! You can attach the pcb to body from inside. Later I will work on the other one.

[Pooya]

Modified - Final

[Warhawk]

Don't you have an extra PCB?
I am too in Germany.

[Pooya]

I have, the domestic shipping cost is between 8 to 10 euro, I paid total 6 euro include shipping VAT everything for 5 pcbs from JLCpcb, If is worth it then I can send.

[AxxAxx]

I wonder if you can merge the two USB ports into one? Can one port provide power delivery as well as be used to upgrade the firmware? That vertical micro-usb looks like an eye-sore. Maybe a USB switch chip with a select jumper? USB3740B could be a candidate. Cheap and readily available.

The reason I did not go with only the USB-C is that I do not know how the firmware upload will work with different USB supplies. If you have a proper USB-C PD capable computer then it should be fine. But if you only have a 5V capable USB-C cable and want to flash new software, as the USB-C is connected to VDD to power everything I'm not sure how the DC/DC converter will handle that as it is outside the specs of min in voltage. By having a separate micro USB I can always know I get 5V from there and feed that through a diode and to the 5V bus on the board.

I found the answer to my own question about what happens when the USB PD countroller can't get the asked for voltage or current, in the datasheet for CYPD3177:

SAFE_PWR_EN Gate Driver
The EZ-PD BCR device has a SAFE_PWR_EN gate driver that can be used to drive an alternate load switch/FET. It is enabled whenever the EZ-PD BCR device is unable to negotiate the requested power contract. In such a scenario, the EZ-PD BCR device negotiates a 5-V/900-mA contract which can be delivered through the SAFE_PWR_EN FET to an alternate power rail in the system. This allows the system to operate in a limited mode when the requested power is unavailable through the USB Type-C port.

Perfect! Thats exactly what I needed. This controller is also cheap, around 2 Euro.

[drksy]

That's a brilliant solution! I was just thinking about a similar thing to switch the 5V with a mosfet if you detect a PD failure over i2c. And since negotiating power delivery over USB-C apparently doesn't involve the data lines D+ and D-, it makes it even simpler. Nice.

[c0d3z3r0]

@drsky Regarding regulations, the fact that gloating supplies are availavle doesn't mean that they are allowed in any use case. (Not all use cases allow choosing any one of SELV, PELV, FELV). However, in this specific one (solder station) both SELV, PELV seem to be ok.

Btw. Having a SELV SMPS and then grounding the secondary of the DC/DC converter still makes the whole construct PELV, even though the grounding is done bc of functional, not safety reasons.

I just had another random idea, that needs further evaluation, though: instead of all that isolation complexity from my post, why not have a second smaller uC, that only does measurements and heating control in the 24V circuit, communicating over isolated I2C with the main uC.

@AxxAxx why not use the native USBPD controller of the STM32G?

Also, I don't think any MOSFET for power switching is required but just two diodes separating power inputs (main power input, USB-C). This way both inputs just go straight to the 5V DC/DC regulator. (4.5V minimum, so ok with the default 5V)