unable to fix faulty capacitor? pls help!

[Peter]

I have an electric piano that recently developed what I believed to be a
faulty capacitor.

The symptoms were that during playing for a few minutes I would start
hearing a sort of crackling sound coming from the speakers.
Then soon after I would hear a quite loud "pop" sound through the speakers
and then the sound would go dead.

I could still hear faint hissing through the speakers (as when the volume is
up but nothing is being played) but hitting the keys would produce no sound.
When I turned the piano off for a few minutes to cool it down, it would play
again.

I could also smell kind of "fish" smell coming from the piano.

Anyway I opened it up and when it went dead again I sprayed all the caps one
by one with a freezer spray.
Strangely that did not help to locate the faulty cap - if that was the
problem at all.

I started removing the caps from the board one by one and testing them with
a multimeter but they all seemed OK except one that seemed to change the
capacitance slightly when warmed up. I replaced it with a new one (same
values, but higher temperature tolerance).
The cap was 6800uF rated at 25V and it was the biggest one on the board,
located near the amplifier IC.

Well the difference is that unlike before when the sound goes dead now, it
quickly comes back within a second or so. No need to even power down the
piano. But I still get the loud "pop" every now and then.

At this stage I am at loss as to what else I could try. Freezer spray will
not help me now at all, since the problem goes away within seconds by itself
(just to reappear a few minutes later)

Should I remove and test all caps? Seems like a lot of work with an
uncertain result.

I am hoping that someone with more experience (I am not a technician) might
have some idea as to what may be causing the problem and how to proceed.

Thank you for any help you could offer. The piano is Yamaha Clavinova
CVP-85A and is way past the warranty.

Peter

 

[David]

So what were the results of the ESR meter test on the capacitors?
I am not even sure how you can test a capacitor with a multimeter and get
any kind of useable result for a failing capacitor. Beyond maybe a shorted
one.

David

 

[Peter]

Sorry, I meant a capacitance meter of course.

The result for all of them were very close to the specifications, except the
one I replaced. That one was showing higher readings after I held it my
hand for a while. But I guess that is not much an indication. However
replacing it did change the symptoms slightly but did not fix the problem.

Peter

 

[The Real Andy]

Sorry, I meant a capacitance meter of course.

The result for all of them were very close to the specifications, except the
one I replaced. That one was showing higher readings after I held it my
hand for a while. But I guess that is not much an indication. However
replacing it did change the symptoms slightly but did not fix the problem.

Peter

<snip>

Sound like a cap that is in the power supply. See if you can locate a
rectifier, or perhaps the four diodes that make a rectifier and
replace. A faulty rectifier can often stress a cap to the point of
failure.

 

[Peter]

<snip>

Sound like a cap that is in the power supply. See if you can locate a
rectifier, or perhaps the four diodes that make a rectifier and
replace. A faulty rectifier can often stress a cap to the point of
failure.

Thank you, that sounds like an idea worth trying.

I can locate the rectifier no problem (a square bit with four legs I think),
but I have a question since as you can probably tell I am still learning as
I go along.
If the rectifier was for the power supply and it was faulty, wouldn't that
stop everything from working?

Or is it that since the rectifier is faulty, it immediately damaged (or
stressed above the limits) the new capacitor I put in?
Secondly, is there a way to test a rectifier to see if it faulty?

Thanks again for your help. I think there may be a light at the end of the
tunnel yet.

 

[Art]

Foprward / Reverse resistance is normally a reliable rectifier check.
However it will not indicate failure under load, if in question it would
probably be best to replace the device in question and the filter caps in
the same circuit. However don't attempt this after a couple "Fosters" Eh,
cheers gov.

 

[The Real Andy]

Thank you, that sounds like an idea worth trying.

I can locate the rectifier no problem (a square bit with four legs I think),
but I have a question since as you can probably tell I am still learning as
I go along.
If the rectifier was for the power supply and it was faulty, wouldn't that
stop everything from working?

If one diode in the rectifier goes open circuit, it can create enough
ripple to damage a cap over time.

The cap will try and smooth the ripple. If you draw to much power from
the cap, it will discharge and will take time to charge up again, the
more ripple, the longer the charge time. Likewise, the more power you
draw from the cap the longer it will take to charge. If you have an
open circuit doide in the rectifier, it may only be charging it half
the time.

In AU (i think you are there) the mains is 50hz. When you rectify a
50hz sinewave you invert the negative part of the sinewave into a
positive, effectively giving you a 100hz voltage, like here
http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/rectbr.html
If one diode is open, you only get half of that voltage, just like the
second diagram on this page
http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/rectifiers.html

Or is it that since the rectifier is faulty, it immediately damaged (or
stressed above the limits) the new capacitor I put in?

The new cap will be fine.

Secondly, is there a way to test a rectifier to see if it faulty?

You can test using the diode range on the multimeter. If you don't
have that you can test it using the ohms range. Each diode in the
rectifier should only conduct one way. If it doesn't conduct either
way it has an open circuit diode.

The bridge rectifier is drawn here
http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/rectbr.html#c2

 

[The Real Andy]

On Thu, 10 Jun 2004 20:08:17 +1000, The Real Andy
<snip>

This page has some info on testing diodes with an ohm meter.

http://www.ibiblio.org/obp/electricCircuits/Semi/SEMI_3.html

 

[The Real Andy]

Foprward / Reverse resistance is normally a reliable rectifier check.
However it will not indicate failure under load, if in question it would
probably be best to replace the device in question and the filter caps in
the same circuit. However don't attempt this after a couple "Fosters" Eh,
cheers gov.

Despite commom belief, Fosters is rarely consumed here in Australia.
In fact, it is the worst tasting beer I have ever consumed, aside from
a few of my experimental homebrewed beers!

PS, I have replaced many a rectifer under the influence of a fine
beer. As suggested though, the test may not prove anything as the
rectifier is not under load on a multimeter. But in my professional
beer swilling opnion I believe that 99% of dud rectifiers can be
checked on a multimeter!

 

[Jerry Greenberg]

You must also use an ESR meter to properly test caps, not a DVM. You
can use a standard cap meter just to know the cap value itself, but
not its performance.

You obviously have another component that is defective. It may not
necessarily be a cap.

You may have to follow up the signal path with a scope to see where
the actual failure is.

Your other alternative is to give the unit out to a service center
that works on the type of unit that you have. This way it will be
serviced properly.


Jerry G.

==

 

[Peter]

Hehe, nah who in AUS drinks Foster's?
Come to think of it I don't even remember seeing it on tap in SYD.

OK, I did measure it and it checks out OK, but as you said it may be a
different story under load.
I will try to find a replacement for it and see if that helps.

Thank you for your help!

 

[Rheilly Phoull]

I have an electric piano that recently developed what I believed to be a
faulty capacitor.

The symptoms were that during playing for a few minutes I would start
hearing a sort of crackling sound coming from the speakers.
Then soon after I would hear a quite loud "pop" sound through the speakers
and then the sound would go dead.

I could still hear faint hissing through the speakers (as when the volume is
up but nothing is being played) but hitting the keys would produce no sound.
When I turned the piano off for a few minutes to cool it down, it would play
again.

I could also smell kind of "fish" smell coming from the piano.

Anyway I opened it up and when it went dead again I sprayed all the caps one
by one with a freezer spray.
Strangely that did not help to locate the faulty cap - if that was the
problem at all.

I started removing the caps from the board one by one and testing them with
a multimeter but they all seemed OK except one that seemed to change the
capacitance slightly when warmed up. I replaced it with a new one (same
values, but higher temperature tolerance).
The cap was 6800uF rated at 25V and it was the biggest one on the board,
located near the amplifier IC.

Well the difference is that unlike before when the sound goes dead now, it
quickly comes back within a second or so. No need to even power down the
piano. But I still get the loud "pop" every now and then.

At this stage I am at loss as to what else I could try. Freezer spray will
not help me now at all, since the problem goes away within seconds by itself
(just to reappear a few minutes later)

Should I remove and test all caps? Seems like a lot of work with an
uncertain result.

I am hoping that someone with more experience (I am not a technician) might
have some idea as to what may be causing the problem and how to proceed.

Thank you for any help you could offer. The piano is Yamaha Clavinova
CVP-85A and is way past the warranty.

Peter

For the low cost of the components, it is the policy of many techs to
replace all electrolytics in the power supply (or anywhere the problem may
be thought to be) in old equipment. I feel that is a practical approach, by
the same token a new bridge wouldnt bankrupt you either. Also that approach
eliminates some components when you are trouble shooting.

 

[Peter]

If one diode in the rectifier goes open circuit, it can create enough
ripple to damage a cap over time.

The cap will try and smooth the ripple. If you draw to much power from
the cap, it will discharge and will take time to charge up again, the
more ripple, the longer the charge time. Likewise, the more power you
draw from the cap the longer it will take to charge. If you have an
open circuit doide in the rectifier, it may only be charging it half
the time.

In AU (i think you are there) the mains is 50hz. When you rectify a
50hz sinewave you invert the negative part of the sinewave into a
positive, effectively giving you a 100hz voltage, like here
http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/rectbr.html
If one diode is open, you only get half of that voltage, just like the
second diagram on this page
http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/rectifiers.html




The new cap will be fine.


You can test using the diode range on the multimeter. If you don't
have that you can test it using the ohms range. Each diode in the
rectifier should only conduct one way. If it doesn't conduct either
way it has an open circuit diode.

The bridge rectifier is drawn here
http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/rectbr.html#c2

Thank you for the info and the links.
I had a look at them and here is what I have found out.

When I looked at the circuit again I got very excited because I noticed that
the rectifier is actually linked to the cap I just replaced.
So I knew I was on the right track.

But when I removed it from the circuit and measured it, the rectifier
checked out OK.
I measured about 0.48V in one direction and 0V in reverse between each two
pins.
However as someone else in this thread has suggested, it may not show as
faulty unless it is under load.

So I guess the best thing to do is to replace it and see if that helps?

Thank you again, really appreciate it.

 

[The Real Andy]

On Thu, 10 Jun 2004 18:32:42 +0800, "Rheilly Phoull"

For the low cost of the components, it is the policy of many techs to
replace all electrolytics in the power supply (or anywhere the problem may

Where is the fun in that?

be thought to be) in old equipment. I feel that is a practical approach, by
the same token a new bridge wouldnt bankrupt you either. Also that approach
eliminates some components when you are trouble shooting.

It's all about learning something new. Whilst I rarely repair anything
these days, I still enjoy getting my hands dirty in the workshop to
see if I can find those faulty parts without replacing anything. I do
admit it wastes time but the boys in the workshop think I'm good!!!
There is nothing like finding the faulty part and telling the
apprentice to replace it, then fire up to see it work!

 

[Peter]

Despite commom belief, Fosters is rarely consumed here in Australia.
In fact, it is the worst tasting beer I have ever consumed, aside from
a few of my experimental homebrewed beers!

PS, I have replaced many a rectifer under the influence of a fine
beer. As suggested though, the test may not prove anything as the
rectifier is not under load on a multimeter. But in my professional
beer swilling opnion I believe that 99% of dud rectifiers can be
checked on a multimeter!

I am having real difficulties sourcing a replacement rectifier.
Not to mention that if 99% of them can be tested out of the circuit and this
one checks OK it probably isn't faulty?

It is S4VB20, manufactured by Shindengen in Japan
(http://www.shindengen.com/content.cfm?contentalias=productsearchresults&pro
ductCode=s4vb20)

I have tried Jaycar, WES and Farnell, but no one seems to stock it.

I was wondering, if I replaced it with a similarly speced rectifier, how
likely is it to affect the circuit?

 

[The Real Andy]

I am having real difficulties sourcing a replacement rectifier.
Not to mention that if 99% of them can be tested out of the circuit and this
one checks OK it probably isn't faulty?

It is S4VB20, manufactured by Shindengen in Japan
(http://www.shindengen.com/content.cfm?contentalias=productsearchresults&pro
ductCode=s4vb20)

I have tried Jaycar, WES and Farnell, but no one seems to stock it.

I was wondering, if I replaced it with a similarly speced rectifier, how
likely is it to affect the circuit?

Thats just a standard 4A rectifier. Take it to dick smith or jaycar
and find a 4a unit that will fit. IF jaycar's ZR1314 will fit, then
use it.

A long link, but it might work. You may have to cut and paste the
lines as the link may spread itself over multiple lines depending on
your newsreader.
http://www1.jaycar.com.au/productVi...d2=&pageNumber=&priceMin=&priceMax=&SUBCATID=


BTW. As you may be aware, the problem may not be the rectifier, but
with out replacing it, you cant rule it out.

 

[Peter]

o

Thats just a standard 4A rectifier. Take it to dick smith or jaycar
and find a 4a unit that will fit. IF jaycar's ZR1314 will fit, then
use it.

A long link, but it might work. You may have to cut and paste the
lines as the link may spread itself over multiple lines depending on
your newsreader.
http://www1.jaycar.com.au/productView.asp?ID=ZR1314&CATID=&keywords=rectifie
r&SPECIAL=&form=KEYWORD&ProdCodeOnly=&Keyword1=&Keyword2=&pageNumber=&priceM
in=&priceMax=&SUBCATID=


BTW. As you may be aware, the problem may not be the rectifier, but
with out replacing it, you cant rule it out.

Thank you for the info. Yes, that looks like it will fit in and if the new
rectifier can be a substitute than it is definitely worth trying.

Assuming that the rectifier is not the problem, which probably is the case,
how would one proceed further?
What I mean is that the board is not very complicated but has a fair number
of caps on it. I can replace all of them, but that would be an overkill
with uncertain result IMO.

I followed links around the rectifier (and the cap) and noticed that other
than several ceramic caps there is not much else immediately near by.
There is something that looks like a triac (I don't think it is one though)
and has 1D2Z1 written on it. I am currently researching what it might be.

Other than the caps I can see a number or triacs attached to a huge heat
sink. They have these numbers on them: 7805A and M3101G.
I don't have enough experience, but something is telling me that since these
are under constant heat stress they are likely to fail at some point.
I am not sure how they would affect the circuit though. I thought that is
one failed the whole circuit would stop working altogether.

The way I see it, something is definitely affecting the cap that I replaced.
And I think triacs either work or they don't so the effect of a failed one
would be constant. Is that a wrong assumption?

I must say that this detective work is really quite interesting, plus I am
learning quite a bit along the way.

 

[Peter]

o

Thats just a standard 4A rectifier. Take it to dick smith or jaycar
and find a 4a unit that will fit. IF jaycar's ZR1314 will fit, then
use it.

A long link, but it might work. You may have to cut and paste the
lines as the link may spread itself over multiple lines depending on
your newsreader.
http://www1.jaycar.com.au/productView.asp?ID=ZR1314&CATID=&keywords=rectifie
r&SPECIAL=&form=KEYWORD&ProdCodeOnly=&Keyword1=&Keyword2=&pageNumber=&priceM
in=&priceMax=&SUBCATID=


BTW. As you may be aware, the problem may not be the rectifier, but
with out replacing it, you cant rule it out.

OK, replaced the rectifier with the above one from Jaycar, but the symptoms
are exactly as they were before replacing the cap. :-(

A few minutes after turning the piano on, I get a "pop" sound through the
speakers and all goes silent. Turning the piano off and then on again
brings it back to life and it works for a while.
And I can still smell the strong fishy smell coming from somewhere but I
cannot locate it. Very strange. No, I don't have a dead fish in there
It must be some component overheating or leaking, but which one?

Is replacing all the caps the next step? It seems like hitting a mosquito
with a hammer.

 

[[email protected]]

There is something that looks like a triac (I don't think it is one though)

and has 1D2Z1 written on it. I am currently researching what it might be.

The part number sounds like a diode. The number crosses to an NTE116
http://www.nteinc.com/specs/100to199/NTE116.html. This doesn't look
like a triac though.

Other than the caps I can see a number or triacs attached to a huge heat
sink. They have these numbers on them: 7805A and M3101G.

The 7805A is probably a 5volt, 1 amp regulator. In a T0-220 case
(tabbed block with three legs) Pin 1 is DC input, pin 2 goes to ground
and pin 3 is output. The 78xx series is very robust and protects
itself from overload/overtemp by shutting down. Pretty cheap, too.

When you have the problem, check if you still have 5-volts at the
7805A output. It's possible the reg is faulty, but usually they just
die altogether from a quick overload that destroys it before the
protection kicks in. Make sure it's still attached to the heatsink if
there is one.

Not sure what the M3101G is.

I don't have enough experience, but something is telling me that since these
are under constant heat stress they are likely to fail at some point.
I am not sure how they would affect the circuit though. I thought that is
one failed the whole circuit would stop working altogether.

The way I see it, something is definitely affecting the cap that I replaced.
And I think triacs either work or they don't so the effect of a failed one
would be constant. Is that a wrong assumption?

A bad filter cap or bad rectifier in front of the 7805 will definitely
heat it up faster than normal and possibly cause shutdowns. It's
possible that the rectifier is bad and killed the original cap.
Following this reasoning, replacing the cap only helped smoth out the
problem, but it's still there.

I must say that this detective work is really quite interesting, plus I am
learning quite a bit along the way.

Kinda fun learning electronics, eh?

-Chris

 

[Asimov]

"Peter" bravely wrote to "All" (13 Jun 04 12:21:29)
--- on the heady topic of "Re: unable to fix faulty capacitor? pls help!"

Pe> From: "Peter" <[email protected]>

Pe> OK, replaced the rectifier with the above one from Jaycar, but the
Pe> symptoms are exactly as they were before replacing the cap. :-(

Pe> A few minutes after turning the piano on, I get a "pop" sound through
Pe> the speakers and all goes silent. Turning the piano off and then on
Pe> again brings it back to life and it works for a while.
Pe> And I can still smell the strong fishy smell coming from somewhere but
Pe> I cannot locate it. Very strange. No, I don't have a dead fish in
Pe> there It must be some component overheating or leaking, but which
Pe> one?
Pe> Is replacing all the caps the next step? It seems like hitting a
Pe> mosquito with a hammer.

One time when I was stuck trying to find a faulty component my little
girl asked me why I didn't simply replace all the parts in the
affected area. She made perfect sense since it was only a handful of
resistors and capacitors to replace. Sometimes a brute force approach
is more profitable than the actual time invested in troubleshooting.
It's a philosophical point that a professional technician can
appreciate since labour is usually more costly than the parts. And, as
an aside, a defective component tends to stress the others around it
too, for example by overheating them, etc.

A*s*i*m*o*v

.... Real techs don't lick nine-volt batteries!