KSGER Mini T-12 station DOA

[Cecil]

This soldering station arrived DOA.  I'm positive my 24V supply has a mating 2X5.5 mm plug with center positive polarity.  There has never been any sign of life.  Upon case disasembly I noticed D2, a two pin chip, approx. 1mm X 2mm package, marked T4 on top positioned next to the coin cell.  It measures 3V at one end but an ohm meter confirms it is open circuit.  In circuit I do not measure a diode junction.  Is it reasonable to assume this is a bad diode or do I need additional information?  What other terminology do I need to shop for such a part?  Should I look for other faults that caused or resulted from the bad diode?

[Algoma]

There are many different versions of that power supply board, are you able to post a picture of the area in question?

.. That DC jack was across the output of the internal 24V supply.. it can be used as either an DC input to power the unit, or act as an output to power other devices.

The device in question could be a fuse and marked like F1 or F2. Many of these units lack a ground to the case, so your measurement reference where you place your probes will affect the readings you see. Beware the power supplies inside these units often exclude all safety in exchange for saving a few pennies of cost.

[Cecil]

     This is the "Mini" version of the KSGER T-12 soldering station.  It does not have an internal 24V supply or any connection to AC mains.  The DC jack is the only source of power except for the 3V coin cell.
     I wish the two pin device I described was a fuse.  However, the silk screen label for the position of the chip is clearly D2.

[Cecil]

     It was easy to find listings for the diode in question: 1N4148W T4 1A 1000V SOD-123 Switch Diode.  But I'm still interested in information about making a thorough diagnosis.
     I've never soldered something so small so that will be an adventure all its own.

[DavidAlfa]

Post pictures ...

[Cecil]

     I just finished reading the dozens of other KSGER T-12 repair threads in the "Product" rather than the "Repair" section. 
     Pictures coming.

[Algoma]

There are so many varients, knockoffs and copycats of these units, its hard to find the "genuine" ones, that may actually care about product quality. Most of them compete to see how many components they can leave off and still function like the others, long enough to have you accept delivery of a "working" fire hazard. They even rip off logos from the more known models. The photos of the PCB inside gives it away if its genuine.

Those hot air stations can be extreemly sketchy for build quality, look for sales listing that actually show the real PCB and inside of the unit.. Signs of acceptable build quality are actually using locking connectors, instead of soldering under sized jumper wires between the Power supply and the PCB. The size of more valuable components like the transformer is a dead giveaway on how much they cut corners.. and if anything is hot glued inside, its not going to last.

[Cecil]

I hope these pictures help. 

[floobydust]

Diode D1 looks like it has a hole in it, so there might be a short circuit downstream.
U2 is LM317EMP as the 3.3V regulator (max. 40V in).
Measure across U2 input-output ohms/diode test to see if it has shorted. This would also kill the MCU and the board is dead.
I would inject say 3.3V power to the U2 output (tab) and see if the current is too high or not, or the controller comes up.
The solder bridge between pins 3,4 on the GX connector is OK, it's connecting two grounds.

[DavidAlfa]

The crystal placement (The metal cylinder over u4) is an absolute recipe for disaster.

[Cecil]

     I saw the "hole" in D1.  It is a chip in the plastic and not very deep.  In diode test mode I measure 1.96V one way and 1.61V the other.
     Measuring U2 in diode test mode: 60mV between Adj. and tab, either direction; between input and tab, 719mV in one direction and 497mV in the other.  In ohms mode: 188 ohms between tab and Adj., the same as the nearest resistor; open between tab and input.
     With 24V power applied: D1 has 24V on one side and 422mV on the other; the regulator output (tab) is 459mV; in both cases using DC ground at the GX connector.
     I'll try injecting 3.3V at the regulator output, measure the current and look for life.

[Cecil]

     I applied 3.3V to the regulator tab.  The circuit drew 34mA, beeped and the display came on with ERROR displayed because no iron was attached.  I can see the regulator output is incorrect but how do I conclude D1 is the fault or U2 is the fault or both are faulty or that I should look for further faults?

[floobydust]

It's good news the controller powers up and is not damaged 
The 3.3V voltage regulator U2 has no GND pin, only two resistors, one to ground- so it could not short in a way to overload D1 and make it explode.
It's a puzzle what happened. Could be just inrush current charging big cap C13- if it is connected after D1. I see a (via) from D1-K to somewhere else. I would beep it out and see where it goes, maybe to a ceramic capacitor as a wild guess. Look for any more parts connected to D1 cathode.
Then I would replace D1 and carefully ramp up voltage/current to see if the new D1 is going to pop or not.

[Cecil]

     The cathode of D1 is connected to the voltage regulator input and the positive trminal of C13.  I tested everywhere and could not find any other connections.
     I guessed at the part number and ordered some diodes.  Is 1N4148W in a SOD123 package correct?
     I gather the hole in D1 obviously indicated the diode had exploded.  I've seen mosfets blow open their package but this is far more subtle than that.  It is a mystery what could have happened.  When I received it I plugged a 24V power brick to the DC jack and concluded it was DOA.  There was no noise, smell or smoke to indicate the application of power caused a component failure.
     floobydust, I have to thank you especially for your attention and your help.  I think I've read all your posts about these soldering stations and gained alot of respect for your expertise.  I'm grateful.
     Should I do anything about the oscillator?  It does look precarious but I can certainly reinstall the board without letting it short to anything in range.  Is that the only concern?  Is a sock made of heat shrink tubing a good or bad idea for such an oscillator?

[floobydust]

If the 24V power brick was on, and then you plugged in T12 power it would blow the diode. This is because of the high inrush current to charge C13 so fast.
1N4148W is not strong enough 500mA surge rated as (yes it's SOD-123) parts are small. I would use a stronger diode that fits the pads...
Try an SMA part (extra 1mm long) S1D 30A surge, or ES1D

The crystal I'm lazy and would leave it, but it is kinda floppy people just hot glue it to something as long as the tube does not short to anything. Or the heatshrink as long as you don't kill it with heat.

[Cecil]

     It is going to take a shoehorn to get an SMA diode on the pads but I can at least try.
     Would you have any concern about the quality of C13?  I happen to have a Rubycon 8,000 hour part with the same values.

[floobydust]

I think C13 has low stress and not important, compared to the capacitors in the brick.

[Cecil]

     I took the advice to substitute a Vishay RS1A-E3/61T SMD diode for the original 1N4148W SOD 123 part at D1.  Soldering was a little tricky because the SMD diode overlaps the pads at both ends.  Using my best reading glasses I got it in place, though not particularly flat, and it works.  I tested it at 12V, then 19V, then 24V with the DC supply that apparently blew the diode in the first place.  I've run it through a couple of dozen power-ups and all the menu options and everything seems normal.  Thanks again to floobydust for the excellent diagnosis and repair advice.

[Algoma]

Actually one trouble with the loose crystal is that the leads are likely to eventually break or crack at the solder joint with vibration and time. If the unit stops working in the future, its an item to keep in mind to check for.

[Cecil]

     Thanks for the heads-up.  There is some blank PCB under the end of the crystal.  Perhaps I'll build up a cradle to support it.  As it is the heat shrimk tube makes solid contact with the underlying chip to support the base of the crystal.