Prev: First-time Builder: Good (v. Bad) Instructions
Next: spilt liquid on laptop.....please, some advice -my wife is going to kill me!
From: John on 7 May 2010 21:24
This is not so much problem as an annoyance, but could be a problem if and
when a problem does happen.

Computer is home built - Asus P6T SE with the latest Bios installed (0808).
The case is Lian Li PC-K58B and came with a small bud like speaker that
fits direct to the motherboard.
When booting, the one beep that happens on boot up is hardly noticeable, I
need to have my ear next to the case to even hear it. I've changed the bud
speaker to an 57mm cone 8ohm speaker in the hope that this would amplify
the sound but to no avail, still hardly noticeable.
I've checked though the Bios settings to see if the sound is adjustable but
as far as I can ascertain this isn't possible. I've also enclosed the new
cone speaker into a container to see if the volume would improve.
Is there a way I can get around this problem by amplifying the volume of
the speaker? Has anyone else come across this and if so how did you rectify
the situation?

From: VanguardLH on 7 May 2010 23:08
John wrote:

> This is not so much problem as an annoyance, but could be a problem if and
> when a problem does happen.
>
> Computer is home built - Asus P6T SE with the latest Bios installed (0808).
> The case is Lian Li PC-K58B and came with a small bud like speaker that
> fits direct to the motherboard.
> When booting, the one beep that happens on boot up is hardly noticeable, I
> need to have my ear next to the case to even hear it. I've changed the bud
> speaker to an 57mm cone 8ohm speaker in the hope that this would amplify
> the sound but to no avail, still hardly noticeable.
> I've checked though the Bios settings to see if the sound is adjustable but
> as far as I can ascertain this isn't possible. I've also enclosed the new
> cone speaker into a container to see if the volume would improve.
> Is there a way I can get around this problem by amplifying the volume of
> the speaker? Has anyone else come across this and if so how did you rectify
> the situation?

A speaker using a magnetically-driven cone won't work. You need to replace
the defective bud speaker with a piezoelectric speaker. An 8-ohm speaker
won't work. The piezo tweeter has an impedance of several hundreds of ohms
at their typical single-frequency for making beep sounds. Some examples of
piezo button speakers are listed at:

http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=720966&k=piezo%20speaker

They carry a lot of electronic parts so you might find a replacement
provided you can some info from the defective part. There are some that
connect to the 4-pin mobo header to which you connect the speaker (if the
mobo doesn't itself have the piezo buzzer), like:

http://www.pcpartscollection.com/mosp.html

If you don't have the 4-pin header (i.e., the buzzer was soldered to the
mobo), you could just clip off the connector, strip the wire ends, solder
the wires to the mobo pads, and use hot-melt glue to hold the speaker bud in
place on the mobo. Or you might find one that solders to the mobo or you
wire it, like:

http://www.radioshack.com/search/index.jsp?kwCatId=&kw=piezo&origkw=piezo&sr=1
From: John on 8 May 2010 03:12
On Fri, 7 May 2010 22:08:43 -0500, VanguardLH wrote:

> John wrote:
>
>> This is not so much problem as an annoyance, but could be a problem if and
>> when a problem does happen.
>>
>> Computer is home built - Asus P6T SE with the latest Bios installed (0808).
>> The case is Lian Li PC-K58B and came with a small bud like speaker that
>> fits direct to the motherboard.
>> When booting, the one beep that happens on boot up is hardly noticeable, I
>> need to have my ear next to the case to even hear it. I've changed the bud
>> speaker to an 57mm cone 8ohm speaker in the hope that this would amplify
>> the sound but to no avail, still hardly noticeable.
>> I've checked though the Bios settings to see if the sound is adjustable but
>> as far as I can ascertain this isn't possible. I've also enclosed the new
>> cone speaker into a container to see if the volume would improve.
>> Is there a way I can get around this problem by amplifying the volume of
>> the speaker? Has anyone else come across this and if so how did you rectify
>> the situation?
>
> A speaker using a magnetically-driven cone won't work. You need to replace
> the defective bud speaker with a piezoelectric speaker. An 8-ohm speaker
> won't work. The piezo tweeter has an impedance of several hundreds of ohms
> at their typical single-frequency for making beep sounds. Some examples of
> piezo button speakers are listed at:
>
> http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=720966&k=piezo%20speaker
>
> They carry a lot of electronic parts so you might find a replacement
> provided you can some info from the defective part. There are some that
> connect to the 4-pin mobo header to which you connect the speaker (if the
> mobo doesn't itself have the piezo buzzer), like:
>
> http://www.pcpartscollection.com/mosp.html
>
> If you don't have the 4-pin header (i.e., the buzzer was soldered to the
> mobo), you could just clip off the connector, strip the wire ends, solder
> the wires to the mobo pads, and use hot-melt glue to hold the speaker bud in
> place on the mobo. Or you might find one that solders to the mobo or you
> wire it, like:
>
> http://www.radioshack.com/search/index.jsp?kwCatId=&kw=piezo&origkw=piezo&sr=1

It hadn't entered my mind that the speaker could be faulty but after
reading your reply it makes sense. I'll need to source a supplier in my
part of the world, of find an one that sells internationally.

Thanks for the information and explanation.
From: Paul on 8 May 2010 16:43
VanguardLH wrote:
> John wrote:
>
>> This is not so much problem as an annoyance, but could be a problem if and
>> when a problem does happen.
>>
>> Computer is home built - Asus P6T SE with the latest Bios installed (0808).
>> The case is Lian Li PC-K58B and came with a small bud like speaker that
>> fits direct to the motherboard.
>> When booting, the one beep that happens on boot up is hardly noticeable, I
>> need to have my ear next to the case to even hear it. I've changed the bud
>> speaker to an 57mm cone 8ohm speaker in the hope that this would amplify
>> the sound but to no avail, still hardly noticeable.
>> I've checked though the Bios settings to see if the sound is adjustable but
>> as far as I can ascertain this isn't possible. I've also enclosed the new
>> cone speaker into a container to see if the volume would improve.
>> Is there a way I can get around this problem by amplifying the volume of
>> the speaker? Has anyone else come across this and if so how did you rectify
>> the situation?
>
> A speaker using a magnetically-driven cone won't work.

That is not true. All of my computers here have magnetic speakers
for "PC Beep" and work just fine.

A sample drive circuit for "PC Beep", can be seen here on PDF page 32.

http://www.intel.com/design/chipsets/designex/BXDPDG10.htm

Redrawing the equivalent circuit, it looks like this.

+5V ("VCC")
|
+------>
"SPKR"
+------<
|
34 ohms
|
Logic level ____|/ 2N3904
|\ silicon switching
> transistor
|
GND

Maximum power transfer, occurs with a 34 ohm speaker. Maybe they selected
that, to match well with a 32 ohm speaker. They use a cheap pair of
resistors in parallel to handle the maximum power/heat expected
of the circuit (like if you short the speaker terminals).

The logic level signal, lacks the output drive, to drive the magnetic
speaker directly. The 2N3904 can be run as a saturating switch,
handling currents of up to 200mA. If it was operated in linear
mode, it would fry. But saturated, with low Vcesat, even the
physically small transistor they use, is sufficient. You have
to check that the 3904 has sufficient current gain, for the
application.

Even when the circuit isn't perfectly matched, such as using
an 8 ohm speaker in the circuit, if you do the math, the
power level isn't significantly reduced. The speaker is
still getting 0.11 watts or so into 8 ohms.

A piezoelectric can also be used. I have no idea exactly what
the equivalent circuit looks like for one. My experience is,
they like lots of driving voltage, at less current than
a magnetic speaker would use. I've driven my piezo
headphones to high levels, using an opamp circuit.

With regard to testing that circuit for correct operation,
I don't see it being particularly easy for an end-user to
test the '3904 transistor is working properly. If there were
utilities that could force a logic 1 or a logic 0 onto
the base of the '3904, it might be easier to characterize and
verify circuit operation. I can't even be sure, looking at
a motherboard, that I can find those components. They
might be near the PANEL header, or they might not. In
a picture of the P6T, I can see a number of transistors near
the PANEL header. And no suspicious large resistors.

To "amplify" that output, you would need an amplifier. And
the amplifier would likely "click or pop" each time the computer
was turned on. I would use this as a means to get an audio signal.
The ceramic cap is present, for DC isolation. No amplifier
necessarily appreciates DC coming at it. The choice of the
8 ohm resistor (or the nearest standard value close to that),
gives a peak to peak amplitude of about 1 volt, which should be
sufficient to drive a line-level set of amplified computer speakers.
A 1/8" jack could be placed, where it says "to audio amp". You could
connect tip and ring to the upper signal, and sleeve to the ground.
If it still isn't loud enough, change the 8 ohm resistor to 33 ohms.
The set of amplified PC speakers, would then be dedicated to the
"PC beep" function.

+5V ("VCC")
|
+------> --+
|
8 ohm resistor, 1/4 watt or 1/2 watt
|
+--- 0.1uF 25V cap ----> to audio amp
| (assume 10K input
+------< --+ +---> impedance)
| |
34 ohms GND
|
Logic level ____|/ 2N3904
|\ silicon switching
> transistor
|
GND

There are other examples of "PC Beep" drive circuits. This one
looks like an emitter follower. See PDF page 74. This would
likely blow your eardrums out. I don't think this one is
safe to short the SPKR terminals, whereas the first circuit above
is short circuit safe.

http://www.intel.com/design/chipsets/schematics/252812.htm

See PDF page 118 here, for another example. This is similar to the
first link, but with a different limiting resistor. This might not
be very loud, but changing the resistor value could fix it. This
might only be about 1/3rd as loud as the first circuit.

http://web.archive.org/web/20061118133939/http://www.winbond.com/PDF/sheet/w83627hf.pdf

Another problem, is figuring out what is driving the logic signal.
I think you can get PC beep from some SuperI/O chips, or
from the Southbridge. There might be more than one potential
source on the motherboard. Trying to trace that circuit, wouldn't
be easy. It would be easy to verify one leg of the SPKR pins
is wired to VCC (+5). But for the other leg, you might have to
ohm around for a while, to find the bit that connects to it, and
then guess at the connected circuit topology.

I checked the vip.asus.com forums, and I don't see a lot of complaints
about problems with PC beep on P6T SE. Some people report they're getting a
beep code, but at the same time, they're not complaining about the
volume level of the beep. It could be, the OPs board has a fault,
like a bad '3904.

HTH,
Paul
From: VanguardLH on 8 May 2010 22:50
Paul wrote:

> VanguardLH wrote:
>> John wrote:
>>
>>> This is not so much problem as an annoyance, but could be a problem if and
>>> when a problem does happen.
>>>
>>> Computer is home built - Asus P6T SE with the latest Bios installed (0808).
>>> The case is Lian Li PC-K58B and came with a small bud like speaker that
>>> fits direct to the motherboard.
>>> When booting, the one beep that happens on boot up is hardly noticeable, I
>>> need to have my ear next to the case to even hear it. I've changed the bud
>>> speaker to an 57mm cone 8ohm speaker in the hope that this would amplify
>>> the sound but to no avail, still hardly noticeable.
>>> I've checked though the Bios settings to see if the sound is adjustable but
>>> as far as I can ascertain this isn't possible. I've also enclosed the new
>>> cone speaker into a container to see if the volume would improve.
>>> Is there a way I can get around this problem by amplifying the volume of
>>> the speaker? Has anyone else come across this and if so how did you rectify
>>> the situation?
>>
>> A speaker using a magnetically-driven cone won't work.
>
> That is not true. All of my computers here have magnetic speakers
> for "PC Beep" and work just fine.
>
> A sample drive circuit for "PC Beep", can be seen here on PDF page 32.
>
> http://www.intel.com/design/chipsets/designex/BXDPDG10.htm
>
> Redrawing the equivalent circuit, it looks like this.
>
> +5V ("VCC")
> |
> +------>
> "SPKR"
> +------<
> |
> 34 ohms
> |
> Logic level ____|/ 2N3904
> |\ silicon switching
> > transistor
> |
> GND
>
> Maximum power transfer, occurs with a 34 ohm speaker. Maybe they selected
> that, to match well with a 32 ohm speaker. They use a cheap pair of
> resistors in parallel to handle the maximum power/heat expected
> of the circuit (like if you short the speaker terminals).
>
> The logic level signal, lacks the output drive, to drive the magnetic
> speaker directly. The 2N3904 can be run as a saturating switch,
> handling currents of up to 200mA. If it was operated in linear
> mode, it would fry. But saturated, with low Vcesat, even the
> physically small transistor they use, is sufficient. You have
> to check that the 3904 has sufficient current gain, for the
> application.
>
> Even when the circuit isn't perfectly matched, such as using
> an 8 ohm speaker in the circuit, if you do the math, the
> power level isn't significantly reduced. The speaker is
> still getting 0.11 watts or so into 8 ohms.
>
> A piezoelectric can also be used. I have no idea exactly what
> the equivalent circuit looks like for one. My experience is,
> they like lots of driving voltage, at less current than
> a magnetic speaker would use. I've driven my piezo
> headphones to high levels, using an opamp circuit.
>
> With regard to testing that circuit for correct operation,
> I don't see it being particularly easy for an end-user to
> test the '3904 transistor is working properly. If there were
> utilities that could force a logic 1 or a logic 0 onto
> the base of the '3904, it might be easier to characterize and
> verify circuit operation. I can't even be sure, looking at
> a motherboard, that I can find those components. They
> might be near the PANEL header, or they might not. In
> a picture of the P6T, I can see a number of transistors near
> the PANEL header. And no suspicious large resistors.
>
> To "amplify" that output, you would need an amplifier. And
> the amplifier would likely "click or pop" each time the computer
> was turned on. I would use this as a means to get an audio signal.
> The ceramic cap is present, for DC isolation. No amplifier
> necessarily appreciates DC coming at it. The choice of the
> 8 ohm resistor (or the nearest standard value close to that),
> gives a peak to peak amplitude of about 1 volt, which should be
> sufficient to drive a line-level set of amplified computer speakers.
> A 1/8" jack could be placed, where it says "to audio amp". You could
> connect tip and ring to the upper signal, and sleeve to the ground.
> If it still isn't loud enough, change the 8 ohm resistor to 33 ohms.
> The set of amplified PC speakers, would then be dedicated to the
> "PC beep" function.
>
> +5V ("VCC")
> |
> +------> --+
> |
> 8 ohm resistor, 1/4 watt or 1/2 watt
> |
> +--- 0.1uF 25V cap ----> to audio amp
> | (assume 10K input
> +------< --+ +---> impedance)
> | |
> 34 ohms GND
> |
> Logic level ____|/ 2N3904
> |\ silicon switching
> > transistor
> |
> GND
>
> There are other examples of "PC Beep" drive circuits. This one
> looks like an emitter follower. See PDF page 74. This would
> likely blow your eardrums out. I don't think this one is
> safe to short the SPKR terminals, whereas the first circuit above
> is short circuit safe.
>
> http://www.intel.com/design/chipsets/schematics/252812.htm
>
> See PDF page 118 here, for another example. This is similar to the
> first link, but with a different limiting resistor. This might not
> be very loud, but changing the resistor value could fix it. This
> might only be about 1/3rd as loud as the first circuit.
>
> http://web.archive.org/web/20061118133939/http://www.winbond.com/PDF/sheet/w83627hf.pdf
>
> Another problem, is figuring out what is driving the logic signal.
> I think you can get PC beep from some SuperI/O chips, or
> from the Southbridge. There might be more than one potential
> source on the motherboard. Trying to trace that circuit, wouldn't
> be easy. It would be easy to verify one leg of the SPKR pins
> is wired to VCC (+5). But for the other leg, you might have to
> ohm around for a while, to find the bit that connects to it, and
> then guess at the connected circuit topology.
>
> I checked the vip.asus.com forums, and I don't see a lot of complaints
> about problems with PC beep on P6T SE. Some people report they're getting a
> beep code, but at the same time, they're not complaining about the
> volume level of the beep. It could be, the OPs board has a fault,
> like a bad '3904.
>
> HTH,
> Paul

My guess is that the problem was with a soldered-on button piezo speaker,
not with one connected to a 4-pin mobo header (which will usually take a
magnetically-driven speaker). With an integrated "buzzer", the driving
circuit probably won't handle the current load required for a magnetic
speaker. That's why the OP needs to replace the piezo speaker with another
one.
 |  Next  |  Last
Pages: 1 2 3 4
Prev: First-time Builder: Good (v. Bad) Instructions
Next: spilt liquid on laptop.....please, some advice -my wife is going to kill me!