Any opinions on the AiXun T3A?

[Ungolian]

Any time something seems to good to be true, human nature begs the question "What's the catch?".  Are there any performance/firmware issues?   Is there something different about the handle?  Maybe it gets hotter than average?  Same tip to grip distance as the JBC?  Other than a few (possibly) questionable joints someone may get in their units, and cheap Chinese filtering caps, could this possibly be a unicorn that does everything as advertised and is cheap??!!

[mastershake]

someone else i know ordered both the 245 and t12 and they will def put them through their paces at work. he runs jbc stuff all day long and he will def tell me if these will keep up lol. i just found out tonight he has them on the way should be a week or so before they arrive. he works with his jbc and metcal all day every day he will for sure nit pick them to death lol. he said he will open them up fully and try to get some psu pics.

[SteveyG]

Any time something seems to good to be true, human nature begs the question "What's the catch?".  Are there any performance/firmware issues?   Is there something different about the handle?  Maybe it gets hotter than average?  Same tip to grip distance as the JBC?  Other than a few (possibly) questionable joints someone may get in their units, and cheap Chinese filtering caps, could this possibly be a unicorn that does everything as advertised and is cheap??!!

The handle is the same as a genuine JBC. It does get warm, but that's why JBC now supply a foam sleeve.

The main consideration is it's supplying the heater directly from a 24V DC supply with no current limiting. If the MOSFET fails, the tip will die, if the temperature sensing fails the tip will die, plus we don't know the long term effects of effectively having a non-current limited source supplying these tips

[FransW]

In view of recent developments with soldering stations I "carefully" conclude that the "classical" stations are pricewise and qualitywise blown from the market.

However, with the exception for the professional side and use.

Frans

[BlackICE]

Any time something seems to good to be true, human nature begs the question "What's the catch?".  Are there any performance/firmware issues?   Is there something different about the handle?  Maybe it gets hotter than average?  Same tip to grip distance as the JBC?  Other than a few (possibly) questionable joints someone may get in their units, and cheap Chinese filtering caps, could this possibly be a unicorn that does everything as advertised and is cheap??!!

The handle is the same as a genuine JBC. It does get warm, but that's why JBC now supply a foam sleeve.

The main consideration is it's supplying the heater directly from a 24V DC supply with no current limiting. If the MOSFET fails, the tip will die, if the temperature sensing fails the tip will die, plus we don't know the long term effects of effectively having a non-current limited source supplying these tips

A thanks to SteveyG the great video review.

I guess the same can be said of the T12 clones like Ksger and Quicko. If nothing goes wrong they will work well. If certain failures happen then you lose a tip or worst start a fire. I have a Ksger and I like it much better than my Weller 1010NA. Quicker to get to temp, easier to change tips, shorter tip to hand length and a sleep function that works under real use. Now I only pull out the Weller to remove some SMD using 2 solder irons like chopsticks.

These T3A look attractive to me, but so far I haven't needed the extra power and buying a set of real JBC tips would add considerably to the total cost of ownership. I doubt if the clone tips will function as well as the real deal.

[Ungolian]

Sounds like a mod to the power source may be in order.  Probably not any room on the board itself, so maybe a daughterboard?  Possibly a more robust MOSFET?  I don't have one yet, so I can't really comment on the one that's already in there.  Just wondering about mods to improve durability. 

And thanks, Steven.  I watch all of your vids, and appreciate the work you are doing.

P.S. I wouldn't mind seeing you do another flux video.  Soldering isn't always pretty.  I do a lot of work on older audio through hole stuff, lot's of oxidation.  Everything in your current flux video is fine for nice, clean, SMD work, but not so great for things I do.  Rosin flux- MG Chemicals, Rectorseal Nokorode, Delcast, SRA, Radio Shack, LA Co, and anything else you think appropriate for problem soldering.   

[thm_w]

Interesting D2 comes from the USB port Vbus, so they might have thought of powering the iron from that. Or perhaps just for during the prototype phase.

Here is the FET: http://www.aosmd.com/pdfs/datasheet/AO4413.pdf
I wouldn't expect it to get much over 70C, probably less, as normal soldering uses low average power. But you could replace it with something bigger if you like, the gate is driven by a transistor.

[DavidAlfa]

Not the best mosfet for the job.
They could have put an AOD403 which handles 55A, TO252 package is marginally larger than SOIC8, so no excuses.
Doing so they save some pennies to invest in proper advertising:

[mastershake]

i did find out (from the company) they do claim you can use them on either 220 or 110 they claim there are not really 2 different models. i went ahead and ordered one to take a look at. i may go ahead and change out the mosfet for something better we will see.

[Ungolian]

As SteveyG said, that MOSFET and the temperature sensing are the only two things keeping the tip from going nuclear.  Makes sense to improve what you can. Seems like a fairly easy fix.

[dreamcat4]

Sounds like a mod to the power source may be in order.  Probably not any room on the board itself, so maybe a daughterboard?  Possibly a more robust MOSFET?

this ^^ 

[DavidAlfa]

I doubt any controller, even jbc, handles such situation.
If the MOSFET shorts out, you'll have a bad day.
But it should be sized to make sure it won't happen easily.
Don't get paranoic!

[wraper]

I doubt any controller, even jbc, handles such situation.
If the MOSFET shorts out, you'll have a bad day.
But it should be sized to make sure it won't happen easily.
Don't get paranoic!

JBC uses transformer, not SMPS, so output voltage is not stable. Current sensing in that case can be helpful to measure actual power delivered and adjust on/off time accordingly to deliver desired amount of power. Or to detect type of cartridge used based on current consumption. Most likely the only thing it can protect is electronics in case when cartridge is shorted.
I don't have JBC but ERSA I-CON has current sense as well. It won't help at all if pass element (TRIAC) is shorted.

[DavidAlfa]

Current detection can't protect anything from disaster. If the mosfet shorts, all the power from the transformer will go into the tip, no matter what.
Sure, the controller will read 20amps.
And what can it do to stop it? Maybe showing "Error" in the screen? Beeping?

[exe]

Beeping?

Beeping is better than nothing. Fire alarms also cannot extinguish fire...

[DavidAlfa]

Fire alarms are usually connected to water sprinklers .
Beeping would be pretty useless. You have very little time to react, in 10 seconds, before you even realise what's going on, the tip will have melted.
*beeeep* What's that? Why is this beeping? Error code... what? Thermal runwaway? Disconnect pow...? Where's that orange glowing coming from?

Burning the mosfet is a really rare condition when driving resistive loads (within mosfet working limits).
That's why you don't see cheap T12 stations burning everywhere. It can happen, but not a common thing.

[exe]

Fire alarms are usually connected to water sprinklers .
Beeping would be pretty useless. You have very little time to react, in 10 seconds, before you even realise what's going on, the tip will have melted.

It's not to save the tip, it's to save the rest of the lab. I don't mind having this feature if it doesn't cost much extra.

[DavidAlfa]

You are supposed to solder in a proper place, some materials can already burn at normal soldering temps.
The tip will glow red hot for 10 seconds and die, simple as that. Nothing will happen if placed in the stand.

[dreamcat4]

Yes this is all true. The tips overheats so fast. So adding a current sense does not help. (My mistake)

I suppose the right type of a solution would be to have an extra independent safety cut off (another 2nd mosfet or relay). That is only for this task. If the MCU detects the tip temperature is exceeding a certain preset safety value (for example 450c or 500c / whatever you like). And also that the has temperature kept increasing during an overshoot time, when the main driving mosfet is actually already been switched off. That the temp kept rising for a time window period that is longer than the longest hysteresis / catch up time of the thermocouple. So to avoid false detection of latent heating to propagate and be seen.

However this is also assuming you are using (for example) a genuine and high quality JBC tip. A tip is not have problems with the TC disconnecting itself. Because that failure would render this method useless. Like inside those cheap T12 chinese tip, which for that happens in those ones very often.

OK so this is the same safety feature is already done in 3d printer world. For the hot end safety cut off, and sometimes the bed heater too. However on a 3d printer thermal mass is typically much larger, so it has significantly more time to detect / react. However maybe it can work here too. Depending on how difficult / easy to catch during these much shorter time period. (The basic fundamental principle is the same however).

Yet to justify the cost of adding this safety feature... you are indeed balancing the chance of failure, against the extra cost of the 2nd safety mosfet (or relay). And extra coding in the firmware. Which adds to the upfront cost of each station. Against the cost of loosing 1 tip (for genuine JBC, might be about $25 YMMV).

Of course MOSFET does fail sometimes, but it's not so often they fail. Making this something of a "luxury feature" i guess? Still, quite an interesting topic to discuss.

But like you say earlier (above)... it's a lot easier. Simpler and more sense to just start with the best possible MOSFET you can use. As your main heater mosfet. And don't use dodgy unreliable mosfet to begin with. And accept the risk for loose $24 one time.

[DavidAlfa]

Sure a simple master switching relay would do it.
Anyways not a major concern unless something else blows the mosfet (bad wiring, defective tip..)

[dreamcat4]

well the way i blew up my mosfet was because there was a fault in the handle. Causing an intermittent short in the handle between the 2 heater terminals. So if you get some a poor connection. Or getd damaged over heavy usage, the cable strain relief etc. Things like that

[DavidAlfa]

I remember that .
One of my tips melted the internal insulator and the two wires shorted.
The mosfet didn't blew, but the tip became almost red hot at the middle!

[thm_w]

well the way i blew up my mosfet was because there was a fault in the handle. Causing an intermittent short in the handle between the 2 heater terminals. So if you get some a poor connection. Or getd damaged over heavy usage, the cable strain relief etc. Things like that

If you are worried about this type of thing in a design, use a smart FET with guaranteed short circuit and overtemp protection. Siliconwizard used TLE9201SG, I'm not familiar with that part. Another user used VN series automotive high side switch, has overtemp, overcurrent, ESD, short to ground, short to VCC, etc. protection. Its what I will try to use, overkill but cheap as its a legacy part.
Not something easily modded into an existing design though, you have to implement it in your own design or an add-on board.

[Ungolian]

Just some quick notes.  The microcontroller is a GD32F303RCT6 which is comparable to the STM32F103RC.  The GD32 has a 120Mhz clock, while the STM32 has a 72Mhz clock.  BUT the STM32 has 5 more A/D controller channels (21 vs. 16 for the GD32).  So while the GD32 can't handle as much analog input , it is 60% faster than the STM32.  Both are perfectly fine for the jobs they are doing.

https://www.gigadevice.com/microcontroller/gd32f303rct6/
https://www.st.com/content/st_com/en/products/microcontrollers-microprocessors/stm32-32-bit-arm-cortex-mcus/stm32-mainstream-mcus/stm32f1-series/stm32f103/stm32f103rc.html

As for the mosfet replacement, doesn't the AOD403 DPAK have the wrong pin out?  The source and gate are inverted.  Alpha and Omega Semiconductors are the only ones that make the D403.  I'm sure there are other mosfets that will work, but for those that know just enough to be dangerous (like me), a direct swap to the AOD403 would have the full transformer voltage going into the gate, not the source.  Which is bad.

[thm_w]

Just some quick notes.  The microcontroller is a GD32F303RCT6 which is comparable to the STM32F103RC.  The GD32 has a 120Mhz clock, while the STM32 has a 72Mhz clock.  BUT the STM32 has 5 more A/D controller channels (21 vs. 16 for the GD32).  So while the GD32 can't handle as much analog input , it is 60% faster than the STM32.  Both are perfectly fine for the jobs they are doing.

https://www.gigadevice.com/microcontroller/gd32f303rct6/
https://www.st.com/content/st_com/en/products/microcontrollers-microprocessors/stm32-32-bit-arm-cortex-mcus/stm32-mainstream-mcus/stm32f1-series/stm32f103/stm32f103rc.html

As for the mosfet replacement, doesn't the AOD403 DPAK have the wrong pin out?  The source and gate are inverted.  Alpha and Omega Semiconductors are the only ones that make the D403.  I'm sure there are other mosfets that will work, but for those that know just enough to be dangerous (like me), a direct swap to the AOD403 would have the full transformer voltage going into the gate, not the source.  Which is bad. (Attachment Link)

Likely the extra clock as you said, for driving that high res LCD. If you watch SDGs other vid, JBC had a PIC33 driving the LCD, and the update was glacial.
Prices used to be good, but now its terrible: https://lcsc.com/product-detail/GigaDevice_GigaDevice_GD32F303RCT6_GD32F303RCT6_C131443.html
And the real STM32F303 is a completely different beast as well, enhanced high speed ADCs, etc. So its not worth most of us using GDF303 IMO, as its not directly compatible.

If you are swapping a SOIC8 for a DPAK, then, its not going to be a "direct swap" regardless, and you should expect to do some bodging .