Reversed polarized electrolytic capacitor (low voltage)

[neel]

Hi there,

In order to try to fix a TV shutting down after a few minutes I wanted to try to replace some capacitors on the power board (there is another possible cause of the shutdown as the tactile button board is also malfunctionning but I will create another thread for this to avoid mixing up with the capacitor subject).

I have salvaged capacitors with the same characteristics and have tested those before doing the replacement. Those salvaged are axial nippon chemicon (16V 1000µf) and the first test was to check the residual charge by checking voltage.

I was surprised to find a negative voltage on both, -0.03V and -0.3V. I have read up to -0.5V is considered safe for such a capacitor, do you think it is safe to continue testing them by discharging then trying a slow charge with the correct polarity (12V + 66Kohms) ? (I just own a cheap meter so I'm measuring time constant, the best test I can do with my current tools)

I have not touched the one charged at -0.3V but tried measuring the resistance of the other one. I got increasing resistance starting from a few Kohms (which is normal range and behaviour for a capacitor in working condition if I'm right) and this resistance test ended with the capacitor holding a small positive charge (0.1V).


I have attached high res pictures of those 2 capacitors so you can have a look at them too. As you can see both have flat bottom side but the upper plastic caps may be suspicious.

One has some (very) small cracks in the plastic around the wire (could be normal aging of the plastic ?)  and the cap seems also very slightly tilted (could be just imperfect manufacturing ?).

The other one has no cracks, no tilting, but the insulator is bent up at 2 locations around the plastic cap (possible result of venting ?)


I know how dangerous those can be if they explode so I prefer asking your opinion before I eventually continue testing them and using them as a replacement on the TV power board. Would you use such salvaged capacitors with the reversed voltage I measured on them ? IIRC they were salvaged from powered desktop loudspeakers which remained unpowered since several years, so they have been holding this negative charge for a few years. 

Your opinion ? Trash bin or can be tested (with gloves and safety googles) and used relatively safely on the power board if the charging test succeeds ?

[HB9EVI]

I wouldn't use salvaged caps to repair another device with them. TVs run switching mode power supplies, so they need caps with a low ESR, which the ones from a desktop loudspeaker set are maybe not.

To avoid further trouble and considering the low prices of new ones, buy some decent new caps from a reliable source. It may save you time, work and frustration

[neel]

Would there be any risk of damaging the board by using caps with higher ESR than required ?
I understand your argument and agree with you, buying new low ESR caps would be the best decision.

A few weeks ago I saved an old monitor using salvaged caps and it still works after a few days.
Of course the risk of failure is bigger with old salvaged caps but it sounds ok if this is just for a short test, no ?

I'm just not sure yet the caps are what is causing the shutdown on the TV, so it would be nice if those salvaged ones could at least let me perform a short test to verify if this is the root cause or not (now if you tell me high ESR would make it not working anyway, then such a test would be useless of course).


The caps I replaced on the faulty monitor were deformed at the top, so more likely bad.

The caps on the TV do not show such bad looking, excepting one of the 2 I would like to try to replace.
This one has some white dust all over the insulation, and I'm not sure this could indicate a bad cap.

Here is a picture of this cap covered by white "dust":



All the highest capacity caps of this TV power board are capXon brand. After some research this brand seems reputed for bad quality electrolytics caps found on computer motherboards or power supply. Any opinion about this statement on this brand ?

In addition to the white dust, the other reason I wanted to try to replace those 2 caps in particular is because they are located against a transistor equipped with a heatsink, the rest of the caps being further away from any identifiable heat source.

Now I have already unsoldered them and the test I have made (check their resistance, and then perform a time constant test by doing a slow 12V charging with a 66K resistor) seems to indicate both still have the nominal capacity. However I have read a cap could be bad despite it still have nominal capacity, so at this point trying a replacement is the only thing I could do to verify if it solves the issue.

[HB9EVI]

In the end it's your decision.

Personally I don't use any used electrolytic caps for reparation or new projects; many of them you find in all kind of devices are from dubious quality with unknown origin, fake/relabled 105 degrees rating a.s.o. Just because it works out once your way, doesn't mean it always works out.

Like you say yourself, there was a lot about faulty caps in the last decade, so it really makes sense to buy brand caps, with genuine 105 degrees rating.

The ESR is a rather important parameter for caps in SMPS; noisy, rippled DC can cause every possible and impossible erratic behaviour in digital circuits.

[PKTKS]

Would there be any risk of damaging the board by using caps with higher ESR than required ?
I understand your argument and agree with you, buying new low ESR caps would be the best decision.

It depends a LOT upon the YEAR and topology of the device.

In rough raw terms you have:
- CRTs (aka TVs)  flyback SMPS works on a fixed 15KHz freq. mostly - may risk say 90%
- in rare cases (when you need high amounts of power for GIANT TUBES AND POWER AUDIO AMPS...
- in that cases the FLYBACK is usually a 2 stage topology - and the two almost certainly don't use same freq.

How do you know you have a 2 stage topology?
The first one uses HIGH POWER stage like PUSH-PULL configuration in which the load is rectified 

The secondary target is the  typical 15KHz regular flyback they are mostly the forward direct type
Regular flybacks for CRT rarely require POWER driver configs like PUSH-PULL or BRIDGEs

And at 15KHz  ESR is not that critical.
As long as your salvage CAPs are not LEAKING  ( you do check them with a uAmp leakage...)...
you are safe.  NON LEAKING CAPS with same rating and same or higher VOLT RATINGs ARE FINE....

Modern CAPS are even better than old ones
Simple flyback topology with MEDIUM POWER output  don't require TOP utmost stuff
they are flexible and can work with a wide variation window.

Paul

[madires]

My advise is to choose the right cap for each job. You can get away with a quick & dirty replacement for some time (AKA temporary solution) but it will cause problems in the long term. If a circuit, like a SMPSU, requires low-ESR caps replacing bad caps with standard electrolytics will increase the temperature by the higher ESR. The higher temperature causes the cap to age faster. Based on ESR, switching frequency, current, ripple and environment this can happen sooner than some might think. Also some circuits can become unstable when using the wrong caps.

[PKTKS]

These dried out CAPs will not pass a leakage test.

As they are mostly aluminum type ELCOs when they dry the
plate oxide becomes a resistor..  a leakage tester should fail...

HIGH leakage currents are definitely associated with dried caps

As long as your replacement subject is not leaking
and presenting a good 'normal'  value it may fit the job

It has become increasingly harder to find replacement parts nowadays

An insane amount of time I have been wasting just.. TO IDENTIFY !!!
some damn small  parts with no data and no clue marks...

Old school parts used to have a more competent way of marking parts.

That also lead us to choose parts which we can identify and test if OK
Obscure parts are becoming harder and harder even to test on the bench

[cncjerry]

Scope it for ripple first then setup a trigger for when it fails before going through and wholesale replacing caps.  If it is shutting down though, you will only see the dropout and not what's causing it.

[neel]

Thanks for the inputs everyone

The TV I'm trying to repair is a philips 47PFL4606H/12 equipped with a 47" IPS panel from LG (lc470wuh-scb2). It was manufactured in Oct 2011.

Here is a picture of the power board (715g4546-p03-h20-003e) 
The red arrow shows the location of the 2 currently removed capXon caps.
The dirt one was the furthest away from the transistor/heatsink:




Also as already mentionned at the beginning of the thread the tactile board is another suspect for causing the shutdown.

Long story short (well, I'm not good at that  ), this TV was meant to go to the bin a few years ago. Someone from my family brought it back from friends (they had it already hanging outside waiting to be trashed...) so I could have a look at it.

When I tried it, it was powering up but with the OSD flashing on screen, the sound volume going up alone, the channel switching up alone too, and it could also go into standby alone IIRC.

I understood it was most likely a faulty tactile panel and decided to try some "voodoo" to fix this. I had the idea (or maybe had read about somewhere) to stick an aluminum foil on top of the front tactile panel and... it worked. With the right positionning the OSD stopped flashing and channels stopped changing, the TV was not going into standby, and the IR was working so I could access all the TV menus. I disabled the internal sound (this was still causing an OSD message) and added external speakers and the TV was usable again. 

It has been working nicely like that since 1 year or 2 until this summer when it started switching off after a few minutes. Repositionning the aluminum foil doesn't seem to prevent the shutdown, hence why I suspected it could be a power supply issue with the high temp we had this summer and the age of the TV.

So another question: do you know if such tactile panel can be by-passed ?
I haven't tried disconnecting it from the mainboard to check what would happen.

The IR receiver is on the same board and there are 2 connectors going to the mainboard.
Is it possible to disable the tactile part while keeping the IR receiver ?
Or disconnect this tactile panel completely and do a mod somewhere on the mainboard to have a ON/OFF switch ? I don't need anything else as the sound and channel switching is controlled through an external dvb-t box anyway.
I have checked yesterday and couldn't find this tactile board to buy (I found 1 on ebay but will ship only to germany. And I'm not sure it would fix the issue).

Here are pictures of the tactile panel board (715G4958-K02-000-004B) and the mainboard it is connected to (bottom left connectors):

 

[PKTKS]

From your OEM CHASSI indication
i have a matching Service here...

This is NOT a CRT  - indeed this SMPS has nothing to do with
'regular CRT' flybacks.

It uses several DC/DC converters and system management ICs
and MOSFETs power switches.

My advice WILL NOT FIT this device as it was for 'regular CRT' flybacks


Care must be taken as this SMPS operates in rather high freq,
and probably high spikes on MOSFETs lines

[PKTKS]

In this CHASSI number <47PFL4606H>

There are several secondary regulators  but the master piece
seems to be the so called "multi DC/DC"

 in which several current sense resistors may be seen
and may cause problems when bad caps appear

not a trivial simple check but you may still probe the BUCK output
if it shuts down at some particular point

[neel]

I never talked about a CRT, idk why this came into the discussion 

With my basic electronics knowledge the only things I initially planned was trying to replace some capacitors and/or check what I could try with the tactile board. I couldn't do any further investigation without being told what to do.

Above you said "These dried out CAPs will not pass a leakage test.", were you referring to the orginal capXon of my picture ? (green ones, in my 2nd post)

Also any opinion about the salvaged caps with negative voltage ? Should those be considered dangerous or I can try to charge them ?

Another thing I haven't mentionned yet, before opening the TV to check the caps shape I tried to lower the contrast to the minimum to check if it would prevent the shutdown. Doing so I observed a flickering of the image, which is also why I began suspecting a possible weakness at the powerboard. Now this may just be a poor design for the backlit dimming and not indicating any power supply weakness.

[HB9EVI]

common electrolytic caps are polarized, so do not alter the polarity; if you measured a small negative voltage after removing them from the pcb, it doesn't mean anything, you can simply ignore that - but it's no invitation to apply opposite polarity voltage on caps during you're testing them.

if the backlight is flickering, I would suspect the psu as very first; but to tell the truth: just armed with a multimeter you'll probably have some problems tracking down the source of the problem; it'll likely be needed to take a look on the voltages with a scope.

[neel]

Don't worry I had no intention to charge a polarized cap using reversed polarity (I have limited knowledge but still some basic knowledge ).

I know I must only replace a cap with one with at least the same capacity and voltage, and the influence of temperature doubling the life every 10°c below nominal temp (so better use 105°c rated) and so the influence of a higher ESR regarding service life.

I just did not figure out how a capacitor could have a (small) negative voltage if not by receiving it when it was mounted in the previous circuit. But if you say I can ignore that.

I understand the problem may also be harder to find and requiring equipment I don't own, I just want to try what is at my reach, and if it doesn't work, well, I will have to sell it to someone who can repair it or trash it if nobody wants it. This is a TV I got for free and was able to extend its life already as previous owner had planned to trash it, and now I'm just trying my best to try to extend its life more. This is a good TV with good image quality and very suitable for large room, thanks to the large viewing angle of IPS panel (my prefered flat panel technology among TN/VA/IPS, despite the poor black level). 

Note it doesn't flicker at all at normal contrast setting, only when dimming it down close to the minimum.

[HB9EVI]

Oh, I didn't want to offend you, sorry about that; sometimes it's better to warn once too much; but I agree if well planned, an exploding cap has its own charm.

Aside from my IT work, I operate a kind of repair-café without coffee; so it's normally my aim not only get the devices back to work, but to last as long as possible. Electrolytic caps are a recurring issue in every kind of device; so when I stocked up caps, I compared prices; the conclusion was that there is no sense in saving a few cents on cheap brands, 85° instead of 105° or minimal capacity.

One example: often you find 820uF caps in PSUs. The manufacturer can save a few cents with not taking 1000u; it's ridiculous, I take 1000uF. Sure the circuit works well with 820u when new; with the time, with increase of leakage current, it maybe turns out, that this size is at the limit. Sure, if it fails and the device stops working, today people buy new - there's no interest from the manufacturer side to improve lifetime.

If you repair, your aim should be a maximal life time for the device without further trouble - so don't save those cents

[neel]

No offense 

And I agree with you, I will certainly buy new caps for the monitor I have repaired now I know for sure it was faulty caps causing the issue.
And same for this TV if I also find the issue is faulty capacitors, but I would like to be sure of that first.

I remember trying to fix a crt tv years ago, also by replacing all caps. This TV was getting harder and harder to power on from standby, had to press several times the button. I replaced all caps with brand new ones I had bought especially for that and then... the TV wasn't even going into standby mode anymore. Failed repair.

This is probably why I'm now less inclined to buy replacement parts without being sure beforehand those components are what causes the issue. I put my hopes on the low probabilty salvaged components would fail during a very short test to identify the root of the issue, and so just need to ensure I don't try something having high probability of failure.

In other words I just try to ensure the risk I take by using those salvaged components for the diagnostic remains very limited. This is the main scope of my questions to esperienced users here. I don't deny a proper final fix will require new and good quality parts.

[PKTKS]

I never talked about a CRT, idk why this came into the discussion 

Apologize my fault
AS you mention "TV"  it came CRT .

I use the word PANEL or FLAT DISPLAY  not TV for the NON CRT devices.
should explain why - THE CHASSI  is the proper reference. no doubt.

With my basic electronics knowledge the only things I initially planned was trying to replace some capacitors and/or check what I could try with the tactile board. I couldn't do any further investigation without being told what to do.

Replacing caps ADHOC will not help.

The diagram that I have which FITS you CHASSI number shows a HIGH FREQUENCY SMPS
with  several secondary BUCK regulators and several MOSFETs power switches

Any part of this chain may be  the culprit.

By "LEAKAGE" I mean  CURRENT LEAKAGE  - not external  leaking fluids.

As the device ages and dry ... the plate form an oxide which acts as a RESISTOR.

Then the CAP itself start to behave like a resistor - formerly a CAP with a parallel resistor
the so called EPR (equivalent parallel resistor)

this oxide resistor is voltage dependent ..
There are som e methods to revert that process by applying carefully chosen voltages
to the plates - a process sometimes referred as REFORMING

IMHO - it worth nothing.  If a LEAKAGE TEST (current leakage test) fails
e.g. the CAP is leaking current and IS NOT BLOCKING DC..  Just dump it.

There are several testers which will perform:
- ESR test - at several freqs.  - EQUIVALENT SERIES
- EPR test - at several VOLTAGES - EQUIVALENT PARALLEL
- NOMINAL CAPACITANCE value.

The 3 must be OK - but the EPR is **REQUIRED** to be 100% ok
Why?  because CAPS can not LEAK current - MUST BLOCK all DC.

If  you have DC current leaking... Ditch the CAP.
Paul

[PKTKS]

and BTW more 2 cents after looking the diagram..

I would start my routine to troubleshoot this device
looking the "CURRENT SENSING" resistors and caps near by
of the MAIN BUCK REGULATOR.

** IF** any current sense is overloading the REGULATOR WILL SHUTDOWN
we need to discard that chance first..

starting point

[perieanuo]

Replace ALL the electrolytic caps, reassemble the thing and live your life.that's how a pro repair a smps.if you don't, after 1 month you can get "lucky" and repair it again for another bad cap


Envoyé de mon iPad en utilisant Tapatalk

[neel]

Hi there,

Just want to give some news about this TV repair.

I have replaced all the caps from the power board with brand new quality ones, including the main 400V cap, and the TV works again.

Now was this really the issue ? Not sure: after trying the TV after the repair I immediately noticed the remote was not working anymore. Indeed when I had the TV opened I have unplugged the small pcb responsible for the tactile keyboard (that has issues) and that is also responsible for the IR for the remote, so it is very possible I just forgot to plug it back after I have replaced the caps and that was the real cause of the TV shutting down alone.

Also I now own a DMM capable of measuring capacity (a nice Brymen BM257s) and have checked all the caps I have replaced from this TV power board and they all have the nominal capacity (although I understand a capacity check doesn't give the full picture about the health of a capacitor).

I may know more when I will open the TV again to plug this tactile/IR reveiver pcb as I need to access the settings to set the brighness/contrast back as I had those set to the minimum when I was trying to see if this would help to prevent the TV from shutting down before replacing the caps.

BTW I took this opportunity to also buy nice new caps for the samsung PC monitor I had previously repaired using used caps. And I have also just saved an old PC motherboard by again replacing caps^^ (6 caps located around the RAM, they had the top inflated and I measured them having much lower capacity than nominal. This repair is currently running with salvaged caps, waiting for the next order I may have to do as I have another TV (plasma) to repair^^).

[madires]

Yep, it's a good idea to keep some stock of low-ESR electrolytic caps with common values. The next bad SMPSU isn't far away.

[Zero999]

I'd recommend you get an ESR tester and perhaps a meter with a high capacitance range of up to 1000µF, so you can test if the capacitors are good.

Leakage current is failry easy to do, without any special equipment. Charge the capacitor, with a 100R and 1M resistor in series with it. Bypass the 1M with a switch, until the capacitor fully charges. Now measure the voltage across the 1M resistor, with it still connected to the supply and the voltage will be 1V per µA of leakage.

Generally always get new electrolytic capacitors for repairs and projects. I do admit I've used salvage parts in the past, but wouldn't recommend it.

[neel]

Just another short update:

I finally have opened the TV again to try to make the IR receiver work again.
Sadly it was already plugged (I was wondering I may have had forgotten to plug it back the 1st time).

So I have checked the few (4) caps I initially skipped on the video board and they were having nominal capacity.

I have cleaned the double-sided tape from the tactile/IR PCB using alcohol, andalso cleaned the IR receiver.

I have also cleaned the pins from the videoboard socket where the wire from the tactile PCB is plugged using sandpaper, also cleaning the connector holes with alcohol and a needle.

Time to test: it works !

Not only the IR receiver works again but the tactile panel doesn't bug anymore ! The TV is just fully functionnal again now, very happy about that. Sadly I just can't tell what was causing the issue(s). Maybe the tactile PCB wasn't well sticked into its housing, or dirty, or the IR receiver was dirty, or the connection to the videoboard wasn't 100% functional.