STB fault - you guessed it...

Do you have a question? Post it now! No Registration Necessary

Translate This Thread From English to

Threaded View
My several year old Digicrystal PVR started playing up some time ago,
taking ages to turn on and lock the image. It got worse recently and
then the PVR function refused to work at all.
Given all the recent similar reports on the group, I knew what the
fault would be. Sure enough, I just opened it up and found most of the
(105 rated) caps in the switchmode supply with bulges - no point
measuring the ESR.
6 new caps later and the box is as good as new.

Are STB's capacitor failure magnets?

Dave.

Re: STB fault - you guessed it...

Quoted text here. Click to load it

Probably not, but cap failures in el-cheapo Chinese SMPS isn't all that
uncommon.
When one looks at the design and specs closely, most are made for 220V
supplies not the 240V nominal supply voltage in Australia, which can drift
up much higher in mains distribution areas with poor voltage regulation.

Cheers,
Alan



Re: STB fault - you guessed it...
Quoted text here. Click to load it


    Not long ago I posted my experience with a brand-new Digicrystal STB
remote control acting up because the electrolytic cap across its 2 x AAA
cell battery was as good as open circuit.

    So the answer seems to be yes. Digicrystal units at least use really
awful cheap electroltyics.


Bob


Re: STB fault - you guessed it...
Quoted text here. Click to load it

I've had to do the same in a couple of cheap DVD players. I dont think
its a design issue (although I have seen those, where designers simply
use massively under-rated caps), these are all small flybacks and a
simple calculation shows the ripple current requirements are not onerous
(rough calc: D = 0.5 so Ipk = (2/D)*Idc = 4*Idc and Irms = Ipk*sqrt(D/3)
= Idc*1.6. for better calc measure actual D) its a shitty cap issue.

Just because it has a label on it that says "105C" doesnt mean its a
105C cap - unscrupulous asian manufacturers long since discovered that
printing labels is far cheaper than manufacturing good products.

also if you look closely at many cap datasheets you will see the shitty
caps specify end of life further down the curve than decentcap mfgs -
who use 200% Ileakage and -20% capacitance, c.f. 300% & -30% for dodgy
caps.

I went thru the exercise a few years back of comparing such caps, and
found that moving the definition of "failure" more than doubled the
alleged lifetime. This after a slaes rep had convinced our purchasing
guy that these caps were "better"

Personally I wont design in an electrolytic unless its made by:

Panasonic
Hitachi AIC
Nichicon
Nippon Chemi-Con
Rubycon

Cheers
Terry

Site Timeline