Yihua 853D soldering station - details about fan circuit repair, incl. schematic

[mosagepa]

Hi there all EE vblogers!

I've seen other threads here regarding internal details and some repairs on this kind of units.

My 853D stopped suddenly blowing hot air, I suspected the fan. So first I rushed to buy a replacement, although later on it turned the TIP122 was to blame. Don't know even why it failed on me!

Anyway, here are the details I've gathered in this experience.

My subsequent posts in this thread include up to now the following:

- First off, internal PCB overall views
- Schematics  (of the fan control section only for now, but I'm recollecting more sections and hand-drawing them for my own notes.)
- Pictures of section in the test bench with lil' breadboard for controlling fan and test
- Full size FRONT and BACK (mirrored for you to align if you wish) views of the full PCB.

In the first 2 pictures we'll see the PCB just as the unit was opened, with all the harnesses to/ from panel and different subsystems still in place.

As anyone familiar with these units might know, they have been several revisions (or should I say clones and re-clones, even between various Chinese suppliers themselves ---oh well, competition is good they say!).

My particular 853D has NO RF sensor in the front panel, but otherwise looks externally identical to many ones with same model designation. I got this from Amazon, not AliExpress or the like (yeah I got Prime and was quite eager those days to get it!)

Here follows --again, just for THIS particular PCB revision, silk-marked as "853D-8S19-V3"-- a summary of the purpose of the connectors you've seen in the previous two pictures:

[ ---- to be updated --- ]

[mosagepa]

Overall pictures re-posted here with lower resolution so as not to deprive resources
 

[mosagepa]

Edited first message to add the PCB silk-name: "853D-8S19-V3".

I haven't seen similar PCBs in all other threads I was browsing thru my initial quest for info.

If anyone out there bought his 853D from Amazon or some other "non directly Chinese distributors", maybe got the same PCB model, let me know please.

CNs on this main board:

Code: [Select]

  --> Left to Right on TOP EDGE:
CON15 (3-pin)  [BROWN-WHITE-GREEN]       from/ to HANDLE FAN
pin 1: [BROWN] (square pad) =  FAN+ (RED from FAN inside HANDLE)
pin 2: [WHITE] GND          =  FAN- (BLACK from FAN ins. HANDLE)
pin 3: [GREEN] CON10 pin 13 =  FAN SENSE (YELLOW from " " ins.")
         (ONLY THESE POINTS, AND NOT ANY OTHER/S, GET CONNECTED)

CON1  (3-pin)  [BLUE-YELLOW-RED]         from/ to HANDLE HEATER
pin 1: [BLUE] =   to mini-PCB inside handle, 'SENSOR' pad
pin 2: [YELLOW] = to EARTH lug for heater inside handle
pin 3: [RED] =    to mini-PCB inside handle, 'HANDLE' pad

CON9  (power,2-pin)   [GREY-TRANSP]      to IRON HEATER  (24VDC ???)
pin 1: [TRANSP.] = to HEATER active voltage input
                pin 2: [GREY]    = to mini-PCB inside handle, 'GND' pad
  (NOTE: also connected inside, to heater's EARTH lug.)

CON14 (power,2-pin)   [BLACK-RED]        AC (post-fuse) to Transf.
CON12 (power,2-pin)   [RED-RED]          from Transf. INPUT
CON2  (power,3-pin)   [RED-WHITE-BLACK]  from Transf. 20V-10V-0V OUT
CON5  (power,3-pin)   [BLACK-WHITE-BLUE-n/c]  from ext. trans. 2N
  --> Top to Bottom on LEFT EDGE:
V2    (3-pin)  [WHITE-WHITE-WHITE, red marks] from PANEL FAN POT
CON11 (4-pin)  [RED-RED-YELLOW-YELLOW]  from Transf. 30V & 10V OUTs
CON8  (2-pin)  n/c ???
  --> Left to Right on BOTTOM EDGE:
CN10  (sic) (long; 13-pin)  from PANEL (identify component lines!!!)
CON4  (5-pin)  [WHITE-WHITE-n/c-BLACK-RED]  to/ from IRON SOCKET
CON6  (2-pin)  [BLUE-BLUE]  from Transf. 24V OUTs
CON3  (3-pin)  [WHITE-WHITE-WHITE, red marks] from PANEL DC SOURCE POT.
pin 1 (square pad): to TL431 volt.reg.ctl. (for DC Source)
pin 2: to U3-A (LM358) Pos. Input
pin 3: GND
CON13 (2-pin)  [WHITE-WHITE, red marks]         to PANEL RED LED
CON7  (power,2-pin)  [BLACK-RED]                to PANEL DC METER



[mosagepa]

Here comes picture of the labelling on the fan that "failed" (I made sure I got the precise same labelled replacement, this time yep it had to be from AliExpress).

As you may notice, this have 3 wires, (red-black-yellow), but it's NOT a brushless motor. The driving circuit as we'll see in my home-made schematics is just a TIP12x-activated circuit, very common and easy to understand.

Although the transformer yields 30V and 10V (AC, rectified by respective 1n4007 bridges), the fan is 24VDc nominal. A convenient zener in the circuit limits the maximum attainable feed into it (see schematic).

 

[mosagepa]

Here comes the hand-drawn schematic of the part of PCB that controls the fan blower.
It includes: TIP120 (replaced the former TIP122, only difference is TIP122 allows for max. 100V, TIP120 "just" 60V.)  Both rated 5A max I think;  2x2N3904-kind SMD for control. The Zener, and the front panel POT (B-10k) which is interposed with a 15k to control the range of fan velocities.

I'll try to pass this hand-drawing to a more decent, say Kicad or some medium if I get the time.

You're free to comment anything, or ask questions about this station. I'm willing to answer or locate anything you need or are interested in this PCB or unit.

(In fact it's way best and easier NOW that I have ti fully opened and in my bench...)

I'll keep posting further details and pictures of it.

Kind regards.

[mosagepa]

Now let's describe a bit operation of this part of the unit, the fan control schematic, for anyone interested (although a lot of forum followers here will find it almost self-explanatory):

I'll paraphrase/ adapt/ reuse here a quite plain description found here for a similar model (858D), for it mostly applies here with a mere re-annotation of components:
https://blog.enbewe.de/posts/2018-06-21-saike-858d-original-schematic/


"The fan is controlled through transistor U9. The fan is rated for 24V, but the supply voltage is a bit more
than 30V. The circuit has to make sure that the fan is not overloaded. The microcontroller can use
transistor Q13 to drain the base current of U9 and turn the controller off this way. If Q13 is turned off,
the speed control is implemented through Q14 and the surrounding circuit. The more Q14 is conducting, the more the base current is drained and the less U9 is conducting. However, Q14 cannot drain U9 completely, as the emitter voltage of Q14 always sits one diode drop above ground level."

"POT (B10K) which is the front-panel potentiometer, can be used to control the voltage divider between R55 and R59.
The higher the voltage at the taper of the potentiometer, the higher the base current of Q14, resulting in
less conductance for U9. In addition the aforementioned voltage divider and the transistor form a feedback loop to control the base current of U9, and in result the voltage on the fan."

[mosagepa]

Some pictures for fun, of my "bench" when I was testing just this part of the PCB.

To test just the fan feeding section, it suffices to connect headers:
- V2 (3-pin) to the Panel POT B10kOhm  (using original harness)
- CON11 (4-pin)  (red-red-yellow-yellow)  which feeds the 30V, 10V AC respectively from the transformer into the PCB.   (again original harness)
- CON14 (2-pin)  (black-red)  (gross wires, original)  These are the mains wires that come from the fuse assembly into the PCB.
- CON12 (2-pin)  (red-red) (gross wires, original)  "copies" CON14 points to the outside of PCB towards the transformer's main AC input
- CON15 (3-pin)  (brown-white-green)  (FAN+, FAN-GND, and FAN "sense" wire, respectively).

Now about the replacement fan wiring (the original one I dismantled even if original TIP122 was suspected and replaced, too).
As it came with just stripped wires, I had to make a JST-to dupont 3-wire harness to feed it adequately. Notice that the yellow wire I am not using in this "experiment" to test the fan.

One of the pictures show a little breadboard which serves me to feed +5V / GND signal using the pushbutton.  If you feed GND to CN10 pin 9 on the PCB,  the fan runs at a speed regulated by the POT. If you feed CN10 pin 9 with +5 volts, the fan stops as explained in the above post.

CN10 pin 11 is GND.  You can see the duponts attached there.

[mosagepa]

Oops, missed the pushbutton little PCB for testing pic. Here it comes.

Along with aligned FRONT and BACK (mirrored) pictures of the PCB for you to analyze and drool over 

[mosagepa]

I present my excuses to anyone that feels my hand-drawn schematic is weird. Also I must point out to the fact that some components seems re-soldered (in a provisional way), see that 2W resistor for example which is tied grossly.

(I had to take some components apart like those big Rs and then refit them, as they didn't allow me to clearly see marking on SMDs and stuff...)

I am yet to clean things up and close the unit one of these days. So feel free to comment anything before that.

About "cleaning", don't get me started about the TONS of flux residue that were on both sides of the PCB!!!  Spent quite a time taking off most of it.  It's really shameful how you can receive some items from Chinese cheap assemblers... I got once an FPGA development board that got so many residue you just couldn't clearly tell out some of the pins!

[fzabkar]

The base of Q14 is held at 8.8V by the zener and Vbe (8.2V + 0.6V = 8.8V). The base is connected to the wiper of the pot. The fan voltage is then determined by the resistances in the potential divider formed by the three resistors.

When the pot is set to minimum, the fan voltage is ...

37K / 22K x 8.8V = 14.8V

When the pot is set to maximum, the fan voltage is ...

37K / 12K x 8.8V = 27.1V

The maximum dissipation in the Darlington pair occurs at minimum fan voltage, so ...

Pmax = V x I = (30V - 14.8V) x 0.2A = 3.04 watts

It seems to me that the transistor is more than adequately rated for this application (65W).

https://www.onsemi.com/pdf/datasheet/tip120-d.pdf

Would it be worthwhile to add a reverse biased 1A 60V (?) Schottky rectifier across the fan terminals as a back-EMF snubber?

[mosagepa]

Many thanks @fzabkar for your input and prospective calculations.

Where do you get/ infer that the Zener is actually 8V2?  Maybe the markings on the pictures?  (I don't have perfect eyesight sorry)  I must confess I didn't measure it  (heck, I am not that used even to *any* testing of Zeners, at least in-circuit,  so a side question is: how could I actually measure it?)   I mean, without powering the PCB, of course.

It's strange that the circuit would allow, according to your computations, more than 24 volts applied to the fan, since it's rated nominal 24V and normally one would not even want to drive it at full voltage, only a tad less  (to allow for tolerances and the like.)

The TIP120 I've soldered replaces a former TIP122. I know both would happily allow this small a load, and all ratings for this precise design. In any case, the fan label warns "0.2A" so it would be good exercise to check that's indeed the case here...?

Do you happen to own a 853D/D+?

[fzabkar]

DZ1 in the original schematic is 8V2 (8.2 volts).

https://blog.enbewe.de/posts/images/858D/858D_Schematic_Fan_Control-big.png

It wouldn't be possible to measure the zener in-circuit without powering it.

I don't own an 853D/D+.


Edit: The calculations assume that the base current drawn by Q4 is much lower than the current in the divider resistors.

[mosagepa]

Oh I see, that schematic I remember, is from 858D ("original" as you mention).

Differences are mainly, that the 853D+ I own uses 30V. The transformer is a bit different and has differing output branches from the 858's. Also the voltage divider around the pot has bit different values. Base resistor to the overall on/off transistor maybe doesn't affect too much.

[ozkarah]

I have a Yihua 853D (15V 2A version). Today the auto-stopping (when the handle is put to the dock) function of the hot air gun stopped working.

It continues to heat the air gun while the handle is in its dock until set temp is reached. Cooling down never starts.

I checked the magnetic sensor and the wire it comes from the sensor to the board (CON 1 blue wire). They seem to be okay. When that handle is in the dock, the blue wire is grounded. And vice versa. It is not grounded when the handle is removed from the dock. 

Apart from the auto-stop function, the rest seems to work okay:
- Heating stops when the set temperature is reached (but cooling down does not start till I manually turn off the hot air switch).
- If I manually turn off the hot air gun switch, the cooling down starts, and the fan stops at 100 degrees Celcius. 

Do you have any idea to fix the problem?

[puterboy]

I have a Yihua 862BD+ Soldering plus Hot Air rework station.
A few weeks ago the fan cut out when I was using the hot air gun on high heat for prolonged time.

I took apart the hot air gun handle and confirmed that the fan was dead (by testing on a separate power supply) and assumed that this was the problem and that the high heat had caused the fan to die.
So I ordered an exact OEM replacement from Yihua on Aliexpress.

When I installed the replacement fan, it worked and the hot air gun heated up for about 3-5 seconds before the fan shut off and the hot air display went from displaying numerical temperature to '--'
If I shut off the hot air portion, let it cool for a few seconds, and turn it on again, the process repeated with the fan turning on and blowing hot air for a few seconds before shutting off.

I don't think it is actually overheating as the temperature is set at 100C and it doesn't blow long enough to even get really hot.

Rather something is causing the station to shut off both the fan and heater circuit prematurely.

Any thoughts on what might be the issue?

[puterboy]

I also noticed, that the magnetic reed switch works BUT it doesn't seem to be cutting off power to the heater -- hence the heater continues to heat even with the reed switch triggered.

The two TRIACS and the BJT controlling the fan both seem to be OK -- so not sure what is wrong...

[puterboy]

Actually, turns out that the fan supply TRIAC (BTA12) was bad. So I replaced it with a higher current BTA16-800B plus added heat sink compount.
All now works.

However, the fan speed display flickers when the speed is less than about 80%.
Any ideas what might be causing such problem?

[roadwarriormax]

IM HAVING EXACTLY THE SAME PROBLEM WITH MY 852D+ ,HOT AIR WORKS FINE .BUT ONCE ON CRADLE .THE HEAT SWITCHES OFF BUT THE FAN KEEPS ON RUNNING .YOU COULD,NT TELL ME WHERE THIS TRIACS  IS IS IT ON THE MAIN BOARD OR INSIDE THE HEAT GUN . ANY INFORMATION WOULD BE GREAT  .THIS HAPPENED 5 MINUTES AGO ,BEFORE THAT ,IT WORKED GREAT .

[ozkarah]

Try to change the state of "Auto" toggle switch at the left bottom of the front panel

[roadwarriormax]

Big thanks to puterboy . my W.E.R  852D+ had lost it,s air flow control ,and would,nt shut off when put into the cradle . my 852D+ is the model without an internal air pump . thanks to puterboy for highlighting that the TRIACS could be the problem .  i,ve noticed some models of 852 or even 853 can have different amounts of TRIACS inside ,my model had 4 inside , BTA 16 600B ,TIP 122,BT137 600E,L7805CV . so after replacing all TRIACS and replacing the 50v capacitor , as mine was bulging , the old W.E.R 852D+ is working perfectly again .everything back to normal ,so thank you to puterboy for replying to my messages ,to be honest i would never have known, what the solution to my problem is ,if i had,nt seen your comments and posts 

[auron]

Another thank you from my end to everyone contributing to this thread. It was the top google hit for my search when my 853D hot air died, and ultimately helped be recover.
Reviving the thread because I believe I also have some more information for others that might find it the same way I did.

I found it very hard to find the replacement fan for the "YHFS HDS-501524H" used in my S218A hot air handle. the price for just the fan was actually almost as expensive as ordering a brand new S218A from Amazon - so I did that. Unfortunately, when I hooked it up to my 853D, the fan was killed within the first 40 seconds as I was testing higher speed. :/

With now two dead hot air handpieces I decided something must be wrong with the base itself. So I took a dive into the schematics from above, and though I have a slightly different PCB version again (853D_USB_ZB_V6) the fan circuit is essentially the same.

Yet, I noticed that the two SMD resistors before and after the potentiometer, R8 was 15k and the other was 12k. The values would match the circuit, but they were swapped against the original design.
12k was going to ground and 15k to the TIP122. This leads to the voltage on the fan exceeding 28V and I believe is the cause for some dead units out there.


When I swapped the resistors, the voltage was reliably limited to 22ish volts.

With that fixed I had to tackle the fans:
24V 5015 fans are easy to get a hold of, but 3-wire ones are pretty rare.
I tried to non-destructively open the fan, but it broke anyway. This revealed the broken chip within, a OCS477H Single Phase Intelligent Hall Effect DC Fan Driver.
I had some other 24v fans as spare for my 3d printer, so I took one of them apart and it had a similar chip within. So I soldered that in my original fan, and sure enough it started working again.
A little superglue plus some decent amount of hot glue and the fan came back together. it might not hold forever, but it's good for the time being. Hot glue seems to be counter intuitive in a hot air station, but as it's on the intake fan, I think it should be fine.

Taking the fan apart, also revealed a second repair approach, by adding a third cable to a two-wire fan.
This is what the motor looks inside



Just drill a small hole in the back of the fan, and solder the sensor cable to one of the coil connector solder points.


It was easy for me to find the right spot with the broken one to use as a template, but the solder points were actually visible from the side.
You just need to make sure to find the one that is not the FAN+

The "ears" on the enclosing for mounting with a screw were easily removed with a saw.

I tested both methods, and they work similarly well.



Edit: formatting trouble with the images

[auron]

Ah, made a mistake in above circuit diagram. Measured the resistance and the FAN+ is actually not connected to anything in the coils, the solder joint seems to be for stability only, not being a simple stepper as I assumed.
So the schematic looks like this:


And that actually matches the 477H datasheet
Repair approach is still valid, though