what causes a switching power supply to not drive a load

• Articles (current)
• Forums (current)
  • Latest
  • Microcontrollers
  • Ability Electronics
  • RF & Wireless
  • Sensors & Actuators
  • Community

• Register
• Log In

Switching ability supply voltage drops under load

Give-and-take in 'Troubleshooting and Repair' started by calmsea, Dec xiii, 2014.

• Search Forums
• Recent Posts

Scroll to continue with content

1. Hi,

   I take an Allen & Heath mixer with a faulty power supply I'm hoping to fix.
   The ability supply puts out +/- ~14 volts with no load (plus a 48v for the phantom power), and I think it'south supposed to run at 12v, but continued to the mixer it drops to a trivial over 5v. Even with a little 5W lightbulb on information technology it dips to less than 7v, and information technology's supposed to supply upwards to 45W.
   It'southward a switch fashion supply based on a UC3842BN controller. I've been learning a bit about SMPSs from repairfaq, but I haven't seen annihilation that describes this problem.
   On visual inspection, zero looks burnt or plain-featured, and solder joints all look good.
   Whatever suggestions every bit to what I should be looking for?

   Many thank you,
   Galen

2. hello Galen
   welcome to EP

   2 possible causes come to mind

   1) the mixer circuit has a fault and is loading down the PSU more than it should
   ii) the PSU itself is faulty .... look at replacing any electrolytic capacitors in the output stages
   you may find some of them are bulging

   evidence u.s.a. pics

   Dave

   Terminal edited: Dec 18, 2014

3. Hi Dave,

   Cheers for your respond.
   I tested the power supply with a pocket-size light seedling. Information technology drew ~200 mA and the voltage fell to vi.4v. The power supply is rated at 45w and then I figure that means information technology must be the power supply that'due south at mistake. Is that reasonable?
   Looking at the power supply, the only thing that I noticed is that a couple of the output caps are very slightly domed on top.
   Testing them with the cap office on my meter they are in range though. Is it possible that they are withal bad? Is there annihilation else I should look at?

   thanks

   
   

4. Hi in that location. Yes, that could very well be the problem. Bulging indicates overheating and venting has occurred, and this reduces the capacitance and increases the ESR of the capacitor, which increases the heating, and so on. So their days have been numbered for a while at present.

   When some electrolytics are showing signs of venting, I replace all electrolytics of that brand, using low-ESR parts from a good-quality Japanese brand (Nichicon, Rubycon, Panasonic, UCC).

   It looks like the faulty ones are "Samxon" brand - Chinese, I expect. If y'all post the complete markings on all Samxon electrolytics, we may be able to find data sheets for them and recommend suitable replacements.

   Where are you located? And do you accept a preferred component supplier?

5. I had a couple of new 1000uF/35v electrolytics lying effectually so a swapped them for the 2 suspect caps and sure enough, that fixed information technology.
   I'll pick up some good 470uF/25v to finish the job (that was the original value). Or should I try a slightly higher value for the replacements? Should I replace all the caps in the output section?
   I'1000 in Vancouver, Canada, btw. I usually selection up components from a local supplier, Principal Electronics, or order from digikey or mouser.

6. first-class well done
   its a very common PSU error

7. Standard electrolytics aren't suitable and will fail. You need low-ESR components.
   As I said, I would replace all electrolytics in the output section that are the same brand as the ones that failed.
   OK, Digi-Key will definitely have suitable replacements. Can you brand a listing of the "Samxon" capacitors in the output stage, and I'll make a list of Digi-Key part numbers for you lot to order. For each Samxon capacitor type, can yous give
   

   • All markings (not just capacitance and rated voltage)
   • Bore
   • Pb spacing

8. Wow, thanks!
   There are half dozen Samxon caps in all.
   5 of them are 470uF, 25v
   There is a GF, which I believe is the Samxon serial, an Chiliad (for +/-20% tolerance?)
   YDA, which looks like an Environmental impact assessment tolerance code, but the D in the middle doesn't seem right.
   The terminal line says 105°C8.
   It'south 10mm across, with 5mm spacing.
   I plant many candidates on Digi-Key from Panasonic and Nichicon, but nearly of them accept a blank ESR field. Not sure how to choose. The one with the longest lifetime rating? The mixer the ability supply is installed in will be on substantially 24/seven until information technology dies.

   The other Samxon cap is 120uF, 63v, GT(M), Z1A, 105°C8, 7mm and 4mm.

9. Right, I Googled Samxon GF series and the commencement result was the PDF data sheet! You accept the 10 mm diameter, 12.5 mm height version. The specifications are 80 mΩ maximum impedance (at 100 kHz) and 865 mA maximum ripple current.

   It looks like the v big capacitors are not the same - three are green and two are black. Are they all Samxon GF?

   Given that Samxon is a no-name Chinese electrolytic manufacturer, these numbers could have been pulled out of thin air (or somewhere darker), or from another manufacturer's data sheet, so they may have been operating within their claimed specifications.

   Y'all should replace them with the same value or slightly college, the same impedance or slightly lower, and the aforementioned ripple current or higher. Information technology'south actually the impedance yous need to match, not the ESR, although these capacitors are called low-ESR!

   Here's the pick filter I used: http://world wide web.digikey.com/production-search/en?FV=fff40002,fff80009,fffc000a,fffc01ed,fffc0235,fffc04a5,3400b7,b8367d,b8387a,b8388a,b8388f,f00339,f0033c,f00340,f00343,f00344,f00345,f00354,f00360,f00367,f00381,f00388,f00393,f003a4,f003a8,f003b2,f003bd,f003c0,f003c1,f003c3,f003c5,f003c7,f003d4,f003f1,f003fb,fc01ea,1140050,2dc0000,2dc0051,2dc1e56,7f00004,b540002,b540004,b540005,b540007,b54000c,b54000d,b54000e,b54000f,b540011,b540014,b540016,b54001d,b540037,b540042,b540046,b540047,b540048,b54004e,b54004f,b540057,b540060,b54006f,14940090,1494010c,14940275,149403c5,149403ec,14940442,1494044b,149404b5,14940567,149405d5,149406b7,149407bd,149408b9,14940c2a&ColumnSort=1000007&folio=one&stock=ane&quantity=1

   This is the 1 I recommend: http://www.digikey.com/product-particular/en/EKY-250ELL471MJ16S/565-1549-ND/756065. Information technology'southward a United Chemi-Con (UCC) KY series part. 470 µF, 25V, threescore mΩ impedance, 1.21A ripple current, xvi mm tall, 7000 hours at 105 °C. USD 0.51.

   The GT ane is specified for 264 mΩ impedance and 362 mA ripple current. The UCC KZN part (http://www.digikey.com/product-detail/en/EKZN630ELL121MH20D/565-4087-ND/4843897) looks pretty good. Its impedance is much lower, at 77 mΩ, and this will increase the ripple electric current, only its ripple current specification is one.2A!

   Actually the highest impedance you lot tin get from Digi-Key for that value is 85 mΩ. Darn them for merely stocking good quality manufacturers! I incertitude there will be a problem with that 1, but if in that location is (e.m. ability supply not stable - makes a sizzling or whining dissonance), you can always add a small-scale resistor (e.g. 0.15Ω) in series with it.

10. Absurd, thank you!
    The 2 darker ones are the same. The colour divergence seems to exist related to their failure.
    I found this pretty proficient document while looking upwards ripple current and capacitor impedance, in example any one is interested:
    http://world wide web.cde.com/resources/catalogs/AEappGUIDE.pdf

11. Look to the Caps first. They can total yous.

12. New here. Did non expect down. LOL! Sorry.

Enquire a Question

Want to reply to this thread or ask your own question?
Yous'll demand to choose a username for the site, which only take a couple of moments (here). After that, you tin can postal service your question and our members will help you out.

jonesmomall.blogspot.com

Source: https://www.electronicspoint.com/forums/threads/switching-power-supply-voltage-drops-under-load.271797/

Share this post