Connect with us

Voltage fluctuation on 12VDC/24VDC dual output P/S

Discussion in 'Power Electronics' started by partyanimallighting, Dec 7, 2019.

Scroll to continue with content
  1. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    Hey wizards, I had a couple non-functioning LED PAR's with blown power supplies which I replaced. I checked out the faulty power supplies and I am did not see anything faulty except for a scorched resistor (23ohm??), highlighted in yellow in the image so I replaced it but the output voltages still fluctuate, the 24V side between 9V and 18V and the 12V side between 6V and 9V. The caps, coils and other components also seem to be fine so I was hoping I could get some assistance with my limited electronic knowledge. I've attached an image of the power supply, front and back (back image reversed) with the caps and some other components labelled. I hope someone out there will offer some assistance.
     

    Attached Files:

  2. Martaine2005

    Martaine2005

    2,738
    738
    May 12, 2015
    D3192E6F-B317-4D1C-91D8-01073AC48EB5.png Looks like you have some dry solder joints on that board.
    Also looks like some cracked solder joints bottom left of board too. Might be worth re-flowing all the joints and testing it again.

    Martin
     
  3. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    Thanks Martin. I'll do a full reflow on all components in the morning, test and report back to the forum.
     
    Martaine2005 likes this.
  4. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    Tried the reflow. Same result. Fluctuating low voltages at output stage.
     
  5. Bluejets

    Bluejets

    4,052
    847
    Oct 5, 2014
    Looks to me like you've made an error posting your photos.
    In the first post, the red arrow should be coming from the 4 pin device at the bottom which you have marked as YG901C2.
    This would then make the burnt resistor feeding the unmarked 4 pin device which will be the switcher element with the feedback opto.
     
  6. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    I do not think there's an error. I reversed the bottom image to match the top image.
     
  7. Bluejets

    Bluejets

    4,052
    847
    Oct 5, 2014

    Fairly certain you have.
    I took the reversal into account.
    As I said, you have the red arrow coming from the device on the RHS of the board but this is a dual diode.
    The 4 pin device is on the left and like I already said, appears it is the switching element with it's burnt reference resistor.

    If you can find out what that is, you may just be well on the way to fixing it.
    Sometimes you will see numbers on a "cooked" device by shining a light on it at an angle.
     
  8. 73's de Edd

    73's de Edd

    2,871
    1,216
    Aug 21, 2015
    Sir partyanimallighting

    Au Contraire . . . .

    Looks like you DEFINITELY reversed some of your parts identifications.

    Specifically . . . .

    The top YG901C2 dual diode with common anode is a 5A fast switching unit which creates your lower voltage supply with the storage of its two 25VDC max electrolytics nearby . It looks to NOT be heatsinked.

    The lower YG901C2 dual diode companion . . .which is heat sink clamped . . . creates the higher voltage supply with its the three - 35VDC max E caps.
    That supply is being referenced to send back any load variance / voltage pull down back to the 4 pin POWER " switch mode power supply in a chip"
    Where, if it detects a voltage pull down, it sends a wider pulse width square wave to the power transformer primary for corresponding compensation.


    Your mismarking is being in the ALSO assigning of the YG901C2 parts designation to the 4 pin POWER " switch mode power supply in a chip" .
    Your "wheezing" power supply condition is telling us that the basic RAW DC high voltage supply and its pigtail fuse is good and HV dc is present at the LARGE HV electrolytic at mid chassis.

    Each of the YG901C2 full wave rectifiers dump into an initial E-cap and then a series ferrite cored choke ( black heat shrink wrapped) and some more E-caps , on past, for final filtering.

    With time and use and ALL of the high frequency hammering *** that those initial E-caps receive, from the strong squarewave waveforms coming into them, will start declining in filtering effectiveness and the developing of an ever increasing ESR, overheating and a consequential electrolyte pressure seepage.With a last gasp, final drying out.
    There is being such a decline in performance, such that they can no longer keep up with the needs of the power requirements. Thus the surging, while it is TRYING to fulfill a keep up level.

    ***
    https://cdn.dribbble.com/users/1181466/screenshots/3592598/02_construction_.gif

    If you are having almost zero techno measurement apabilities of analytical ripple and waveform situations, minimally, you could take a full leaded length ~ 1000 ufd at 35VDC new / known good E-cap and shunt across the first cap of the lower power supply and fire up to see if the surging abates to a disappearing or slowing down a bit of the surging.
    Then pull it and move thru all of the other cluster of E-caps in that area . . . one at a time.
    PLUS there is one more E-cap being involved, and that is a value that I can not read, it is being black and physically located at the 11:00 clock position from your YELLOW marked up heated resistor .
    ( And that resistor probably will be inserted series wise, within the HOT ground return of the units 4 pin " SMPS in a chip"

    You do not need to pull existing E-caps . . . just solder tack your substitute E-cap in parallel across them. TRIPLE CHECKING POLARITY MATCH.

    ALSO . . . .if the power supply has run for 5-10 minutes . . . marginal / bad caps tend to start warming up, so compare temperature between caps with a fingertip.

    Au Fin . . .
    Sometimes you will (/ CAN) see (LASER ETCHED ) numbers on a "cooked" device by shining a (VERY BRIGHT ) light on it at an (SPECIFIC) angle.
    . . . .( with the viewing help of a magnifier)

    True . . . true . . .VERY TRUE !

    Thaaaaaaassssssit . . . . .

    73's de Edd . . . . .


    A boiled egg is . . . is . . . is . . . . hard to beat.

    .
     
    Last edited: Dec 9, 2019
    partyanimallighting likes this.
  9. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    Thanks 73's de Edd!! You guys were so correct. I flipped once when I should have flipped twice!! :(:( I'll decipher and translate the electronic jargon of which you speak, double check to be sure I understand and follow your instructions thoroughly and report back soon. Now, I am no electronics wizard, I just do a little tinkering and I do a lot of work involving LED's and their power supplies and from what I've seen, the majority of problems are caused by failing capacitors. Get back to you soon.
     
  10. Bluejets

    Bluejets

    4,052
    847
    Oct 5, 2014
    Your photos are quite ok, no problem.
    Just pointing out your interpretation of which part is which, just so you are clear which device is which.

    Again, for your attention, the red arrow you use in the photo is coming from the wrong device AND as such, your marking of the switcher as YG901C2 is incorrect.
     
  11. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    Hi guys. Six months later but never too late, especially with what's happened to our world for these first few months. Anyway, I'm at it again with these four small power supplies. I took 73's de Edd's advice and went a step further. I pulled all the 25V and 35V caps and tested them and all had good readings. So I still replaced them all and powered up and the pulse was still there. I'm going to focus on the area with the scorched resistor now and that E cap next to the resistor. I'll pull the cap and the diode, test both, replace if necessary and report back my findings. See you soon.

    12VDC 24VDC POWER SUPPLY.JPG
     
  12. Harald Kapp

    Harald Kapp Moderator Moderator

    10,025
    2,138
    Nov 17, 2011
    450 V 100 mF? Unlikely if ot impossible. More like 450 V 100 µF.
    Watch your units: m <> µ and a huge difference.
     
  13. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    Yup. Sorry about the typo. :(:(:(

    12VDC 24VDC POWER SUPPLY REV. II FW.jpg
     
  14. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    Soooooo.....it's been a while but I'm trying to get four of these power supplies repaired. Two of them carry 450V 100μf caps and the other two carry 450V 220μf caps (maroon brown color) but otherwise are identical. As I mentioned before, I swapped out all the 25V and 35V caps on one of these power supplies and the problem persists. 73's de Edd's comments had me looking at the area of the burnt out resistor closely so I pulled the cap (50V 47μf) and it tested fine and the diode next to it too. There's what seems to be a component next to the scorched resistor and I pulled this but there is nothing marked on it, nothing marked on the pcb except "8MM" and it tests continuity both ways. There's a wire jumper below this marked "12MM" so my question is, is this an actual component (no marked value, I tried bright lights at an angle, lacquer thinners etc) or another "jumper"? The resistor is really scorched too so I'm not sure of the actual value either.

    12VDC 24VDC POWER SUPPLY 002.jpg
     
  15. Harald Kapp

    Harald Kapp Moderator Moderator

    10,025
    2,138
    Nov 17, 2011
    The "components" labeled 8 mm and 12 mm are usually wire jumpers of that length (as you can clearly see for the 12 mm part). These may be used to simplify PCB layout or to enable optional assemblies.
    From the image in your post #13 it looks like the "8 mm component" could be a ferrite bead, a length of wire through a ferrite core. This should register as a short circuit on your multimeter's Ohms range. These are veeeery unlikely to break down unless you physically damage the ferrite.

    R4 could be 10 Ω 5% (color code brown, black, black gold), but the colors are difficult to see on the images.
    What does a measurement show?

    The part labeled IC102 (817 in post #13) is a photocoupler (EL817?). The photocoupler is typically (and I assume this is the case here too) used in a feedback loop to regulate the output voltage. A bad photocoupler may also be a reason for malfunction of the power supply. Try to replace it. Suitable parts are EL817, LTV817, PC817, whichever you can get.
     
  16. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    Thanks for the prompt feedback Harald. I was thinking along the lines of "ferrite bead" myself or some sort of line filter similar to those used on AC lines. As for the resistor, the image is of the least scorched one in the lot on the four power supplies so that's the reason why I included the image, to get a more experienced opinion as it's so badly discolored. I didn't take a reading because I figured that, because they are all so scorched, the reading would not have been accurate anyway but I'll check and confirm in the next post. I did some reading up on the 817 a while back (infrared diode on one side, phototransistor on the other) so I'll be able to diode check across the legs between anode and cathode to confirm if it's functional, correct?

    IMAG1222-20200615-010438690.jpg IMAG1223-20200615-010437927.jpg
    Pictures with and without flash. o_Oo_O
     
  17. Martaine2005

    Martaine2005

    2,738
    738
    May 12, 2015
    If you have four scorched resistors from the same four items from the same PCBs. I would swap them for higher wattage resistors. Also, check all four for resistance. Hopefully a couple will be good and read the same.

    Martin
     
  18. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    Back at it again. I got similar results from 2 of the resistors and it looks like a 22Ω from the readings so to avoid future scorching I should replace it with a 1/2W 22Ω. Am I on the right track here?
     
  19. partyanimallighting

    partyanimallighting

    309
    4
    Oct 22, 2012
    I replaced the resistor with a new 27Ω that I found lying around and I also replaced the 817 opto-coupler with another. Also to 25V and 35V caps were replaced before but the pulse is still there. I realized that the labeling on the picture was wrong (YG901C2 and unmarked component needed to be switched). That unmarked component........I'm really starting to have some doubts about that particular component. I checked all 4 power supplies and the markings were removed on each of these components and I tried the norm, lacquer thinners, different angles, extreme magnification under bright lighting etc but it's unreadable. Is this a primary power MOSFET similar to what I encountered in another post?

    https://www.electronicspoint.com/forums/threads/triple-output-power-supply-blowing-main-fuse.292403/

    If this component is faulty, would it cause the pulsing voltage? All four power supplies have the same identical pulsing problem so I'm thinking that it's one particular faulty component common to all the power supplies. Revised image posted below....

    12VDC 24VDC POWER SUPPLY REV. III.JPG
     
  20. Harald Kapp

    Harald Kapp Moderator Moderator

    10,025
    2,138
    Nov 17, 2011
    With 4 pins this is not likely to be a simple MOSFET. I suspect it is a controller IC, possibly a linear voltage regulator similar to one of these. You would have to analyze the circuit (create a schematic from the PCB) to find out more.
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-