APEX AD-1010w DVD Player repair

[(*steve*)]

In general, things that are built in quantity are built using the cheapest components that will do the job. As long as the device lasts longer than the warranty period in average use then it's not coming back.

The more cynical of us would suggests that the use of certain components may be designed to cause failure after warranty so you have to buy another one.

So yes, you will be unlikely to see the more expensive components where a cheaper part will do, even if it will do only for a relatively limited time.

When you are repairing something, the economics of scale are rarely such an issue. The difference in cost for the cheapest vs the most expensive components will be swamped by other costs (your time being foremost).

The failure of capacitors in SMPS can probably be timed quite well as long as the environment the device is in and the load remain constant. In cases where I've replaced capacitors with what turned out to be exactly the same type this exact issue has crossed my mind.

 

[Resqueline]

Life is rated at max temp, max ripple, & max voltage - all at once. Reducing temp & ripple will exponentially increase life.
That translates into that the base life rating is quite unimportant, whereas the "headroom" of the ripple has an overshadowing great impact on lifespan.
2000 hours is just 3 months, and 9000 hours is just 1 year, and that difference is significantly less than the 10-fold increases seen by the above effects.

So running a (longlife) cap at max ripple may give you a year service, whereas running a short-lived ultra-low-esr cap at 1/3rd ripple might give you 10 years of service.

There is more difference between caps than most people realize, and those parts are the ones having the greates impact on service life of equipment.

E-lytic caps are not made in 20V. The next step over 16V is 25V.

Polymer caps can be put directly in place of e-lytics, if their often limited capacitance range is not an issue. Their cost is why you don't see them more often than you do.

Did you try to check out what the original cap's (the ones that failed) in your application were rated at?

 

[Tesla]

Thanks guys for the additional explaination.

Yes, I saw those low hours but then I realized that's running at max. I'm sure they last much longer at lower temps.

Ok, I found the 25v ones. Again, PW look to be the best ones DigiKey carries (Mouser has the others). I did notice Rated Ripple on the PWs is nothing great.

Yes, I noticed the Polymer caps have limited values, so that's another thing holding back their use.

How would I find out what they are rated at?

 

[(*steve*)]

There is more difference between caps than most people realize, and those parts are the ones having the greatest impact on service life of equipment.

I think they're the "wax coated paper capacitor" of our times :-(

I also think it's true that people are far more likely to look up the specs of a semiconductor they replace than a capacitor.

Having said that, the capacitor specs can be rather hard to find.

And given some recent experience it's hard to know if capacitor failure was due to

1) Reaching end of "design life"

2) Initial choice of poor (or badly manufactured) capacitors, or lack of concern

3) Deliberate choice of cheapest possible components (China, I'm looking at you)

As an example of the former. I have some monitors that failed after 5 years of being on 24x7. The warranty period was 3 years, and the monitors were probably not designed to be left on continuously. If one assumes that 40 hours per week is an assumption made by the designer, then these lasted the equivalent of almost 20 years -- and you can't argue with that. Would that perhaps be considered over-designed?

edit: The easiest way is if the supplier has links to the specs. I find Digikey to be pretty good in this respect. If it's the specs on the capacitor being replaced, then (if you're lucky) you can identify it and find it at Digikey or Mouser etc with specs. Far more likely you'll have to go to the manufacturers web site and look for the specs for that capacitor series. It may take some time to determine which series it is, and for many surface mount components you're going to be really battling because even if you identify the manufacturer, there's likely to be precious little space to have further series identification, or even in some cases the value (think ceramic capacitors). To further compound this problem is that it is often difficult to find specs on "obsolete" components, even if you know the manufacturer and the full part number. To be fair, capacitors under stress tend to be physically larger, so generally there's something to read.

 

[(*steve*)]

Let's look at a random case.
====================

I have some capacitors identified as TPSE107K016R0. I know that they're a 100uF 16V 10% low ESR capacitor. What else can I find...

Going to Digikey and sticking in the part number tells me that the part number is incomplete. It should have a few more digits on it e.g. TPSE107K016R0125. From Didgikey's catalogue I can see that they come in 125, 100, and 55 milliohm ESR and that they are manufactured by AVX.

Clicking on one at random I see that they are TPS series and that there is a link to the series datasheet. Searching through that I note that they're all in an E case and they have significantly different ripple current ratings. But it doesn't tell me what part markings I should see, and whether that will identify the ESR.

However it does say "For part marking see page 132". This document goes from page 33 to page 43, so clearly this is part of a much larger document. Digikey also has links to a photograph and technical drawings, but these are totally generic, the photo having no markings shown at all). So it's off to AVX...

The obvious site is http://www.avx.com, and that turns out to be correct. Looking at "technical Papers" and "Tantalum Capacitors" yields a document called "AVX TPS Series III..." which after a brief synopsis leads you to a very technical paper describing these capacitors and includes a subset of the specs we've already seen. Possibly of interest if you need some construction details, but not what I'm looking for...

After some searching around, I again search for TPSE107K016R0 on the AVX site and look at the "Master Catalog". Sure enough pages 33 to 43 correspond to the other document, and page 132 tells you what you'll see on the devices themselves.

According to the manual, you cannot tell the ESR value from the case (only capacitance, voltage, and lot code).

So how do I tell if I should use the 55, 100, or 125 milliohm version as a replacement?

For a single capacitor you're looking at between $US8.50 and $US13.75, so they're not cheap.

Let's try another example...
====================

An electrolytic capacitor, surface mount, with the markings 47 / 35A / P25 over 3 lines. It's a small aluminium cylinder. There is no manufacturers logo. The only other marling is a dark segment painted on the top to indicate polarity.

A good guess is that it's 47uF 35V, but it could also be 25V. Can we find any specs?

I happen to know the part number of this capacitor, but going on what I have here, I would have extreme trouble unambiguously identifying it if it were on a board.

Third example
==========

Some through-hole electrolytic capacitors in a power supply labelled Suntan / CD286 / 105 degC / 25V / 470uF.

It's pretty unambiguous what the capacitance and voltage rating are. In addition, this being a 105C part amongst other 85C parts tells me that it's not rated 105C because of elevated ambient temperature -- so it's probably a low ESR capacitor.

I go to digikey and quickly look up 470uF 25V, and after I tell it that I'm looking for aluminium electrolytic capacitors I note that "Suntan" is not a brand that is listed.

http://www.suntan.com/ clearly isn't the right place to go. A quick Google tells me that I want http://www.suntan.com.hk/. A couple of clicks gets me a datasheet on CD293 capacitors, which isn't overly helpful, but does tell me that CD286 is the series code.

Back to Google and a quick search for CD286 yields a page on the suntan site with a PDF. This has the capacitor in question with height and diameter of the package (for confirmation), and it's low ESR. It has ESR at 2 temperatures and max ripple current.

 

[Tesla]

Here's an example of what things used to be like ...

The other day, I got my Amiga 1080 RGB CRT monitor down out of the attic. It is 22 years old. It was used for 6 years then boxed up. About 8 years ago I setup the system and it worked fine. I boxed it all up and put it in the attic this time.

When I powered up the monitor the other day, all I got was a single 1/4" horizontal scan line. I figured it was going to smoke. But over the next 5 minutes, the picture got bigger and bigger. I'm not sure if the caps were re-forming or just charging but finally the picture was full size. The color, alignment ... everything was fine. If I powered it off and back on, it came on instantly. I ran it for several hours and the guy I gave it to has been using it and it's fine now.

"Good capacitors" I thought to myself... I wonder if todays (consumer device) caps would pass this test? Somehow, I don't think so.

 

[Tesla]

RE: Examples

Example 1:
I'm guessing you got the original part number from a schematic. So what's the application that needs $10 caps?

Example 2:
Yes, SMT is the worst. I know there are some letter charts (that equal large numbers) that are used on small disc caps (because there is so room). Possibly something like that.

Example 3:
All the bad caps I pulled (2 different Chinese makers) had the CDxxx codes on them.

 

[(*steve*)]

I'm guessing you got the original part number from a schematic. So what's the application that needs $10 caps?

No, from an ebay Auction. Half a reel went for about $4.

I don't know of any design that needs $10 caps, but there are many that require low ESR capacitors

Yes, SMT is the worst. I know there are some letter charts (that equal large numbers) that are used on small disc caps (because there is so room). Possibly something like that.

They were Panasonic ECEV1VA470P, but I only know that because I know that I'm not sure how I'd get back to that from the device itself.

All the bad caps I pulled (2 different Chinese makers) had the CDxxx codes on them.

These were replacements for caps in an Acer monitor. So similar story.

 

[Tesla]

Here is a list I found ... but nothing too surprising ... I think we all see the pattern here:
http://www.capacitorlab.com/low-esr-capacitor-manufacturers/

I noticed they mention Digikey and Nichicon (even HE Series) here:
http://www.capacitorlab.com/where-to-buy-capacitors/index.htm

 

[Tesla]

DigiKey shipped my caps today so I should have them later this week I think.

 

[Tesla]

I fixed it!

I went ahead and replaced all (8) 16v caps on the low-voltage side (the 4 bulging from the bottom and leaking ones, plus the other 4) with Nichicon caps (PW & HE Series). The parts cost $4.

The first time I powered it up the SMPS came right on. Under load (with everything connected) here are the old and new voltages being outputted from the repaired SMPS.

Marked: Old -> New
3.3v : 2.2v -> 3.3v
5v : 5.0v -> 5.1v
12v : ? -> 11.1v
-12v : -14.5 -> -11.1v

Obviously, the processor (and/or motherboard in general) needed a higher voltage (like it's intended +3.3v) to boot up. DVDs play and look fine.

Thanks again all for your help. It's been very educational and I'm ready for the next AV component or display now.