Newer 12V 5A Power brick no longer putting out full (enough) current

[born2fly2]

I have a 7 month old power brick for my security camera dvr that stopped putting out more than 1.5 amps, suddenly. I've gone through it twice with a MM and everything seems to be as it should. I was not able to test the opto coupler, so I don't know if that could be the problem. The inductors seemed to be within range when I ran an ohms test on them. The 'yellow' guy between inductors I did not test either as I am not sure what it is and what it's function is. Basically - I am trying to determine where the point of failure is and what would cause the current to be reduced/restricted.  I'm getting 12V and 1.5A. I decided I would challenge myself on this one and what a challenge it is turning into. 

[Rasz]

was it in someplace hot like an attic? the sad boring answer seems to be always output capacitors
do you have a scope to look at the output voltage? or ESR meter? if not then just quickly solder additional output cap and retest the supply to se if there is any difference

[born2fly2]

I don't have a scope yet, just a MM and workstation stuff ...it's on my list, but I did remove the electro caps and they all tested good. I checked all of the smd resistors & caps, all good. Really stumped. I know a scope would really help out but holding back pulling out the credit card for the moment, but....  VERY tempted. Thank you for your input.

[born2fly2]

Interesting info - thank you. They sell for about $15+ online, from what I've seen. YES, for fun and also education, still learning all of the newer tech. Was an ET in the navy 4 decades ago. Will try the safer and cheaper route. Curiosity has me and the fact that it cost me $20 to get another brick (hopefully better quality this time)

[sleemanj]

How you are testing this power brick?

Never exclude an error in methodology as a reason for apparent failure.

[born2fly2]

I need 12V and aprox 2.3 A to run the DVR. What kind 'device' would be required to trick the unit into supplying the 12V?

[TimNJ]

Interesting info - thank you. They sell for about $15+ online, from what I've seen. YES, for fun and also education, still learning all of the newer tech. Was an ET in the navy 4 decades ago. Will try the safer and cheaper route. Curiosity has me and the fact that it cost me $20 to get another brick (hopefully better quality this time)

Do you really need 5A? If not, try getting a USB PD adapter, those 60W/65W ones, and you can get 3A~3.25A max from them, more than your 1.5A need.

You also need a tricking device to trick the adapter that you are a computer needing 12V (or 9V or 15V or 20V, depending on which one you chose), so it will output 12V instead of the default 5V.

Those are still relatively higher tech, so the technology is only available to few companies, thus you don't get utter crap like this one.

Xiaomi has a 65W gallium nitride-based unit selling for some $20 in China, and I just bought one. It is surprisingly small, but surprisingly well built.

It is based on two Navitas GaN transistors configured in active flyback topology, controlled by a presumably ST controller, and it utilized some pretty cool technology like polymer capacitors and planar transformers.

Unfortunately I can't show you a teardown as we are not yet to publish it. We took apart a few GaN adapters available in China and did thorough tests -- load characteristics, drain waveform and component level analysis.

The bottom line is, the Xiaomi "lipstick" adapter is the best performer, and not the most expensive one.

Isn't this already public?

http://www.chongdiantou.com/wp/archives/46552.html

Or is there another one? The teardown above shows a nice quality product. Only thing I can fault is the SMD Y-caps, and that's really just a preference. Although they are Y-rated, and supposed to fail safe, big surface mount ceramic caps are a little scary with respect to cracking during manufacture, or further down the line.

[station240]

Have you pulled the output diode out and tested it ?
It's the TO200 device bolted to aluminium, top right in your photo.
Usually these have 2 diodes connected to 3 pins, if one of those diodes is dead, then it would cut the output current in half.

[TimNJ]

Although they are Y-rated, and supposed to fail safe, big surface mount ceramic caps are a little scary with respect to cracking during manufacture, or further down the line.

Those Y caps are made by HEC, specifically for Y cap applications and certified for that applications. The internal construction is actually 2 caps in series, sharing an intermediate plate that s not brought out.

Also, just to be safe, Xiaomi used two safety rated Y caps in series, and to make up with the loss of capacitance, also 2 in parallel, so they have 2*2 of 1nF caps.

However, one thing I don't like about it is its higher Y cap leakage, which is unfortunately needed to pass EMI testing since they used 0.8mm PCB planer transformer, so there is quite a large pri-sec capacitance that needs to be suppressed.

Ah, so each SMD Y-cap is 2 in series. Is this the way all Y-caps are made, or just SMD type?

So, will planar ever be useful for a low leakage application? Medical? Bias/cancellation windings in the transformer to reduce interwinding capacitance?