On 14/01/2011 05:25 PM, Darren New wrote:
> scott wrote:
>> The audio DAC is probably one of the simplest parts in the phone for
>> the designers to do.
>
> You probably have separate DACs for the MP3 player and the rest of the
> phone anyway.

You reckon? I would have thought space and power consumption 
considerations would preclude that.

Post a reply to this message

>> You probably have separate DACs for the MP3 player and the rest of the
>> phone anyway.
>
> You reckon? I would have thought space and power consumption
> considerations would preclude that.

Well for power consumption the DAC for the the phone conversations can 
be much lower quality.  It wouldn't surprise me if there was a clever 
DAC that could be switched between low-quality-low-power and 
high-quality-high-power mode.  There is all sorts of stuff like that in 
phones.

Post a reply to this message

> I bought the
> content, and I'm not breaking the DRM to profit from it - but to preserve
> my access to the content in the event that I cannot get it back again.

Presumably they factored in to the original price that a certain % of 
customers will re-buy the same material due to losing access, changing 
hardware or whatever.  They also probably put something in the license 
describing that too.  If everyone removed the DRM they would probably 
need to sell it for a higher price.

Post a reply to this message

On 17/01/2011 11:30 AM, scott wrote:
>>> You probably have separate DACs for the MP3 player and the rest of the
>>> phone anyway.
>>
>> You reckon? I would have thought space and power consumption
>> considerations would preclude that.
>
> Well for power consumption the DAC for the the phone conversations can
> be much lower quality. It wouldn't surprise me if there was a clever DAC
> that could be switched between low-quality-low-power and
> high-quality-high-power mode. There is all sorts of stuff like that in
> phones.

I wouldn't have thought there's any relationship at all between quality 
and power consumption. Not for a DAC, anyway. Either way it's a pretty 
low-power device. Presumably the transmitter and backlight use vastly 
more power...

Post a reply to this message

> I wouldn't have thought there's any relationship at all between quality
> and power consumption. Not for a DAC, anyway.

In general more power means you have more possibilities to do stuff to 
increase the quality.  If you've designed what you think is a great DAC, 
and someone tells you to reduce the power by 25%, chances are the 
quality will suffer as a result.  All sorts of things used to increase 
quality (eg digital filters, analogue filters, voltage regulators) draw 
power, and usually can be removed/replaced to reduce the power.

 > Either way it's a pretty
> low-power device.

Only if it's designed that way, according to Wikipedia a typical DAC for 
PC soundcards is the CS4382, which draws almost half a Watt during 
normal operation - not really suitable for a portable device!  Believe 
me, the phone makers will be pushing very hard to get the power for all 
components as low as possible, every little helps!

 > Presumably the transmitter and backlight use vastly
> more power...

Backlights usually use around 100 - 200 mW depending on the screen size, 
but I don't think they usually stay on (at least not at full power) 
whilst music is playing, do they?  The transmitter should be on almost 
zero power whilst playing music, and judging by the talk times and 
battery capacities, looks around half a Watt whilst in a call.

Post a reply to this message

On 17/01/2011 01:30 PM, scott wrote:
>> I wouldn't have thought there's any relationship at all between quality
>> and power consumption. Not for a DAC, anyway.
>
> In general more power means you have more possibilities to do stuff to
> increase the quality. If you've designed what you think is a great DAC,
> and someone tells you to reduce the power by 25%, chances are the
> quality will suffer as a result. All sorts of things used to increase
> quality (eg digital filters, analogue filters, voltage regulators) draw
> power, and usually can be removed/replaced to reduce the power.

I still rather suspect that rather than put in a high-quality DAC and a 
low-quality DAC, they'll just not bother with the high-quality one at all.

>> Either way it's a pretty low-power device.
>
> Only if it's designed that way, according to Wikipedia a typical DAC for
> PC soundcards is the CS4382, which draws almost half a Watt during
> normal operation - not really suitable for a portable device!

OK. But also not *designed* for a portable device either. I'm sure 
somebody has come up with much more frugal designs.

>> Presumably the transmitter and backlight use vastly more power...
>
> Backlights usually use around 100 - 200 mW depending on the screen size,
> but I don't think they usually stay on (at least not at full power)
> whilst music is playing, do they?

I would hope not.

> The transmitter should be on almost zero power whilst playing music

I don't know that much about GSM, but doesn't it have to periodically 
broadcast to keep in touch with the network?

> and judging by the talk times and
> battery capacities, looks around half a Watt whilst in a call.

Mmm, maybe.

Post a reply to this message

> I still rather suspect that rather than put in a high-quality DAC and a
> low-quality DAC, they'll just not bother with the high-quality one at all.

And then your MP3s sound like phone conversations .... ummm I doubt it. 
  IIRC phone conversations are up to 3.4 kHz or something, whilst MP3s 
are usually up to 20 kHz or so.  As a very simple solution you could 
just run the same DAC at a slower clock rate for the voice calls to save 
power, but I bet there are far more sophisticated solutions actually 
used in the handsets.

>> The transmitter should be on almost zero power whilst playing music
>
> I don't know that much about GSM, but doesn't it have to periodically
> broadcast to keep in touch with the network?

Hence "almost" zero, and not actually zero.

Post a reply to this message

On 17/01/2011 03:03 PM, scott wrote:
>> I still rather suspect that rather than put in a high-quality DAC and a
>> low-quality DAC, they'll just not bother with the high-quality one at
>> all.
>
> And then your MP3s sound like phone conversations .... ummm I doubt it.
> IIRC phone conversations are up to 3.4 kHz or something, whilst MP3s are
> usually up to 20 kHz or so. As a very simple solution you could just run
> the same DAC at a slower clock rate for the voice calls to save power,
> but I bet there are far more sophisticated solutions actually used in
> the handsets.

Is power consumption directly related to clock speed in some way?

>>> The transmitter should be on almost zero power whilst playing music
>>
>> I don't know that much about GSM, but doesn't it have to periodically
>> broadcast to keep in touch with the network?
>
> Hence "almost" zero, and not actually zero.

Oh, OK.

Post a reply to this message

> Is power consumption directly related to clock speed in some way?

Pretty much always.  On the digital side of things transistors use most 
power whilst switching (as opposed to being in steady state on or off), 
so the more switching they do per second the higher the average power 
consumption.  On the analogue side the details probably vary depending 
on exactly which DAC architecture you're using, but in general I can't 
imagine power going down at higher clock speeds.  Higher clock speeds 
mean faster rate of change of voltage, which usually means higher 
currents are involved, so more power gets dissipated in various components.

Post a reply to this message

On 17/01/2011 03:33 PM, scott wrote:
>> Is power consumption directly related to clock speed in some way?
>
> Pretty much always.

Mmm, OK.

> On the analogue side the details probably vary depending on
> exactly which DAC architecture you're using, but in general I can't
> imagine power going down at higher clock speeds.

No, but it could stay roughly the same.

I suppose the next question is how much power changes with frequency, 
but I guess that's going to depend intimately on exactly what you're 
doing...

I still rather suspect that the difference between a 30,000 sps DAC and 
a 20,000 sps DAC is rather minute, but whatever.

Post a reply to this message