Aixun T3A misbehaving on grounded PCBs

[tony359]

Hi all,

Just got this Aixun T3A and I'm experiencing a pretty severe issue. Basically if the PCB I am working on is grounded, some (not all) joints will cause the station to lose connection with the tip (NO TOOLS displayed on screen).
This happens with multiple tips.

Also the handle is buzzing which is... concerning?

I am not on very latest FW, I downgraded to 1.33 as with 1.34 the temperatures are all over the place.

This is brand new so I can ask for warranty support (wish me luck) but I'm curious if anybody else is experiencing the same or has an idea of what is going on?

Thanks!

[slavoy]

Touching the grounding point with the tip of the soldering iron shorts one of the thermocouple leads in the tip, causing the station to be unable to read the correct temperature during that cycle, which results in reporting a tip error or behaving erratically. It's the same in Quick soldering stations, for example. This is due to the design and temperature measurement method, and apparently, during the temperature reading period, the thermocouple cannot be shorted to the ground.

[tony359]

That’s not the case though. If I touch ground with the tip, nothing happens.

Only if I touch some solder joints of a grounded PCB.

So are you saying this station cannot be used on grounded equipment? I had that board on a pre-heating station - which is grounded. Are you saying I can’t use it in that scenario? It limits the case use of the station massively.

Also, what is the purpose of the ground alligator lead enclosed in the box of grounding the PCB causes that issue?

I’m still very puzzled.

[slavoy]

I think this way you create a loop to ground, which introduces interference into the thermocouple signal.

Also, what is the purpose of the ground alligator lead enclosed in the box of grounding the PCB causes that issue?

It's meant to connect to the ESD point, but internally, it's directly connected to the ground in the case without any resistor.

[tony359]

T3A owners: have you seen this on your station?

[drksy]

T3A owners: have you seen this on your station?

I tried it on my unit. Grounded some board to mains earth and touched the tip of the iron to some ground point. I couldn't reproduce the effect. Nothing happened.

[tony359]

I appreciate the time you spent on this, thanks!

However, it's not ground that triggers the issue.
It's SOME pins on a GROUNDED board. I can touch ground with my tip and nothing happens as well.

As I am measuring 0.95VDC at the tip, that is enough to activate some junctions and I suspect that some combination of components might trigger some unexpected feedback.

[thm_w]

Looks like you've cross posted here: https://forum.allaboutcircuits.com/threads/aixun-t3a-issue-on-grounded-pcbs.196084/
You are measuring 1VDC from the tip wrt ground right? What is the tip resistance to earth?

I tried with Aifen that I've earthed the tip on, and it gives a similar voltage and a heater error if I ground the tip via another path. The stock config was floating tip which was not nice, I prefer this configuration. Its probably a common issue with JBC clones. JBC itself has a ton of circuitry inside to deal with this, in addition to the linear transformer which might have less noise. I remember bskull dealing with a similar issue: https://www.eevblog.com/forum/projects/jbc-clone-with-good-ground-noise-immunity/

What you probably want is a proper ESD mat (not silicone) and then ground the ESD mat and not the circuit board itself.

[tony359]

Yes apologies for cross-posting but I don't get much feedback on this

Tip to earth is zero ohms but only when not heating. When heating I'm reading OL - to measure that I had to go from cold to 400C so the tip is being heated for several seconds.
During the "trickle" heating I see the resistance jumping around but that's my Fluke not fast enough to read OL I guess.

Yes it's 1VDC measuring from tip to ground.

Thanks for the link. Things are getting clearer. However, if I touch ground with the tip nothing happens. It's when I touch SOME joints on a GROUNDED PCB.

Many have commented on why I would need to ground a PCBs and I don't normally do that. But as I said I discovered this issue while soldering a PCB which was sitting on a (obviously) grounded pre-heater. The PCB touches the metal chassis of the pre-heater so it becomes grounded.

If this behaviour is expected, I can take precautions and shield the resting points of the pre-heater. It just baffles me that nobody has seen this before, I noticed that 20 minutes into using this station

Also, wouldn't 1V/1.5A be potentially dangerous for some more delicate components?

Thanks for your input!

[thm_w]

Thats a good point about the grounded preheater, if you have traces near the edge. I guess options for you are:
- Add teflon washers or rails to insulate the board on the preheater, and rely on the station being earthed
- Get another soldering iron

If you measure 1V and shorted out providing 1.5A then yes there is a very small chance some components could be damaged. But if you are also measuring zero ohms to earth, I don't think the current would be that high. There are some measurements in the aixun thread where 0.4A short circuit is seen: https://www.eevblog.com/forum/reviews/any-opinions-on-the-aixun-t3a/msg4417474/#msg4417474 still fairly high I guess...

One of the JBC schematics was reverse engineered here: https://www.eevblog.com/forum/testgear/jbc-soldering-station-cd-2bc-complete-schematic-analysis/
You can see two 0.22 ohm resistors and a fuse from the tip to earth ground, then a circuit that senses earth leakage. I don't know if an Aixun schematic exists anywhere, and if any mods can be done.

[tony359]

thank you @thm_w

You're the first in the whole internet who has actually pointed me to what I think is the right direction.

c0d3z3r0 is measuring the same things I am and he's experiencing the same issues I am. It sounds like there is a reason for this station to cost a fraction of a JBC - what annoys me a bit is that none of the reviews I've watched have mentioned any of this potential issues.

Let's see what Aixun comes back with (meanwhile they have confirmed that my station is genuine at least and they will help if they can), then I'll draw my conclusions.

I did put some kapton tape on the rails where the PCB sits while on the pre-heating plate but inevitably you end up with some metal parts touching something. I guess I'll have to decide whether to live with it or not.

Thanks for now!

[c0d3z3r0]

Thats a good point about the grounded preheater, if you have traces near the edge. I guess options for you are:
- Add teflon washers or rails to insulate the board on the preheater, and rely on the station being earthed
- Get another soldering iron

If you measure 1V and shorted out providing 1.5A then yes there is a very small chance some components could be damaged. But if you are also measuring zero ohms to earth, I don't think the current would be that high. There are some measurements in the aixun thread where 0.4A short circuit is seen: https://www.eevblog.com/forum/reviews/any-opinions-on-the-aixun-t3a/msg4417474/#msg4417474 still fairly high I guess...

One of the JBC schematics was reverse engineered here: https://www.eevblog.com/forum/testgear/jbc-soldering-station-cd-2bc-complete-schematic-analysis/
You can see two 0.22 ohm resistors and a fuse from the tip to earth ground, then a circuit that senses earth leakage. I don't know if an Aixun schematic exists anywhere, and if any mods can be done.

Schematics can be found here: https://github.com/c0d3z3r0/aixun_t3a_rev/blob/master/schematics/aixun_t3a.pdf

[c0d3z3r0]

I believe that the grounded tip + heating through DC/PWM instead of AC + the way T3A is measuring temperature is the root cause.

This is how the cartridge is connected:

GX12-5 pin                              handle / tip
5 +            ---------------- heater (middle)

4 GND        ------+--------  GND (top metal)
                          |
3 ID           ------+
 
2 T(EMP)    ----------------  temp (lower spike)


Temperature is measured between the outer connector (GND) of the cartridge and the spike / center connector.
Heating happens through outer connector (GND) and the middle connector.

JBC stations measure between the outer connector (GND) and the middle connector, while heating is done through middle connector and center connector. (Oh well, there are different opinions on that topic btw. Some people say JBC is heating through middle connector - outer connector. When I looked at the reversed engineered JBC schematics, I had the impression that it's middle + center)

[tony359]

@c0d3z3r0

Thanks for your reports on this issue - as I said you are the first I am reading about that.

What is your opinion on this, do you see this as a major issue with the T3A or just something I should try to mitigate by insulating the PCBs I am soldering? And what about the voltages and currents you measured? I measured 1.5A in DC between tip and ground, I believe you only measures 0.4A.

Thanks!

[c0d3z3r0]

@c0d3z3r0

Thanks for your reports on this issue - as I said you are the first I am reading about that.

What is your opinion on this, do you see this as a major issue with the T3A or just something I should try to mitigate by insulating the PCBs I am soldering? And what about the voltages and currents you measured? I measured 1.5A in DC between tip and ground, I believe you only measures 0.4A.

Thanks!

Wow, 1.5A o.O

Well, I usually don't ground any boards I am soldering. Since I also solder highly-ESD-sensitive stuff and I have not yet fried any board (or... at least I'm not aware of any issues yet - ESD might cause damage that only shows at some later time), I would say it's fine for me.

However, if anyone finds a proper solution/fix/hack/... for these weird behaviours of the T3A, I'd be more than happy

[tony359]

I don't either - my pre-heating board happens to be grounded though, that's how I discovered that!

Thanks for your input, gives me a little peace of mind!

Are all those copycat cartridge stations like that? Is the famous KSGER the same? Just wondering.

[thm_w]

As noted above, Aifen was worse if anything (tip is ungrounded), and then the same behavior after the mod. Ksger I have one portable one, same issue where the temperature can bounce around due to induced noise.
With a pure T12 clone you wouldn't have this issue as the tip earth and heater are completely separate, so the T12 KSGER can be OK (assuming the tip is earthed inside).
There is an "i2c" station with a toroidal transformer but I don't know the results of that one. Aixun has their high end model as well.

Its a tradeoff with the JBC design I guess, to get that performance the thermocouple has to be bonded electrically to the tip, making isolation difficult.

c0d3z3r0 is measuring the same things I am and he's experiencing the same issues I am. It sounds like there is a reason for this station to cost a fraction of a JBC - what annoys me a bit is that none of the reviews I've watched have mentioned any of this potential issues.

Yes, as you can see with codezeros schematic, its built to be about as cheap as possible in terms of control and measurement. Single opamp single FET.
If you are a professional, don't mess around with cheap clones, get a name brand station.

I'm still not sure why the current is this high though.. the secondary is isolated right? The tip should be earthed via an isolated supply so it can float. But I guess the issues are all related to inductive or capacitive coupling in this case?

The filtering in the Aixun schematic: C27 then R33, if that is right it seems backwards, I would expect RC filter on the temperature sense input not CR. In the JBC schematic you can see two op-amps, and an normal RC filter at the output of the second one.

[tony359]

Thanks.

I'm not a professional, I really cannot justify spending £500 on a JBC.

As long as I am not damaging anything, I'm happy with the T3A.

The T420 looks interesting - but will it be free from bugs? The T3A has stellar reviews online, nobody has mentioned any issues. So not sure I want to experiment with another mode.  As I said, as long as I am not damaging things, I can live with the T3A. Opinions?

[tony359]

stumbled into this comment on a T3A review video.

So clearly not just me.

[c0d3z3r0]

Its a tradeoff with the JBC design I guess, to get that performance the thermocouple has to be bonded electrically to the tip, making isolation difficult.

Well, the genuine JBC station uses AC for heating, which probably simply does not interact with DC GND / earth. I guess the main problem is using common GND for earth, heating and the TC. It's just a gut feeling, though.

The filtering in the Aixun schematic: C27 then R33, if that is right it seems backwards, I would expect RC filter on the temperature sense input not CR. In the JBC schematic you can see two op-amps, and an normal RC filter at the output of the second one.

[c0d3z3r0]

The filtering in the Aixun schematic: C27 then R33, if that is right it seems backwards, I would expect RC filter on the temperature sense input not CR. In the JBC schematic you can see two op-amps, and an normal RC filter at the output of the second one.

Not sure if that is meant to be a RC filter. Check out this one: https://taaralabs.eu/tag/t245/
"Series resistor and reverse diode on the input of the Op-Amp are used for the protection against possible voltage spikes from the iron."

[LuisBe]

Hi, I don't know if my contribution will help but I recently had the following problem with the ground of my workbench.
The output of my signal generator (it was on but the output was disabled) gave me a voltage of 3Vac (measured with a multimeter). The problem is that an old power supply (switched off but connected to the same socket) was feeding 10Vac into my ground and some of it was going to my generator. I removed the power supply cable and my ground was clean. I don't know if you have tried just having the soldering station in the main socket and doing the test.
Best regards and good luck.

[thm_w]

Thanks.

I'm not a professional, I really cannot justify spending £500 on a JBC.

As long as I am not damaging anything, I'm happy with the T3A.

The T420 looks interesting - but will it be free from bugs? The T3A has stellar reviews online, nobody has mentioned any issues. So not sure I want to experiment with another mode.  As I said, as long as I am not damaging things, I can live with the T3A. Opinions?

Yeah you'll probably be good then.
I wouldn't spend £500 on new JBC either, the T420 is $230, will it be bug free, who knows. I already buy used Metcal for ~$240 so I'm not interested in the aixun at that price point.

Not sure if that is meant to be a RC filter. Check out this one: https://taaralabs.eu/tag/t245/
"Series resistor and reverse diode on the input of the Op-Amp are used for the protection against possible voltage spikes from the iron."

OK, makes sense. Though any external signal coming into an onboard ADC I would expect an RC filter to also be present. The signal we care about is very slow (temperature).

Well, the genuine JBC station uses AC for heating, which probably simply does not interact with DC GND / earth. I guess the main problem is using common GND for earth, heating and the TC. It's just a gut feeling, though.

Yeah most likely you are right.
Whether it matters if its DC or AC, I don't think it does as long as switching is not too hard. With DC, you would have a main 24V supply that only goes to the heater. Then an isolated 24V:5V brick to run your mcu and control. And differential sensing of the TC, not using common ground. Might work.

[mastershake]

i will not be back to the office till at least tues or wed but i will test this on the 420 and see. also we have 14 of the t3a here so ill test a few of them and see if i assume it will but more then one will do this same thing. i dont ground like code says i will to the esd mat but that does not seem to effect it like this.

[thm_w]

Good teardown of a Weller station was just posted: https://www.eevblog.com/forum/reviews/weller-wxr-3-rework-station-teardown/
They are using DC and have a FET for each output (24V, 12V, and 12V). So clearly DC itself is fine. Although they've still gone with the linear transformer instead of a switch mode.