Repairing (?) a busted TiVo power supply

[Xylker]

The lead-in: My TiVo was off. Thinking that was odd, I removed the plug, and attempted to reboot it. Click, clunk, clunk. Awww, damn! I thought, "failing HDD." So, in preparation to watch the SuperBowl, I unplugged it to let it cool off. Plugged it back in and up comes the UI. 30 minutes later I came back to another dead box. This weekend I had some time to pull things apart and I had a 1TB WD Green AV drive that I purchased for eventual installation in this system. Well, after pulling the old drive out and trying to try to image it (why not, right?) I discovered that the HDD was fine.

Attached you will find a picture of what appears to be a damaged capacitor.

ninja edit: fixed pic

So, the question is, is that enough to account for the symptoms? The cap is listed as 2200 uF and 25V on the label. I think I can buy one for $2 at Fry's. Anything special I should know before I start heating up the soldering iron?

What about using a computer PSU? The only wire in the bundle I am not familiar with (and a quick search will fix that...) is the gray one.

ninja edit: fixed pic

Do you suppose that this PSU used standard wire colors? If it still worked, I could test. But, alas, it does not still work, so I am potentially killing my MB if those wires aren't ATX PSU-like...

Thoughts? Dire warnings of death and worse?

[JustAnEngineer]

http://www.weaknees.com/tivo-power-supply.php

[just brew it!]

Yeah, if you can get an exact replacement part for the PSU that's your best bet.

If you decide to go the recap route I would also replace any other caps that are the same type, since they are now suspect too even if they are not bulging.

[notfred]

The black and gold ones are low ESR types and those are usually the ones that go rather than the standard electrolytics. Match voltage and uF (go bigger in voltage if you can't match exactly) and only replace low ESR with low ESR. If it's more than just 1 or 2, it may be worth getting them from Digikey rather than Frys.

[tanker27]

http://www.weaknees.com/tivo-power-supply.php

BAH! Just remove and add the caps yourself! Its not hard.

FWIW before my new cablecard enabled TiVo, I have fixed a series one and two this way.

[Captain Ned]

I've got a spare Series 2 power supply sitting on my shelf if you want it.

[Xylker]

Thanks all.

I am familiar with weakknees, but $129 vs. $50 (if I pay myself for the trouble and getting the $2 spare part) I think you can see where that leads...

Ned, I am grateful for the offer, but I have one of those (series 2 PSU) as well. Doesn't work for series 3.

Oh, and for the fun of it... viewtopic.php?f=1&t=56455&p=1084677#p799079
Same TiVo. Maybe I should take the hint that TV is bad for you and just leave this one to die in peace.

Nah!

[sluggo]

The black and gold ones are low ESR types and those are usually the ones that go rather than the standard electrolytics. Match voltage and uF (go bigger in voltage if you can't match exactly) and only replace low ESR with low ESR. If it's more than just 1 or 2, it may be worth getting them from Digikey rather than Frys.

A modest addition. The film wrap color is entirely a vendor's choice and is not a wholly reliable indicator of performance. Nichicon (I believe) started the trend of using more identifiable colors for their products and it caught on with other vendors. Black and gold may mean one thing for a particular vendor and something else for another. The lowest ESR caps are made by Mallory, Sprague, and Sangamo in their computer grade lines, and they all use very plain (blue or white) film.

To the OP: the supply in question is not an ATX-spec supply and so the wire colors are not indicators of ATX-compliant voltages. Red might be +5, or it might not. There's no guarantee unless you know the Tivo spec. Also, the supply provides 30V, which you will not find on an ATX supply, so retrofitting an ATX supply will not be possible.

I'd replace every electrolytic in the supply that's made by the failing part's maker. Just take pictures of the unit before you replace so you can be sure of matching the polarity on the new caps.

[Captain Ned]

I'd replace every electrolytic in the supply that's made by the failing part's maker. Just take pictures of the unit before you replace so you can be sure of matching the polarity on the new caps.

Also, if the Series 3 PSUs are anything like the Series 2 units, there's potting compound all over the place.

[ludi]

I'd replace every electrolytic in the supply that's made by the failing part's maker. Just take pictures of the unit before you replace so you can be sure of matching the polarity on the new caps.

Emphasis on this point, or even better, take a fine-point Sharpie marker and physically mark the negative (-) polarity stripe on the top side of the circuit board before proceeding. It's not uncommon for the silkscreen to have labeling errors that were discovered and corrected somewhere in between producing 100k bare boards and final component assembly.

[Xylker]

Fry's looks like it is going out of business.

At least from the capacitor selection I saw last night. They had a little poster on the aisle endcap: Common capacitors for your HDTV power supply, listed was the 2200 uF 25V part that I need. On the shelves (ok, wires sticking out of pegboard) there were ~120 varieties and 15 of those were actually stocked...

[Xylker]

Booya!

Thanks for the help and suggestions. It was good to know that a 50V cap would not hurt me.
(Radio Shack also was nearly sold out... only one way too big cap remained)

If you are interested, I can show more pics of the repair itself. But it is U-G-L-Y!

[just brew it!]

It was good to know that a 50V cap would not hurt me.

Yeah, the thing people who haven't worked with component level electronics frequently don't realize up front is that cap voltage isn't like voltage on a power supply or battery. Having a higher voltage cap won't fry anything, because a cap only puts out as much voltage as is put into it; the voltage rating is just the maximum safe voltage the cap can be used at.

[notfred]

the voltage rating is just the maximum safe voltage the cap can be used at

Before it blows the body off the end, possibly taking an eye out on the way past, and sprays the rest of the components with its insides.

[Xylker]

the voltage rating is just the maximum safe voltage the cap can be used at

Before it blows the body off the end, possibly taking an eye out on the way past, and sprays the rest of the components with its insides.

Awesome!

[just brew it!]

Installing a polarized cap backwards will result in some "interesting" things happening as well...

[thegleek]

If you are interested, I can show more pics of the repair itself. But it is U-G-L-Y!

Yes, please!

[DLHM]

I would just replace the caps. I'm sure a local radio shack would have the part. Also I have one of these and it comes in handy and works pretty well. http://www.ebay.com/itm/6013-autoranging-capacitance-capacitor-tester-up-47mF-/320746967595?pt=LH_DefaultDomain_0&hash=item4aae02522b

Edit: Fixed some horrible spelling..

[ludi]

It was good to know that a 50V cap would not hurt me.

Yeah, the thing people who haven't worked with component level electronics frequently don't realize up front is that cap voltage isn't like voltage on a power supply or battery. Having a higher voltage cap won't fry anything, because a cap only puts out as much voltage as is put into it; the voltage rating is just the maximum safe voltage the cap can be used at.

At the same time, it is worth noting three points about electrolytics that a novice user might not know:

1. The capacitor's energy storage rating is proportionate to the voltage rating, so way over-spec'ing a capacitor's voltage rating relative to its application results in lost capacity, and therefore less filtering/storage effect. RETRACTED - SEE CONTINUING DISCUSSION BELOW.

2. ESR (equivalent series resistance) usually increases proportionate to both the energy storage rating and the voltage rating, so once again, massively over-spec'ing the unit relative to its application can actually harm performance rather than help it.

3. The capacitor's ripple current rating usually increases proportionate to the energy storage rating and the voltage rating, so some over-spec'ing can improve the circuit's reliability.

Fortunately, the physical size of the unit usually increases proportionate to its energy storage and voltage ratings, which usually prevents anyone from going too wild because the capacitor soon won't fit in the location designed for it. I usually figure a +20% over-spec on either of those ratings is tolerable for most applications, especially since electroyltic capacitors are not very precise devices anyway.

[just brew it!]

@ludi -

Agree with #2 and #3, but #1 doesn't make sense. The lost capacity is only relative to the maximum energy the capacitor can theoretically store if charged to its maximum rated voltage, which does not affect the amount of energy stored at a given (lower) voltage. Energy stored in a capacitor is a function of the actual voltage across the capacitor and the capacitance.

[ludi]

@ludi -

Agree with #2 and #3, but #1 doesn't make sense. The lost capacity is only relative to the maximum energy the capacitor can theoretically store if charged to its maximum rated voltage, which does not affect the amount of energy stored at a given (lower) voltage. Energy stored in a capacitor is a function of the actual voltage across the capacitor and the capacitance.

Okay, after chewing on it a bit, I think I follow what you're saying but it's been a while since Physics I, and Wikipedia's description of the Farad is ambiguous. Supposing, for example, two simple capacitors rated 1000uF (0.001F) at 25V and 50V, respectively. Do both charge or discharge at the rate of 1V/s for 1mA of current, but the latter has the capacity to charge to (and discharge from) twice as much voltage?

[just brew it!]

@ludi -

Agree with #2 and #3, but #1 doesn't make sense. The lost capacity is only relative to the maximum energy the capacitor can theoretically store if charged to its maximum rated voltage, which does not affect the amount of energy stored at a given (lower) voltage. Energy stored in a capacitor is a function of the actual voltage across the capacitor and the capacitance.

Okay, after chewing on it a bit, I think I follow what you're saying but it's been a while since Physics I, and Wikipedia's description of the Farad is ambiguous. Supposing, for example, two simple capacitors rated 1000uF (0.001F) at 25V and 50V, respectively. Do both charge or discharge at the rate of 1V/s for 1mA of current, but the latter has the capacity to charge to (and discharge from) twice as much voltage?

Yup, that pretty much sums it up.

All else being equal (ESR, leakage current, parasitic inductance...), a 25V capacitor and a 50V capacitor will behave identically in circuits up to 25V. Above 25V the lower-rated capacitor is at risk for going BZZZT (or BOOM), while the 50V one keeps on truckin'. That's pretty much it.

[ludi]

Okay, I "get" it, then. Modified my earlier post to note it.

[danplaysbass]

Hi guys. I stumbled upon this thread while trying to repair my series 3 tivo power supply. I read all about the bad capacitors so I was prepared to change those out and be back in business. My 12v rail is only reading about 7.5v so I started there. I replaced 2 of the 2200uf caps.. One was rated 16v and the other 25v, and these were the only 2 rated above 12 volts. The original caps didn't show any signs of busting.

This didn't fix the output and I can still hear the slight ticking from the switching supply. I can't find a schematic and I'm not sure where to look next. Any guidance?

BTW, I am an electrical engineer so I should be able to handle anything you can throw at me.

Thanks in advance.

Dan

[just brew it!]

I suppose you could try checking all the other caps for shorts.

[Xylker]

Interesting observation about the "ticking." I had the same issue after re-capping. In my case I chalked it up to the increased "capacity" (ugh) and let it sit for ~5 minutes, maybe less, then the noise seemed to stop.

[Irepoder]

Any luck with your 3rd gen ps? I have the same ticking after replacing all the caps.

Thanks,

Mark