Panasonic
This Power Supply is dead. Can I fix it?
This Power Supply is dead. Can I fix it?
#Power #Supply #dead #fix
“Tony359 | Tony’s Tinkering Shop”
Thanks PCBWay for sponsoring this video: #smps #electronics #repair Second channel!
source
To see the full content, share this page by clicking one of the buttons below |
Let me chime in with another vote for "the Vcc cap for the chip might be the real culprit as well". Everything you saw on your scope with the supply not operating is consistent with the chip failing to start up and ending up in hiccup mode:
@13:19 – we are seeing the Vcc cap being charged through the startup resistor (rising part of the "sawtooth"), but then being discharged quite quickly when the chip reaches the startup threshold, until Vcc drops to the under-voltage lockout (UVLO) threshold. The Vcc cap needs to power the chip until switching starts to produce energy on the auxilliary winding, which does not start till UVLO kicks in.
@14:55 – While the chip is not running (because the startup voltage has not yet been reached or UVLO kicked in), the VREF regulator is turned off. So you would expect to see 7.5V while the chip is running and ~0V when the chip is suspended. The width of the pulses here is approximately consistent with the width of the falling slop at the Vcc pin (the time the chip is running).
@16:30 – This might be the death knell to the theory the Vcc capacitor is to blame. For the Vcc to be powered from the Aux winding instead of the startup resistor, there need to be switching pulses. On the other hand, you are looking at just one of the two switch outputs, possibly the first pulse happens on the other one, and UVLO already kicks in before the pulse on this output is generated.
The startup resistor that charges the Vcc cap *does not supply enough current to power the chip while it is operating*. So there is a "take-over" period in which the chip draws its higher operating current instead of the really low standby current, but does not yet get powered from the aux winding. During that period, the Vcc cap is the only source of power for the chip. If the capacity is too low or the ESR is too high, the time period that the Vcc cap is able to power the chip is not sufficient for the supply to start up and generate its own supply voltage from the aux winding.
I see three possible explanations why swapping the controller chip and the broken 100nF cap fixed the supply:
– The controller chip was acutally broken.
– The replacement chip you installed uses less power during startup than the old one, so a marginal Vcc cap is sufficient for the replacement chip but unsufficient for the old chip.
– The Vref cap was already broken (i.e. the leg disconnected) before you desoldered the old chip, and the missing capacitance on Vref caused the actual issue. Your work on the PCB just caused the broken capacitor to be more visible.
Whenever you decide to recap a SMPS, in my oppinion you should always recap the Vcc capacitor for the controller chip as well as the output capacitors. I've seen so much SMPSs failing due to the startup cap getting bad that I would never skip that one on preventive maintainance.
I noticed you keep saying 'cage'. Out of curiosity: is that a common thing? I'm not an native english speaker but isn't "housing" a better word for it? Enjoyed the video nonetheless Tony! Always do! Keep up the good work!
This is among the best, well explained SMPS repair videos I've seen yet. Great stuff Tony.
If you already replacing caps – at least check the high voltage one. People often do not even check them as they're mostly good but that's more "good enough" than good after 20 years. Even with low frequency of main network transformer still stress them with high frequency depletion and sitting there in the middle of heatsinks doesn't help. They're working closer to rated voltage (sometimes just 10-30V under it) also putting additional stress on them.
Also small cap (usually 10-22uf/50V) next to control IC is common point of failure, replacing it with new quality cap is more important then replacing good measuring output capacitors.
Symmetrical powersupplies are not only for balanced outputs. Nearly every opampcircuit needs it to function properly. You could use voltage dividers and have a so called "virtual earth", but this isn't exactly how good audio circuits should be built.
Nice fix, but I think you should have also replaced the small electrolytic that does the startup supply for the IC, small caps usually die first, I’ve fixed several power supplies just by replacing the start up capacitor.
@10.00 that behaviour is probably the small capacitor near the chip, heat can bring back a capacitor a little when it is marginal.
Could also be something as simple as a bad solder joint in that area.
Can you do a tutorial on your oscilloscope since you explain very nicely and are understable
I know it's your style, and it's fine, but you don't need the 3-2-1-go countdown. we can pause and rewind, no problem
This is the most I ever learned in 35 minutes. Bravo
I really must get a differential probe its a lot saver than just an isolation transformer. Excellent video thank you
Great video as always. Cheers!
The only pin you did not measure after you got it working was the vcc, I suspect this will show you the cap is duff on VCC and you saw the ripple on VCC when it was not starting I am upto 20mins so far.
I stick by my diag, just the new IC is more tolerant.
I repair quite a lot of PSU's and the cap is the most common replacement, in fact I repaired a medical grade PSU tonight and it was a 100uf 35V on pin 7 of a 3842 IC.
I rarely replace PSU IC's unless it has blown the output FET then it usually kills the ic.
On the other hand if a TOPxxx ic is not working that is much more common.
Nice job. Enjoy your videos.
Great job, well explained, thank you.
I had a similar case where the power supply started during measurements, but the culprit was a small electrolytic decoupling capacitor for controler IC. The golden rule with capacitors is that the smaller the size, the sooner it needs to be replaced. In application notes, such capacitors have the shortest lifespan.
как же круто когда ты можешь смотреть видео на 4 языках. poti sa treci in romina. Украинску мову. and english… Its time to learn one more.
It could have been a dry joint
the short basic theory behind why the SMPS is much more compact is that transformers change the electrical energy into magnetic energy then back to electrical — but for the magnetic energy the transformer's core should have enough capacity to "take it up" (not get saturated) which is depending on the power and the length of the electromagnetic cycle — so for big power you need a big core if the cycle is slow or for the same power use faster cycle that requires smaller core thus smaller transformer
And to do that it first makes DC, then a switching transistor chops it up and feeds the transformer with high frequency electricity. Then the secondary side just rectifies and smooths it (then there are feedback and protection and filtering and yada-yada….)
Nice diagnostics and fix.
Must say thats a very neatly designed power supply.
Regarding the differently labeling on the IC's they could have be pulled from E-waste and indeed different years/batches.
Considering you mention its around 20 years old replacing the capacitors is certainly a good idea as its around the capacitor plague era.
Hi Tony! – I follow your advice and dont mess with power supplies
Not because of the explanation you and other people give about the basics of its operation
That I understand… the big components I understand, the bridge rectifier, the switching transistors etc
But it is all that ARMY of small components there, resistors, ICs, capacitors that really scares me… it seems there is so so so much more going on there
Dont tell me all those dozens of tiny componentes are all just for the voltage sensing and feedback circuit….