EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: DAVEISGREAT on December 23, 2022, 05:51:07 am
-
By sound I've isolated these two very suspicious unmarked components, and would like to give a go at, first time, replacing SMD parts. Buzz/whine is emanating from one or both. When I touch them the wine almost goes away and voltage climbs to between 1-3V from zeroish. I am not much of an EE guy but have had good success repairing through hole stuff. Will practice on my collection of boards first and hopefully find a similar FET in the heap.
Most concerned about matching parts since the markings are not there, rubbed or worn off. The back of the case says output is only 200mA. What FET part number would you all recommend?
-
Try reading the marking by shining a light or torch on it from different angles and use the 'zoom' magnifier from your phone in picture mode. It has helped me out several times already.
As you write that the noise stops when touched try reflowing those FETs. With some flux and a soldering iron, don't need an oven or heating plate for this. It might be a simple bad contact issue.
There seems to be some manual rework done already near DZ202. The solder on the pin of some through hole component nearby might have create an unintended short.
Can't really suggest a FET type but it seems these are small signal FETs in SO23 package.
-
I don't think the sound originates from them. It most likely comes from the transformer on the other side of the PCB. Your touching of a switched mode power supply in operation is VERY dangerous and you might die, even after it was disconnected from the mains! Be aware of that.
What happens when you touch those FETs is that some values for the feedback change, so the transformer sound changes as well when switching frequency and duty cycle change.
I do not recommend anyone not familiar with electronics to work on such power supplies. They are dangerous and people have died from it.
-
Try reading the marking by shining a light or torch on it from different angles and use the 'zoom' magnifier from your phone in picture mode. It has helped me out several times already.
As you write that the noise stops when touched try reflowing those FETs. With some flux and a soldering iron, don't need an oven or heating plate for this. It might be a simple bad contact issue.
There seems to be some manual rework done already near DZ202. The solder on the pin of some through hole component nearby might have create an unintended short.
Can't really suggest a FET type but it seems these are small signal FETs in SO23 package.
Thx for torch tip. I can now read them but it is just two big letters(SW) and my google search came up empty.
Previously, I did pull D202 diode to check out of circuit(no short just shoddy soldering), and most of the the electrolytic can caps. All tested good. I will eventually pull one or two of the FETs to test out of circuit (can't find my chipcrete solder, and with my greenhorn SMD soldering skills don't want to try without).
FETS package size is very close to SO23 but width is slightly larger. No luck in finding any FETs in similar package, yet?
Will reflow FETs when, as mentioned, numbnuts finds his chipcrete. :palm:
Thank again - much appreciated :-+
-
I don't think the sound originates from them. It most likely comes from the transformer on the other side of the PCB. Your touching of a switched mode power supply in operation is VERY dangerous and you might die, even after it was disconnected from the mains! Be aware of that.
What happens when you touch those FETs is that some values for the feedback change, so the transformer sound changes as well when switching frequency and duty cycle change.
I do not recommend anyone not familiar with electronics to work on such power supplies. They are dangerous and people have died from it.
Transformer was my first guess too but then read that Ts are often blamed in error so before trashing decided to check other components. I still want to test FETs out of circuit but also don't want to trash em while doing so, especially since I don't know part numbers, and so far have not even been able to source FETs in similar packages?
Thanks for safety warnings. Those higher volt caps on the AC side are definitely scary but I do check for voltages on em without power. Before touching the FETs on the DC side I did cringe and close my eyes(aka numbnuts). Next time I will probe with plastic - thank you. :-+
-
Frequently it is possible to test FET's in circuit with a multimeter. Ohm's setting, enough video's on the tube about how to do it. Only if unsuccessful measurement in circuit it is time to remove and confirm.
I meant to write SOT23 (with a T like transistor)
About the SW marking: https://smd.yooneed.one/code5357.html check out the last one in the list (SOT323 package)
-
Frequently it is possible to test FET's in circuit with a multimeter. Ohm's setting, enough video's on the tube about how to do it. Only if unsuccessful measurement in circuit it is time to remove and confirm.
I meant to write SOT23 (with a T like transistor)
About the SW marking: https://smd.yooneed.one/code5357.html check out the last one in the list (SOT323 package)
That is awesome! How did you find that part number when google shoots blanks?
The datasheet for part number gave the me the source/drain/gate and youtube video tells me both are good so you guys saved me from de-soldering them, nice.
So the whine/buzz is loudest near the fets, and on the other side of board is loudest at the transformer but not as loud as near the FEts. Attached is a picture of the small T. It has eight pins on the other side of board. Numbers on Google comes up with diddly squat, again?
Even if you again can magically find a replacement I'm wondering if it would be cost prohibitive. Maybe though if you could find one with its specifications, in my "heap" of dead boards I may find a similar size/configuration. Not too familiar with these smaller switch mode transformers though. Any suggestions on how to proceed would be greatly appreciated.
-
Careful! The transformer is on the primary side. A small inadvertence and it's gonna tickle you soo seriously that you won't be able to login anymore to tell us about it.
This guy explains how IPD chips work and shows a schematic. Your primary side will be very similar.
https://www.youtube.com/watch?v=FgUzKy9sdIM (https://www.youtube.com/watch?v=FgUzKy9sdIM)
It is not because that transformer makes noise that it is dead. Likely it is just being loaded more or differently than it was designed to be. Search for a (near) short on the secondary side. The FET's you checked are beyond the Vout/GND of the schematic shown in the video.
-
So if I understand correctly you are saying the transformer is blown because another component(s) failed on secondary near transformer?
I did not replace big cap c201 there before, because it read 407, should be 470, and ESR was very low, at least from cheap component tester. Tried replacing this time with known good cap and unfortunately got the same result (led indicator flashes on for a second or two, then goes off, and have same noise). The large diode next to it that I tested out of circuit before, and you thought may have been shorting also tests good. I did clean up the solder blob very well there, but again still same results. Most of the components nearby on secondary are resistors that all look good and the caps are small and don't even have markings on them? There is a large four legged part that bridges the primary and secondary that reads L1124, then on 2nd line reads 817b. 817b datasheet is a photocoupler but I don't know what the function would be on a charger or how to test it?
Would appreciate further advice, and really appreciate all your help, but understand if you are getting tired of helping this novice out.
UPDATE: larger resistor #R230 with marking of R470 on its case, or 47 ohms reads .8 ohms on my meter (not smart enough to know if this resistor is affected by nearby components)?
Also, the two small resistor on side of board r233,r219 with markings that look like 010 read 10 kilo Ohms. Again, don't know if nearby affect or even I am reading these two marking correctly?
As for all the near upstream small caps, no clue cause no markings?
-
I'm not saying the transformer is dead, but it might be.
Interesting is that you write the LED blinks for a second as that would indicate this stuff works at least for that amount of time. What is the purpose of the LED, showing power on status or charging status? It has 3 leads, so bi-color. Have you tried charging the battery again? It looks as it is a smart charger, so don't expect to measure a big potential on the output terminals till a battery is connected.
The optocoupler is a EL817 (not certain how relevant the 'B' is. In the schematic on the video this optocoupler is IC301. It is shown as a transistor on primary and LED on secondary, both with lightning arrows. It is a feedback loop to tell the IPD chip what is happening on the secondary.
-
Oh boy, maybe barking up the wrong tree again. Unfortunately, I don’t have the battery. I’m trying to charge a different 18v lithium pack whose charger was run over by car and smashed to bits by my wife. Still, the noise has to indicate a problem, no? Also, I originally tried shorting charger pins for 18v but that had no effect. Maybe if I put a resistor load on one of the pins that will yield a useable charging voltage?
There is only one led and not sure if current behavior is correct. Initially, I just hooked up my good 18v battery to only this charger’s pos/net terminals thinking it should charge. Sorry, if I omitted relevant info, not familiar with smart charging?
-
First of all, let’s sum up the symptoms:
Your charger makes a kind of unhealthy noise and the output voltage is zero except a short LED flash.
When you touch certain areas of the PCB, an output voltage of ~3V appears.
Does the tone/noise also change?
The noise can be generated by magnetic components or ceramic capacitors which may not be faulty.
Strange noise can be generated due to an unstable regulation loop as switching power supplies usually operate above the audible range.
Also open load, overload or short circuit conditions can result in audible frequencies.
Semiconductors are unlikely to be touch sensitive, opposite to ceramic components, especially capacitors.
I suspect a broken ceramic capacitor that either overloads the secondary voltage or influences the regulation loop.
Or a broken solder joint, although they seem to look good as far as I can see.
You can try to locate the culprit by poking the ceramic capacitors with an isolated probe (chop stick) and the required precautions.
Update: I just saw your last posting…
Chargers often have additional features to maintain safe operation.
For example a temperature sensor is checked whether the battery is in a healthy temperature.
Also the battery voltage might be checked before starting the charge cycle.
A load resistor might not be sufficient, depending on the intelligence of the charging circuit.
-
First of all, let’s sum up the symptoms:
Your charger makes a kind of unhealthy noise and the output voltage is zero except a short LED flash. (BASICALLY YES, and when outlet turns on LED flashes for a sec or two.
When you touch certain areas of the PCB, an output voltage of ~3V appears.(Before when I touched FETs with my fingers the voltage climbed as high as 3V depending on pressure, very hard=3V. Now I am using a plastic probe, and voltage is not budging. So in the name of science I fingered em again. Same result fluctuating voltage, and get this -- I witness one brief 20V reading! before calling it off it off. Probably should do as Swank suggested last week: reflow the FET solder joints and maybe others in same area?
Does the tone/noise also change? (with fingers, yes. at 3V the noise was much less.
The noise can be generated by magnetic components or ceramic capacitors which may not be faulty.
Strange noise can be generated due to an unstable regulation loop as switching power supplies usually operate above the audible range.
Also open load, overload or short circuit conditions can result in audible frequencies.
Semiconductors are unlikely to be touch sensitive, opposite to ceramic components, especially capacitors.
I suspect a broken ceramic capacitor that either overloads the secondary voltage or influences the regulation loop. (can you tell me where the ceramic capacitors are, numb nuts doesn't see any?)
Or a broken solder joint, although they seem to look good as far as I can see.(Thinking on the same lines now
You can try to locate the culprit by poking the ceramic capacitors with an isolated probe (chop stick) and the required precautions. (hum, no amount of poking with a plastic spudger has affect on. does that shot out the bad joint theory?
Update: I just saw your last posting…
Chargers often have additional features to maintain safe operation.
For example a temperature sensor is checked whether the battery is in a healthy temperature.
Also the battery voltage might be checked before starting the charge cycle.
A load resistor might not be sufficient, depending on the intelligence of the charging circuit.
ANSWERS TO QUESTIONS ABOVE ARE IN PARENTHESIS AND ITALICISED (last week I did hook up my similar 4 prong weed-eater battery to the same four prongs and it did not charge. If I recall correctly it also temporarily "killed" the battery as battery read no voltage. I had to take off battery top and briefly short battery main terminals to "revive" back to 20V (dam battery management crap -like Scotty on Star Trek used to say "them more plumbing they put it the more messed up things get" or something like that.
-
All the SMD capacitors are ceramic in this circuit.
Are you sure your similar battery has the same pin-out?
The charge pins seem to be on opposite sides, so check on your battery whether anything is connected to the inner pins. I would expect a resistor in between or two to minus.
I already have seen battery packs that had a coding resistor for voltage or capacity and a thermistor.
Shorting the battery terminals to reset the BMS is a brute force method, applying the correct charge voltage should do the same.
-
Have you tested/replaced the small C near the optocoupler P101?
This was my first guess anyway.
-
If you are open to taking suggestions from random people on the internet, my first choice (after visual inspection and sniff test) would be to measure the output rectifier D202 using the diode test function on your multimeter.
-
People that read the entire thread?
-
Oops, fair enough - too much skimming.
Well, if it's not the obvious then I'd start doing more measurements. I'd use a oscilloscope on the low voltage side just to see what is still working, but I guess that may not be an option in this case. I don't think there is a short if pressing a finger to the board can make a difference.
-
I was just a little bit ;-)ing