TV power supply repair - diode replacement options

92saturnsl2

92saturnsl2

Joined
Mar 2, 2004
Messages
4,102
Location
Kentucky
Father in law brought over a dead Vizio flatscreen LCD TV that wasn't showing any signs of life. He bought a new one so it'll be a free TV once I get it working.

I've fixed TV's before and the majority of the time it's the power supply board that flakes out, so I started there. Unplugged all electrical connectors to the logic/CPU board, and tested for voltage at the output connector. Where I'm supposed to have 19V for the standby circuit, I've got nothing. I go around the board testing diodes and I immediately find a pair that are shorted. I remove them from the board and test them again and only one is shorted. So evidently they are wired in parallel because they were both testing as shorted before removing them from the circuit. Just to confirm my suspicions, I reinstall the good diode and repeat my test looking for voltage at the output, and SUCCESS, I have 19V at the standby pin now, and TV powers up fine.

The diode number is SB5150, which some Googling shows to be a 5A 150V schottky doide. I'm trying to find something on Amazon but they don't have an exact match to that. Given that this diode is on the secondary side of the power supply, is a 150V part really necessary? There's plenty of 45V parts on Amazon to choose from. Do I just need ensure the amperage is at least 5A and correct voltage rating, or are there other variables I need to be looking at? LT4116 is another number printed on the diode but Google doesn't turn up anything helpful. I guess my question is, is a schottky diode a schottky diode, or do I need find an exact match for this part? Hoping someone can chime in that knows a little more about the intricacies of this stuff than I do!
 
I'm dubious that they would spec a part at a rating higher than necessary--the higher the voltage, the worse the other parameters, typically. When you build a million of something, you have the means to pressure vendors into shaving fractions of a cent. Then again, maybe BOM reduction. Who knows.

You sure SB5150 isn't coming up? I quickly found a match at both Amazon and Ebay. The Amazon one claims to be a cross.
 
supton said:
I'm dubious that they would spec a part at a rating higher than necessary--the higher the voltage, the worse the other parameters, typically. When you build a million of something, you have the means to pressure vendors into shaving fractions of a cent. Then again, maybe BOM reduction. Who knows.

You sure SB5150 isn't coming up? I quickly found a match at both Amazon and Ebay. The Amazon one claims to be a cross.
Click to expand...

Isn't that already heavily shaved off? Part of it would be a lot of stuff has already been integrated into ICs rather than as discrete components. The discrete components left are now smaller/cheaper surface mount packages where assembly is automated. And typically if something doesn't test out they're just going to toss the entire board rather than investigate what part might need to be reworked.

I would expect that there are a lot of materials where it doesn't make sense to make different grades. A lot of ICs these days are identical, and features are enabled by hardwiring inputs to access different settings.

So what's left these days that easy enough to replace? I suppose power diodes are something that are still available as discrete components.
 
y_p_w said:
Isn't that already heavily shaved off? Part of it would be a lot of stuff has already been integrated into ICs rather than as discrete components. The discrete components left are now smaller/cheaper surface mount packages where assembly is automated. And typically if something doesn't test out they're just going to toss the entire board rather than investigate what part might need to be reworked.

I would expect that there are a lot of materials where it doesn't make sense to make different grades. A lot of ICs these days are identical, and features are enabled by hardwiring inputs to access different settings.

So what's left these days that easy enough to replace? I suppose power diodes are something that are still available as discrete components.
Click to expand...
Everything is as cheap as possible, but you are right, one way for an OEM to shave costs is to make one die and then make it programmable--one trip through R&D versus multiple projects.

Not a lot today that isn't SMT. Easier to assemble, sure, although the PCB is often part of the heat gets out of components. That said, while integration is huge, there is still quite a demand for discretes, if nothing else because of heat, and sometimes because of the voltages & currents encountered.
 
supton said:
Everything is as cheap as possible, but you are right, one way for an OEM to shave costs is to make one die and then make it programmable--one trip through R&D versus multiple projects.

Not a lot today that isn't SMT. Easier to assemble, sure, although the PCB is often part of the heat gets out of components. That said, while integration is huge, there is still quite a demand for discretes, if nothing else because of heat, and sometimes because of the voltages & currents encountered.
Click to expand...

I'm really more a digital guy, but some of my work and education has been in test. I remember the first time I asked about how boundary scan test or any kind of error with a board and its components went, and was told that means just scrapping the whole thing unless it's a prototype. It just costs too much to fix something. Everything is automated, and surface mount is much faster than drilling thru holes, bending wires, etc. It's not terribly friend for DIY repairs though. It's not really worth trying to fix an iPhone unless it's for data recovery. I think LCD TVs these days use a lot less power and don't require terribly beefy power supplies. My 55" TV says typical 60W power consumption, 140W max, and that's way less than my 12 year old plasma screen.
 
I
supton said:
I'm dubious that they would spec a part at a rating higher than necessary--the higher the voltage, the worse the other parameters, typically. When you build a million of something, you have the means to pressure vendors into shaving fractions of a cent. Then again, maybe BOM reduction. Who knows.

You sure SB5150 isn't coming up? I quickly found a match at both Amazon and Ebay. The Amazon one claims to be a cross.
Click to expand...
For the power supply to pass UL certification, the primary and secondary side of the power supply must be physically separated; since that diode lives in the secondary side, there should be no high voltage in that area. The highest secondary voltage is 19V, so I would assume the diode doesn't need to handle 150 volts. But I'm not 100% sure how diodes are rated, so I'll still with the original rating. I'm guessing they use it for parts commonality, only needing to stock one part for multiple applications.

I checked Amazon and the part that comes up for SB5150 has a delivery date in mid-March. Hoping to get it sooner than that. I ended up picking up a 20 pack on Ebay that shipped from a US warehouse. They are most certainly off-brand China components, but probably as good or better quality than the original diode which only lasted about 18 months. These are discreet "old-school" diodes that are very accessible and easy to test/replace.
 
The longer lead time components/products for sale on Amazon are typically being shipped from China.
Since the entire TV was manufactured in China, my guess is that every component in it was manufactured in China, so replacing a failed Chinese component with a new Chinese component wouldn't be the end of the world. The vast majority of electronic components are being manufactured in China now days.
 
y_p_w said:
I'm really more a digital guy, but some of my work and education has been in test. I remember the first time I asked about how boundary scan test or any kind of error with a board and its components went, and was told that means just scrapping the whole thing unless it's a prototype. It just costs too much to fix something. Everything is automated, and surface mount is much faster than drilling thru holes, bending wires, etc. It's not terribly friend for DIY repairs though. It's not really worth trying to fix an iPhone unless it's for data recovery. I think LCD TVs these days use a lot less power and don't require terribly beefy power supplies. My 55" TV says typical 60W power consumption, 140W max, and that's way less than my 12 year old plasma screen.
Click to expand...
Not terribly surprising, it costs "a lot" for a tech to pull up schematics and figure out what went wrong. After an hour or two one likely has exceeded the raw cost to just toss the assembly and move on.

I live on the opposite end where everything is prototype. I can't count how many times my modification list starts with "using wirewrap wire..."
 
92saturnsl2 said:
I

For the power supply to pass UL certification, the primary and secondary side of the power supply must be physically separated; since that diode lives in the secondary side, there should be no high voltage in that area. The highest secondary voltage is 19V, so I would assume the diode doesn't need to handle 150 volts. But I'm not 100% sure how diodes are rated, so I'll still with the original rating. I'm guessing they use it for parts commonality, only needing to stock one part for multiple applications.

I checked Amazon and the part that comes up for SB5150 has a delivery date in mid-March. Hoping to get it sooner than that. I ended up picking up a 20 pack on Ebay that shipped from a US warehouse. They are most certainly off-brand China components, but probably as good or better quality than the original diode which only lasted about 18 months. These are discreet "old-school" diodes that are very accessible and easy to test/replace.
Click to expand...
Interesting; I avoid UL as much as I can (just don't do much with it) so it's an unknown world to me.

Wild guess: Shotkeys are not low leakage devices. In order to keep leakage and thus power loss low, perhaps they went with a high voltage diode--at 150V the leakage current is likely not far below the maximum package dissipation (current * voltage, gets Watts, then times thermal resistance to find junction temp rise). But using well below the max voltage means leakage drops off significantly, and lower leakage means less loss.

I'm guessing this supply might be live always? thus this affects standby current, and I think there are regs about how much electronics can pull in standby mode.

Couldn't find a good datasheet for SB5150's nor MBR5150 but this was what I could find, and you can see the dramatic drop off in reverse leakage as a function of voltage.
MBR_diode.PNG
 
I'm scratching my head why the use of a Shotkey diode? Whenever I built, or repaired, power supplys I used plain 1004 or 1005's.
 
Schottky diodes are used for their fast switching (Trr) , and low forward voltage (Vf). We have used them for decades on the secondary side as rectifiers in Switching power supplies.
 
Generic 1N4005 etc diodes (60 Hz mains frequency) won't last long in the high freq secondary side of switch mode power supplies. For that, you need fast recovery/Schottky diodes.
 
You must log in or register to reply here.

Similar threads

1
Lost a Good TV last night... 2014 Vizio TV failure
Replies
17
Views
2K
Barkleymut
1 SX
Klipsch home theater Sub-10 stopped working. Looking for an upgrade or maybe repair..
Replies
13
Views
3K
1 SX
JHZR2
Pinpointing a wire break location? AC power technique?
Replies
13
Views
955
meep
J
  • Locked
looking for multiple battery protection design.
Replies
7
Views
793
ironman_gq
clinebarger
  • Locked
I get all the good ones......
2 3
Replies
52
Views
11K
clinebarger
Back
Top