I currently have a Peavey PVX-P12 amp chassis on the bench. There was no
audio at all from either the HF or LF digital output ICs. The IC in question
has a 'go / mute / standby pin and this pin on both ICs is connected to a
little control circuit. That in turn is driven by the "AC Detect" circuit on
the power supply. This bit of circuitry is very straightforward, and
comprises a bridge fed with AC from the line input, via a 0.47 uF 275 v ~
x-class cap. Across the output of the bridge, are two caps in parallel,
shown on the schematic as being 22uF at 25 v working. But here's the thing.
They are not shown as being polarised, nor is there any marking that I can
see on the caps themselves, that indicate any polarity. They are surface
mount and small - approx 3.6 x 2.6 x 1.8 mm - and look just like a typical
sm ceramic cap. Pink-y glazed body with conventional 'end caps'. One of
these two capacitors is short circuit. With it removed, everything returns
to normal, and the amps both un-mute. The circuit only produces a few volts,
and this is used to drive the LED in an opto via a 47 ohm R, the transistor
side of the opto being the "AC Detect" signal that connects to the mute
control circuit on the amp board.

So what type of caps are these ? They seem awfully small for any kind of
solid dielectric non-polarised cap of that value, and that sort of voltage
rating. I've had a look around at a few component supplier's offerings, and
can't find anything that seems to match. I checked the capacitance of the
one that isn't short, and it came up at 18 uF, so a bit low, but in the
ballpark of what it says on the schematic. Given that these caps only seem
to serve as the filter for the bridge output, and that there is only a few
volts across them in normal operation, can anyone see any potential problems
with replacing them with a pair of 'conventional' 22 uF 16 v polarised
tants, obviously taking care to put them the right way round ?

Arfa

Arfa Daily wrote:

I currently have a Peavey PVX-P12 amp chassis on the bench. There was no
audio at all from either the HF or LF digital output ICs. The IC in question
has a 'go / mute / standby pin and this pin on both ICs is connected to a
little control circuit. That in turn is driven by the "AC Detect" circuit on
the power supply. This bit of circuitry is very straightforward, and
comprises a bridge fed with AC from the line input, via a 0.47 uF 275 v ~
x-class cap. Across the output of the bridge, are two caps in parallel,
shown on the schematic as being 22uF at 25 v working. But here's the thing.
They are not shown as being polarised, nor is there any marking that I can
see on the caps themselves, that indicate any polarity. They are surface
mount and small - approx 3.6 x 2.6 x 1.8 mm - and look just like a typical
sm ceramic cap. Pink-y glazed body with conventional 'end caps'. One of
these two capacitors is short circuit. With it removed, everything returns
to normal, and the amps both un-mute. The circuit only produces a few volts,
and this is used to drive the LED in an opto via a 47 ohm R, the transistor
side of the opto being the "AC Detect" signal that connects to the mute
control circuit on the amp board.

So what type of caps are these ? They seem awfully small for any kind of
solid dielectric non-polarised cap of that value, and that sort of voltage
rating. I've had a look around at a few component supplier's offerings, and
can't find anything that seems to match.

** I found lots - eg:

http://www.newark.com/murata/grm32er...x5r/dp/24R6350



I checked the capacitance of the
one that isn't short, and it came up at 18 uF, so a bit low, but in the
ballpark of what it says on the schematic. Given that these caps only seem
to serve as the filter for the bridge output, and that there is only a few
volts across them in normal operation, can anyone see any potential problems
with replacing them with a pair of 'conventional' 22 uF 16 v polarised
tants, obviously taking care to put them the right way round ?


** Should work fine.

The SMD caps are cheaper than tants of the same ratings.


..... Phil

"Phil Allison" wrote in message
...
Arfa Daily wrote:

I currently have a Peavey PVX-P12 amp chassis on the bench. There was no
audio at all from either the HF or LF digital output ICs. The IC in
question
has a 'go / mute / standby pin and this pin on both ICs is connected to a
little control circuit. That in turn is driven by the "AC Detect" circuit
on
the power supply. This bit of circuitry is very straightforward, and
comprises a bridge fed with AC from the line input, via a 0.47 uF 275 v
~
x-class cap. Across the output of the bridge, are two caps in parallel,
shown on the schematic as being 22uF at 25 v working. But here's the
thing.
They are not shown as being polarised, nor is there any marking that I
can
see on the caps themselves, that indicate any polarity. They are surface
mount and small - approx 3.6 x 2.6 x 1.8 mm - and look just like a
typical
sm ceramic cap. Pink-y glazed body with conventional 'end caps'. One of
these two capacitors is short circuit. With it removed, everything
returns
to normal, and the amps both un-mute. The circuit only produces a few
volts,
and this is used to drive the LED in an opto via a 47 ohm R, the
transistor
side of the opto being the "AC Detect" signal that connects to the mute
control circuit on the amp board.

So what type of caps are these ? They seem awfully small for any kind of
solid dielectric non-polarised cap of that value, and that sort of
voltage
rating. I've had a look around at a few component supplier's offerings,
and
can't find anything that seems to match.

** I found lots - eg:

http://www.newark.com/murata/grm32er...x5r/dp/24R6350


Thanks, Phil. Oddly, I searched in Farnell, which is basically the same
company as Newark, and these caps did not appear. But sticking the Murata
part number into the Farnell search engine - there they are. Just goes to
show that what I have always said - that Farnell's product searcher is
useless - has been borne out again ...



I checked the capacitance of the
one that isn't short, and it came up at 18 uF, so a bit low, but in the
ballpark of what it says on the schematic. Given that these caps only
seem
to serve as the filter for the bridge output, and that there is only a
few
volts across them in normal operation, can anyone see any potential
problems
with replacing them with a pair of 'conventional' 22 uF 16 v polarised
tants, obviously taking care to put them the right way round ?


** Should work fine.

Yeah, I thought so. The guy is hoping to use it tonight so I think I will
just go with the tants

Arfa



The SMD caps are cheaper than tants of the same ratings.


.... Phil

On 19/06/2015 02:31, Arfa Daily wrote:
I currently have a Peavey PVX-P12 amp chassis on the bench. There was no
audio at all from either the HF or LF digital output ICs. The IC in
question has a 'go / mute / standby pin and this pin on both ICs is
connected to a little control circuit. That in turn is driven by the "AC
Detect" circuit on the power supply. This bit of circuitry is very
straightforward, and comprises a bridge fed with AC from the line input,
via a 0.47 uF 275 v ~ x-class cap. Across the output of the bridge,
are two caps in parallel, shown on the schematic as being 22uF at 25 v
working. But here's the thing. They are not shown as being polarised,
nor is there any marking that I can see on the caps themselves, that
indicate any polarity. They are surface mount and small - approx 3.6 x
2.6 x 1.8 mm - and look just like a typical sm ceramic cap. Pink-y
glazed body with conventional 'end caps'. One of these two capacitors is
short circuit. With it removed, everything returns to normal, and the
amps both un-mute. The circuit only produces a few volts, and this is
used to drive the LED in an opto via a 47 ohm R, the transistor side of
the opto being the "AC Detect" signal that connects to the mute control
circuit on the amp board.

So what type of caps are these ? They seem awfully small for any kind of
solid dielectric non-polarised cap of that value, and that sort of
voltage rating. I've had a look around at a few component supplier's
offerings, and can't find anything that seems to match. I checked the
capacitance of the one that isn't short, and it came up at 18 uF, so a
bit low, but in the ballpark of what it says on the schematic. Given
that these caps only seem to serve as the filter for the bridge output,
and that there is only a few volts across them in normal operation, can
anyone see any potential problems with replacing them with a pair of
'conventional' 22 uF 16 v polarised tants, obviously taking care to put
them the right way round ?

Arfa

Why would they have to be unpolarised? Just 2 paralleled together to get
the required capacitance. Any polarising mark on the SM carrier strip
perhaps

Why would they have to be unpolarised? Just 2 paralleled together to get
the required capacitance. Any polarising mark on the SM carrier strip
perhaps


I don't think that they *do* have to be un-polarised, as such, given the
ones that Phil has found. I think it's just because they *are*, if you see
what I mean. To be honest, I never knew that you could get multilayer
ceramics in such high values. Presumably, they don't go up as high as 47 uF
which the designer felt he needed, so he just put 2 x 22 uF in parallel
instead. As I said, the circuit is absolutely un-critical in that it only
needs to produce a small amount of DC from the incoming mains to power the
LED in the opto that produces the muting signal. Because of time constraints
with ordering-in those ceramics, I think I'm just going to go with a pair of
small tants, fitted to observe their polarity.

Arfa

On Friday, June 19, 2015 at 4:04:09 AM UTC-4, Arfa Daily wrote:


I don't think that they *do* have to be un-polarised, as such, given the
ones that Phil has found. I think it's just because they *are*, if you see
what I mean.

That's correct. They are not polarized because that's the technology. Electrolytics are generally polarized because that's the tradeoff in the technology at the time. You can always use a non polarized cap in place of a polarized.


To be honest, I never knew that you could get multilayer
ceramics in such high values.

The first time I ran across these caps was about 10 years ago in LCD TVs on the tcon board. Six to ten paralleled in a bypass on the main dc-dc convertor to give several hundred uf of capacitance. Take up no room at all. Not as short happy as tants but they short more than typical electrolytics.

Because of time constraints
with ordering-in those ceramics, I think I'm just going to go with a pair of
small tants, fitted to observe their polarity.

If they fit and you can keep the leads short, no problem.

On 19/06/2015 09:04, Arfa Daily wrote:




Why would they have to be unpolarised? Just 2 paralleled together to
get the required capacitance. Any polarising mark on the SM carrier
strip perhaps


I don't think that they *do* have to be un-polarised, as such, given the
ones that Phil has found. I think it's just because they *are*, if you
see what I mean. To be honest, I never knew that you could get
multilayer ceramics in such high values. Presumably, they don't go up as
high as 47 uF which the designer felt he needed, so he just put 2 x 22
uF in parallel instead. As I said, the circuit is absolutely un-critical
in that it only needs to produce a small amount of DC from the incoming
mains to power the LED in the opto that produces the muting signal.
Because of time constraints with ordering-in those ceramics, I think I'm
just going to go with a pair of small tants, fitted to observe their
polarity.

Arfa

So to get that capacity in that package , they must be seriously multi
MLCC. Was it actually dead short or very low resistance. I suspect the
usual metal migration failure mode of MLCC , then paralling-up just
increases the chance of failure , for the circuit. Someone mentioned
paralling ten , asking for trouble.

So to get that capacity in that package , they must be seriously multi
MLCC. Was it actually dead short or very low resistance.

Very low resistance, I guess. Decimals of an ohm. Low enough for me to
consider it to be 'dead short', anyway ...

Arfa


I suspect the usual metal migration failure mode of MLCC , then
paralling-up just increases the chance of failure , for the circuit.
Someone mentioned paralling ten , asking for trouble.

On 19/06/2015 09:04, Arfa Daily wrote:




Why would they have to be unpolarised? Just 2 paralleled together to
get the required capacitance. Any polarising mark on the SM carrier
strip perhaps


I don't think that they *do* have to be un-polarised, as such, given the
ones that Phil has found. I think it's just because they *are*, if you
see what I mean. To be honest, I never knew that you could get
multilayer ceramics in such high values. Presumably, they don't go up as
high as 47 uF which the designer felt he needed, so he just put 2 x 22
uF in parallel instead. As I said, the circuit is absolutely un-critical
in that it only needs to produce a small amount of DC from the incoming
mains to power the LED in the opto that produces the muting signal.
Because of time constraints with ordering-in those ceramics, I think I'm
just going to go with a pair of small tants, fitted to observe their
polarity.

Arfa

So to get that capacity in that package , they must be seriously multi
MLCC. Was it actually dead short or very low resistance. I suspect the
usual metal migration failure mode of MLCC , then paralling-up just
increases the chance of failure , for the circuit. Someone mentioned
paralling ten , asking for trouble.