PDA

View Full Version : SB2: reduce analog out voltage by 10x?



usualsuspects
2005-09-07, 07:42
I would like to run SB2 analog out direct to my amp (Parasound HCA-750) - my amp manual says "0.775v input for full output". At 6v on the analog outs - I am afraid it will be way too high. Is there a simple analog resistor network I can build that will reduce the voltage by a factor of 10 to 0.6 volts? The input impedance on my amp is 33K ohms. I tend to listen at low volume and having the most range in the volume control is desired. My speakers are fairly efficient – 90db. Thanks!

cbemoore
2005-09-07, 08:04
I've copied my post from another thread:

You need to vote for this bug:

http://bugs.slimdevices.com/show_bug.cgi?id=1601

I REALLY want this functionality, but so far it looks like I'm the only person who's actually voted for it!

Running directly into a power amp (and using the digital volume control) gives a far cleaner signal path than connecting via a pre-amp. But the signal out of the SB2 is far too hot, and its impossible to turn the digital volume down low enough - 1 out of 40 is still really quite loud!

The DAC used in the SB2 already has a built-in attenuation capability, but its not used. If the Slim guys could add an option to make use of it, it would solve all my problems....

Chris

mazurek
2005-09-07, 08:05
Parts express sells line level attenuators, look at them if you are in the US

Davey
2005-09-07, 08:12
I'm not sure where you're getting the 6 volt figure from. The maximum output from the SB2 is approximately 1.85 volts RMS. I would try using a simple volume control between the SB2 and your amp. That should work fine.

Cheers,

Davey.

Robin Bowes
2005-09-07, 08:22
usualsuspects wrote:
> I would like to run SB2 analog out direct to my amp (Parasound HCA-750)
> - my amp manual says "0.775v input for full output". At 6v on the
> analog outs - I am afraid it will be way too high. Is there a simple
> analog resistor network I can build that will reduce the voltage by a
> factor of 10 to 0.6 volts? The input impedance on my amp is 33K ohms. I
> tend to listen at low volume and having the most range in the volume
> control is desired. My speakers are fairly efficient – 90db. Thanks!

Hi,

You don't actually need to reduce the output by that much.

The 0.775v figure is almost certainly a "standard" 0dBu level.

dBu measurement is referenced to 0.7746v (often rounded to 0.775v), i.e.:

dBu = 20log10(v/0.7746v)

where v is the rms voltage, 0.775v in this case (rounded from 0.7746v).

The 6v figure for the Squeezebox is a peak-to-peak figure.

Converting this to rms gives:

Vrms = (Vpp / 2) / sqrt(2)
= (6 / 2) / 1.414
= 2.12v

This corresponds to:

dBu = 20log10(2.12 / 0.7746)
= 8.75 dBu

So, to get an rms level of 0dBu you would need to reduce the output
voltage by 2.12 - 0.7746 = 1.35v rms

In actual fact, it is highly unlikely that your amplifier can't cope
with this signal level. So I would just plug it in and see how you go.

R.
--
http://robinbowes.com

If a man speaks in a forest,
and his wife's not there,
is he still wrong?

usualsuspects
2005-09-07, 20:52
Just wanted to say thanks to everyone for responding. I connected my SB2 directly to the Parasound amp, adjusted the gain pots on the amp a little, and it works great!

PS - the DAC in the SB2 is not bad at all. Not Classe Dac1 range (what I am used to) but not bad.

Jacob Potter
2005-09-08, 12:57
On 9/7/05, occam <occam.1v0utz (AT) no-mx (DOT) forums.slimdevices.com> wrote:
> A man walks into a doctor's office, and says to the doctor-
> "Doctor, when I do this with my arm, it really hurts."
> and proceeds to show his arm all the way up his....
> The Doctor looks at the contorted patient, ponders a moment, and says -
> "Well, have you considered not shoving up ALL the way?"
>
> Have you considered not shoving the volume up all the way?

Nice analogy :)

The problem is, there's nowhere near enough control with the low
volume ranges. On my Grado SR-80s driven directly from the headphone
jack, with some music, a volume level of 3 is a bit on the loud side
but 2 is too quiet. Sometimes I need to crank it up to 4 or 5 though.
If I could scale everything down (up?) by a factor of 5, it would be
much easier to find a happy medium.

- Jacob

occam
2005-09-08, 14:15
Dang Jacob..... I was editing that post and I see you copied the edited, more 'tasteful' version of the anecdote. I saw Usual's reply, while editing, mentioning that the Parasound had attenuators on its inputs which seemed to address the problem quite well. So I deleted my post. But you're right, inevitably, you loose resolution as you've only got the 40 set volume levels that are done in software. It would be nice if Sean, et al, would implement 256 levels of .5db attenuation. Barring that, if one doesn't want external attenuators, you could replace the input 4.7k resistors, R37 & R39, of the output opamps, with something close to 11.22k which would bring down the max to about .775Vrms.
http://www.seanadams.com/dac.pdf
This would still give those 39-40 steps up to the max desired output, which begs the question of whether X fixed steps is ever enough.
But I'm old and uncoordinated, and replacing smt resistors is a daunting task for me.

cbemoore
2005-09-08, 15:24
I still think Sean's solution (option 2 in bug 1601) is the best solution.

When I hook my SB2 direct to my power amp, volume 1/40 is fairly loud. Volume 2/40 is very loud. Volume 3/40 is too loud! And there's no way I can make it quiet enough for background music.

The best solution would be to add a a per-player option to attenuate the signal. That way, its up to the user whether they want to select the option or not.

And since the DAC has an inbuilt attenuation capability (which can be controlled by the firmware), I'm guessing this would be the obvious place to perform the attenuation.....

So if you're having problems with the analog output being too loud, make sure you vote for bug 1640!!

Chris

JJZolx
2005-09-08, 15:30
Qustion: _Why_ is the SB2 output level so high?

seanadams
2005-09-08, 16:06
Qustion: _Why_ is the SB2 output level so high?

Squeezebox1 was 3vpp and people complained it was too low.

Now it's 6vpp (~2v rms) which is in line with nearly all high-end equipment. Higher voltage == better SNR along the whole path. This isn't an "oops we made it too loud" - SB2 was deliberately designed (at significant added component cost) to have 2vrms levels.

This $20,000 cd player, for example, is 2vrms:

http://www.stereotimes.com/trans120699.shtm

I don't get what the problem is. If the output were hot enough to clip anyone's receiver that would be an issue, but it's not - the only thing this would potentially clip is really cheap amplified speakers - even a $100 receiver will handle 2vrms. If you're going directly into a standalone amp... well, the only reason you cat do that at all is because we have a volume control. Of course at full volume you're going to drive it to the max - the same would happen with a CD player.

The ideal solution is to

a) add a preamp
b) open up the amp and tune its input pots (asuuming it has them)
c) add in-line attenuators as suggested above.

The internal volume control is not ideal for this, but sure we can tweak the software to adjust the range and make it work.

pfarrell
2005-09-08, 16:18
On Thu, 2005-09-08 at 16:06 -0700, seanadams wrote:
> > Qustion: _Why_ is the SB2 output level so high?
>
> Squeezebox1 was 3vpp and people complained it was too low.

You can never make everyone happy.
I'm happy.

> I don't get what the problem is.

Me neither.
A voltage divider circuit has three resistors.
You'd need two, so it would be six resistors. Or
two pots. Clutch coupled, concentric, audio taper
pots aren't that big a deal for the hand full of folks
that have a problem. Less than $10 worth of parts.

--
Pat
http://www.pfarrell.com/music/slimserver/slimsoftware.html

cbemoore
2005-09-08, 16:28
If you're going directly into a standalone DAC... well, the only reason you cat do that at all is because we have a volume control.

Since we're referring to the analog outputs, I assume you mean a standalone power amp, not a DAC?

As you say, the only reason you can do this at all is because of the internal volume control. But this is exactly why I want to do it!

If I connect the SB2 to a preamp and on to my power amp, the *only* thing the preamp does is control the volume. Since volume is already controlled in the SB2, it makes sense to remove the preamp completely, and connect the SB2 direct to the power amp. This makes the signal path much cleaner, and to my ears at least, it sounds much better. It also cuts down on the number of boxes, cables and power sockets in my system, which is always nice....

Chris

Robin Bowes
2005-09-08, 16:36
seanadams said the following on 09/09/2005 00:06:
>>Qustion: _Why_ is the SB2 output level so high?
>
> Squeezebox1 was 3vpp and people complained it was too low.

Are you sure? 3vpp is 1.06v rms and I'm pretty sure my SB1 has an output
that is less that 0dBu (which would be 0.7746v rms).

> Now it's 6vpp (~2v rms) which is in line with nearly all high-end
> equipment. Higher voltage == better SNR along the whole path. This
> isn't an "oops we made it too loud" - SB2 was deliberately designed (at
> significant added component cost) to have 2vrms levels.
>
> This $20,000 cd player, for example, is 2vrms:
>
> http://www.stereotimes.com/trans120699.shtm
>
> I don't get what the problem is. If the output were hot enough to clip
> anyone's receiver that would be an issue, but it's not - the only thing
> this would potentially clip is really cheap amplified speakers - even a
> $100 receiver will handle 2vrms. If you're going directly into a
> standalone DAC... well, the only reason you cat do that at all is
> because we have a volume control. Of course at full volume you're going
> to drive it to the max - the same would happen with a CD player.

The problem is that most domestic equipment is designed to operate at
0dBu, or 0.7746v rms for max level. Sure, the preamp can often handle
more, but the gain stages are setup so the volume control operates
sensibly with this level of input. When the input is "hotter" the useful
range of the volume knob is reduced.

> The ideal solution is to
>
> a) add a preamp

....designed to operate with 2.12v input

> b) open up the amp and tune its input pots (asuuming it has them)
> c) add in-line attenuators as suggested above.

....with attenuation of approx. 2.58 dBu

d) open up the Squeezebox and replace R37 & R39 with 12.7k or 12.9k
resistors.

Here's my calculations:

Standard Configuration
R37 4700
R38 8250
Gain 1.755319149
Standard Voltages
Vout(p-p) 6
Vout(p) 3
Vout(rms) 2.121320344
Vin(rms) 1.208509771
Desired Voltages
Vout(rms) 0.7746
Gain 0.640954685
Required R37 12871.42475
Attenuation
For 0dBu -2.585548412


> The internal volume control is not ideal for this, but sure we can
> tweak the software to adjust the range and make it work.

What effect would this have on audio quality?

R.
--
http://robinbowes.com

If a man speaks in a forest,
and his wife's not there,
is he still wrong?

seanadams
2005-09-08, 16:38
Since we're referring to the analog outputs, I assume you mean a standalone power amp, not a DAC?
oops - edited



If I connect the SB2 to a preamp and on to my power amp, the *only* thing the preamp does is control the volume. Since volume is already controlled in the SB2, it makes sense to remove the preamp completely, and connect the SB2 direct to the power amp. This makes the signal path much cleaner, and to my ears at least, it sounds much better.

The thing is... it actually doesn't! The SB2 has a very low noise floor down around -120db. However when you reduce the signal level, the noise floor doesn't change so your SNR becomes effectively worse. Less signal, same noise == less SNR. That's why a fixed analog attenuator set your maximum listening range is the cleanest way to go - it attenuates the signal and the noise together, meaning that the noise floor in your amplifier is the limiting factor (which it might be anyway, in which case you're right and all this is moot).



It also cuts down on the number of boxes, cables and power sockets in my system, which is always nice....


I agree. It is also easier to set up and doesn't require buying or soldering together an attenuator.

cbemoore
2005-09-08, 16:58
Thanks for the explanation, Sean. That makes it much clearer.

So do I need to go out and buy some attenuators, or is there a teensy weensy chance you might be implementing something in software???

Cheers
Chris

seanadams
2005-09-08, 17:02
seanadams said the following on 09/09/2005 00:06:
>>Qustion: _Why_ is the SB2 output level so high?
>
> Squeezebox1 was 3vpp and people complained it was too low.

Are you sure? 3vpp is 1.06v rms and I'm pretty sure my SB1 has an output
that is less that 0dBu (which would be 0.7746v rms).


I'm rounding here... but yes I'm pretty sure I measured Squeezbox1 is in the 2.9-3v range. That's taking the outputs straight off the Micronas' built-in DAC. It can actually go almost to 3.3 (Vcc) but it starts to clip at some point and that's where I set the max. Sorry I don't have the exact figures handy.

seanadams
2005-09-08, 17:03
Thanks for the explanation, Sean. That makes it much clearer.

So do I need to go out and buy some attenuators, or is there a teensy weensy chance you might be implementing something in software???

Cheers
Chris

It's on the list. The list is long. I will move it up the list. :)

seanadams
2005-09-08, 20:12
OK I added this. The setting goes from 0 to 63 db of attenuation in 0.5db steps. Firmware will be updated soon - need to sync up with Vidur on that first.

usualsuspects
2005-09-08, 20:58
Great! In further reading of my amp manual it says (re the gain pots) “Your amplifier sounds best with these level controls set to maximum…” So – it looks like I can turn them all the way up and set attenuation on the SB2 to match the “correct” level for my amp. Earlier in this thread you stated: “However when you reduce the signal level, the noise floor doesn't change so your SNR becomes effectively worse…” So – from a “best sound” standpoint – would I be better off using the new built in attenuation – or using an external attenuator?

seanadams
2005-09-08, 23:50
Great! In further reading of my amp manual it says (re the gain pots) “Your amplifier sounds best with these level controls set to maximum…” So – it looks like I can turn them all the way up and set attenuation on the SB2 to match the “correct” level for my amp. Earlier in this thread you stated: “However when you reduce the signal level, the noise floor doesn't change so your SNR becomes effectively worse…” So – from a “best sound” standpoint – would I be better off using the new built in attenuation – or using an external attenuator?

Give it a try - if you don't hear any hiss/hum when the system is inactive, it should be fine. Personally I would prefer limiting the amp, for safety reasons if nothing else.

cbemoore
2005-09-09, 00:01
OK I added this. The setting goes from 0 to 63 db of attenuation in 0.5db steps. Firmware will be updated soon - need to sync up with Vidur on that first.

Wow - didn't expect it to be *that* quick! Thanks Sean!!

Robin Bowes
2005-09-09, 01:15
seanadams wrote:
> OK I added this. The setting goes from 0 to 63 db of attenuation in
> 0.5db steps. Firmware will be updated soon - need to sync up with Vidur
> on that first.

Just to add...

I reckon 2.5dB of attenutation will bring the level down to 0dBu.

R.
--
http://robinbowes.com

If a man speaks in a forest,
and his wife's not there,
is he still wrong?

Deaf Cat
2005-09-09, 11:06
Errrrm, I do feel a little daft joining in this conversation about attenutation and dB and 0dBu's as I have not got the faintest idea what you rather technical people are all talking about, but would like to learn a little if possible...

The Preamp Volume Control is default Zero, which means full volume..... approx. 2Vrms? - (normal CD o/p)

I guess attenuation has something to do with adjusting the signal level which in this case equates to the volume.

The range 0.0 to 63.0 - So the setting 0.0 being approx. 2Vrms and 63db being 0V, if I had a power amp that likes 0.8V, a setting closer to the 63.0 than the 0.0, would please the power amp...?

So, the firmware lets SB2 know what these settings mean, or does the firmware help with SNR?

If you do understand how I am thinking, and what I would like to understand in the simplest form, I would love to hear nice simplistic explanations if possible, thanks for your patience.

Robin Bowes
2005-09-09, 13:36
Deaf Cat said the following on 09/09/2005 19:06:
> Errrrm, I do feel a little daft joining in this conversation about
> attenutation and dB and 0dBu's as I have not got the faintest idea what
> you rather technical people are all talking about, but would like to
> learn a little if possible...

Good attitude!

>
> The Preamp Volume Control is default Zero, which means full volume.....
> approx. 2Vrms? - (normal CD o/p)

The std SB2 output is 6v(peak to peak) which is 3v(peak) or 2.12v(rms).

"Normal" CD output tends to be 0.775v(rms).

> I guess attenuation has something to do with adjusting the signal level
> which in this case equates to the volume.

Correct. Attenuation means reducing the level.

> The range 0.0 to 63.0 - So the setting 0.0 being approx. 2Vrms and 63db
> being 0V, if I had a power amp that likes 0.8V, a setting closer to the
> 63.0 than the 0.0, would please the power amp...?

Not quite. dB (decibels) are a logrithmic measurement, usually
referenced to some standard value. dBu are referenced to 0.7746v.

For the stock SB2, output is 2.12v(rms). In dBu, this is:

20 x log10(2.12 / 0.7746) = 8.75dBu

Say you want to reduce your output to 0dBu, you need to add 8.75dBu of
attenuation. This would reduce the max Vout(rms) to 0.7746v.

> So, the firmware lets SB2 know what these settings mean, or does the
> firmware help with SNR?

The firmware is the bit that implements this functionality by
controlling the DAC chip in the SB2.

> If you do understand how I am thinking, and what I would like to
> understand in the simplest form, I would love to hear nice simplistic
> explanations if possible, thanks for your patience.

I hope the above helps.

BTW, the firmware has just been relased. I've downloaded it and can
report that it's working well.

I'm not sure that the way of setting the attenutation is the most
intuitive of user interfaces, but it does the job.

R.
--
http://robinbowes.com

If a man speaks in a forest,
and his wife's not there,
is he still wrong?

seanadams
2005-09-09, 14:00
I'm not sure that the way of setting the attenutation is the most
intuitive of user interfaces, but it does the job.


I just put something simple in there to make it work... open to better ideas.

Dean, Vidur and I have been talking about volume control in general - there's a lot more to it than meets the eye, once you consider crossfading, replaygain, headphone use, amp use, etc. Right now the sofware volume control uses an 8-bit linear value which by itself doesn't give enough levels on the low end of the scale. With this additional "preamp" control (7 bits of half decibels, done in the DAC) it should solve the problem for those who've been stuck on the low end of the bar with not enough resolution. But yes, the UI is a bit abstruse. How about just a pop-up list with "quiet", "not so quiet", "loud", "eleven"? :)

JJZolx
2005-09-09, 14:27
I just put something simple in there to make it work... open to better ideas.

Dean, Vidur and I have been talking about volume control in general - there's a lot more to it than meets the eye, once you consider crossfading, replaygain, headphone use, amp use, etc. Right now the sofware volume control uses an 8-bit linear value which by itself doesn't give enough levels on the low end of the scale. With this additional "preamp" control (7 bits of half decibels, done in the DAC) it should solve the problem for those who've been stuck on the low end of the bar with not enough resolution. But yes, the UI is a bit abstruse. How about just a pop-up list with "quiet", "not so quiet", "loud", "eleven"? :)
If you want to make use of all 128 steps in the digital control then a more intuitive interface might be to simply enter an integer from 1 to 128 (maximum) (or perhaps 0 to 127) and explain that at the maximum setting the signal is unaltered and that each step below the max gives 0.5db of attenuation.

If the DAC offers 128 discrete steps of volume control, have you considered the possibility (maybe as an option) of reversing the roles of the software volume control and the DAC's digital volume control so that we'd have the option of controlling volume in the digital domain?

seanadams
2005-09-09, 14:44
If the DAC offers 128 discrete steps of volume control, have you considered the possibility (maybe as an option) of reversing the roles of the software volume control and the DAC's digital volume control so that we'd have the option of controlling volume in the digital domain?

Eh? They're both digital - just one's done in the CPU and one's done in the DAC.

Deaf Cat
2005-09-09, 14:55
Cheers for the explanations, All Clear, except, hope you don't mind another (maybe daft) question:

So, take 8.75dBu off of 2.12V to end up with 0.7746V (Which is what my power amp would like-roughly 0.8V-cool)

So, if we add 8.75dBu attenuation (it would be safest to round it to 9dBu I suppose) into the Option Box in the Audio settings for the SB2 we should end up with approx 0.8Vrms o/p from SB2.

But how does the 8.75dBu relate to the 0.0 - 63.0dB scale?

Oh Boy! Takes alot of brain power to almost understand that taking off is addin on and 0 is 0.7, quite intresting how logs relate though.



>I hope the above helps.

Absolutly!


>BTW, the firmware has just been relased. I've downloaded it and can
report that it's working well.

Is this in the 6.2, 2005-09-09 01:26 download?


>I'm not sure that the way of setting the attenutation is the most
intuitive of user interfaces, but it does the job.

Your telling me!! Who would have thought it a! Buy a bit of audio equipment to replace a cd player and then starting to want to know about Logarithms!!

Thanks again for the explanations, much appreciate it.


> How about just a pop-up list with "quiet", "not so quiet", "loud", "eleven"? :)

""eleven"" - FAB!

rme
2005-09-09, 16:41
Does this change for volume control degrade the quality of the stream? I saw a prior thread on analog volume reductions dropping bits.

This sounds like it is changing voltage instead so I have no idea of whether it's changing the data or is truely like a preamp volume control.

Thanks for any insight into this.

cbemoore
2005-09-09, 17:00
""eleven"" - FAB!

Have you noticed how the player volume in the web interface goes up to 11? That's not a mistake..... :-)

cbemoore
2005-09-09, 17:03
Hi Sean,

Just downloaded the latest firmware, and it works perfectly. Thanks!

(BTW, I've set the SB2 preamp control to 30dB attenuation, *and* my power amp to 10dB attenuation, and the volume is just about perfect! I must have a particularly loud amp!!!)

Chris

seanadams
2005-09-10, 01:29
Does this change for volume control degrade the quality of the stream? I saw a prior thread on analog volume reductions dropping bits.

This sounds like it is changing voltage instead so I have no idea of whether it's changing the data or is truely like a preamp volume control.

Thanks for any insight into this.

I don't know how the volume control is implemented in the DAC except that it is a digital function. Here's the data sheet:

http://www-s.ti.com/sc/ds/pcm1748.pdf

Deaf Cat
2005-09-10, 04:07
It works lovely, thanks everso for getting this extra enhancement up and running.

Started with attenuation of 30 just to be extra safe (knowing it should be 8.75 for my particular power amp-0.8V-thanks very much to Robin for explaining), then 20 and I'm on 10 at the mo and do not think I need to go to the 8.75.

Next is to get a decent interconnect to tighten up the bass a bit more and smooth the voices.

Thanks again for this enhancement.

Music Machine
2005-09-10, 09:30
"Preamp Volume Control" not working in my setup.

Have updated to nightly SlimServer_v2005-09-09.exe and installed firmware update.

SlimServer Version: 6.2b1 - 4218 - Windows 2000 - EN - cp1252
Player Firmware Version: 20

W2K, wired only SB2

Trying a few setting from 0 to 60 in "Preamp Volume Control" made no difference. I'm using SB2 analog outs.

I just got SB2 about a week ago and there is no going back to disk players. Would like to connect directly to my power amp so any suggestions that might help me use this feature are greatly appreciated.

Regards
Music Machine

Robin Bowes
2005-09-10, 09:38
Music Machine said the following on 10/09/2005 17:30:
> "Preamp Volume Control" not working in my setup.
>
> Have updated to nightly SlimServer_v2005-09-09.exe and installed
> firmware update.
>
> SlimServer Version: 6.2b1 - 4218 - Windows 2000 - EN - cp1252
> Player Firmware Version: 20
>
> W2K, wired only SB2
>
> Trying a few setting from 0 to 60 in "Preamp Volume Control" made no
> difference. I'm using SB2 analog outs.
>
> I just got SB2 about a week ago and there is no going back to disk
> players. Would like to connect directly to my power amp so any
> suggestions that might help me use this feature are greatly
> appreciated.

Preamp Volume is in firmware > 20, i.e. from v21.

I got it sooner than the 10/09 nightly because I track svn.

R.
--
http://robinbowes.com

If a man speaks in a forest,
and his wife's not there,
is he still wrong?

JJZolx
2005-09-10, 09:55
Eh? They're both digital - just one's done in the CPU and one's done in the DAC.
Eh, brain fart.


Right now the sofware volume control uses an 8-bit linear value which by itself doesn't give enough levels on the low end of the scale.
If volume control is done in the CPU couldn't it have been made nonlinear to address the problem at the lower volumes? Or would that be difficult/confusing for users?

Music Machine
2005-09-10, 10:05
Thank you Robin Bowes. Latest nightly working as expected.

dean
2005-09-10, 10:27
On Sep 10, 2005, at 9:55 AM, JJZolx wrote:
> seanadams Wrote:
>
>> Right now the sofware volume control uses an 8-bit linear value
>> which by
>> itself doesn't give enough levels on the low end of the scale.
>>
> If volume control is done in the CPU couldn't it have been made
> nonlinear to address the problem at the lower volumes?
Yes, we'll be fixing this.

> Or would that
> be difficult/confusing for users?
Nope, linear is the one that doesn't make any sense.

Robin Bowes
2005-09-10, 12:11
Deaf Cat said the following on 09/09/2005 22:55:
> Cheers for the explanations, All Clear, except, hope you don't mind
> another (maybe daft) question:
>
> So, take 8.75dBu off of 2.12V to end up with 0.7746V (Which is what my
> power amp would like-roughly 0.8V-cool)

Correct.

> So, if we add 8.75dBu attenuation (it would be safest to round it to
> 9dBu I suppose) into the Option Box in the Audio settings for the SB2
> we should be fine.

8.5 would probably be OK, or even 8.

> But how does the 8.75dBu relate to the 0.0 - 63.0dB scale?

0.0 is 0dB attenuation so the output is 8.75dBu.

8.5 is 8.5dB attenuation so the output is 0.25dBu.

63 is 63dB attenuation so the output is -54.25dBu

> Oh Boy! Takes alot of brain power to almost understand that taking off
> is addin on and 0 is 0.7, quite intresting how logs relate though.

Well, I expect the UI to change - this is something Sean has knocked out
quick and there is almost certainly a better way to specify the
attenuation. I'm sure he'll come up with something better.

Any suggestions?

>>BTW, the firmware has just been relased. I've downloaded it and can
>
> report that it's working well.
>
> Is this in the 6.2, 2005-09-09 download?

No, but it's probably in 2005-09-10.

if you're downloading regularly I recommend using svn to get the latest
updates.

>>I'm not sure that the way of setting the attenutation is the most
> intuitive of user interfaces, but it does the job.
>
> Your telling me!! Who would have thought it a! Buy a bit of audio
> equipment and then starting to want to know about Logarithms!!
>
> Thanks again for the explanations.

Unfortunately, your ears are logarithmic. It's just you didn't know
about it before now!

>
>>How about just a pop-up list with "quiet", "not so quiet", "loud",
>
> "eleven"? :)
>
> ""eleven"" - FAB.....

;)

R.
--
http://robinbowes.com

If a man speaks in a forest,
and his wife's not there,
is he still wrong?

Deaf Cat
2005-09-10, 15:31
Cheers Robin for the simple explanations for a simple brain.

Getting the hang of this attenuation business now...!?
Actually changed attenuation to 12, as vol no.20 was rather loud and as the bass is a tad flabby at the mo did not want to push it any more - will wait till I get a better analogue cable.


Well, I expect the UI to change - this is something Sean has knocked out
quick and there is almost certainly a better way to specify the
attenuation. I'm sure he'll come up with something better.
Any suggestions?
What about being able to type in the max Voltage you want on the output (Range from 0Vrms to 2.15Vrms) so I would simply put in my 0.8V that my power amp likes-(in its spec), and the software does all the dBu log calcs? Would this work?



But it's probably in 2005-09-10.
Cool, got it, cheers.

if you're downloading regularly I recommend using svn to get the latest
updates.
This svn sounds rather good, where would I find this?

Ears work on logs...oh! so nature understands it - that means my brain understands them too.....hmmmmmmm

Cheers again for your help,
Nice to have a rough idea now, what is going on!

cbemoore
2005-09-14, 08:06
Still on the subject of dBu.....

My power amp has a power-save function. If the input signal is less that -74dBu, the power amp will automatically switch off after 5 mins. If it detects a signal louder than -74dBu, it will automatically turn back on.

My problem is that when I connect the SB2 directly to the power amp, the power amp never turns off. This would seem to indicate that even when the SB2 is turned off, its noise floor is louder than -74dBu.

Is this to be expected, or do I have a duff unit?

Cheers
Chris

Robin Bowes
2005-09-14, 08:23
cbemoore wrote:
> Still on the subject of dBu.....
>
> My power amp has a power-save function. If the input signal is less
> that -74dBu, the power amp will automatically switch off after 5 mins.
> If it detects a signal louder than -74dBu, it will automatically turn
> back on.
>
> My problem is that when I connect the SB2 directly to the power amp,
> the power amp never turns off. This would seem to indicate that even
> when the SB2 is turned off, its noise floor is louder than -74dBu.
>
> Is this to be expected, or do I have a duff unit?

Sean can probably answer this better but 74dBu is approx. 150uV rms,
i.e. not much so it's possible that there may be some residual noise
that is keeping your amp switched on.

R.
--
http://robinbowes.com

If a man speaks in a forest,
and his wife's not there,
is he still wrong?

seanadams
2005-09-14, 10:17
That is a really low threshold. I'd have to measure what the noise floor is in absolute terms - I've only looked at it in the frequency domain which isn't going to indicate an exact pk-pk figure needed to determine if it's always below 150uV.

fuzzyT
2005-09-14, 11:10
seanadams wrote:
> That is a really low threshold.

Is it possible that the new attenuation controls could lower the noise
floor level enough to get around this?

If so, there's at least a workaround. A bit clunky for manual
operation, but perhaps could be coded to raise and lower on power on/off.

--rt

seanadams
2005-09-14, 14:36
seanadams wrote:
> That is a really low threshold.

Is it possible that the new attenuation controls could lower the noise floor level enough to get around this?

Well this is fascinating - the short answer is yes we can probably make it work.

Our noise floor in the audio band is down at -127 dBu. However delta-sigma DACs inherently generate impulses (high frequency content) of some couple hundred microvolts even when fed a DC level - i.e. you trade much higher performance in the audio band for some out-of-band noise. Even for all-zeroes the DAC is actively working to hold the output between two levels.

So what is tripping your amp is probably this component of the noise floor. Looking at the rms level of noise floor it is about -60dBu when the DAC is active. I tried setting the attentuation to -64dB (mute) but this apparently does not actually disable the DAC, it's the same as playing zeroes. Next I tried inhibiting the clocks completely, which worked - it goes down to -90dBu, which is the same as the analyzer measures with RCA cables not connected to anything. Next I tried disabling the DAC by setting its reset bit. This worked too. So with some smarts in the SB2 to shut off the DAC when appropriate we should be able to do this.

http://www.seanadams.com/sb2_noise_floor.gif
http://www.seanadams.com/dac_disabled.gif

seanadams
2005-09-14, 15:07
Hmmm looks like using 128x (instead of 64x) oversampling brings the out-of-band noise way down, mostly under -115dBu all the way up to 96KHz.

fuzzyT
2005-09-14, 15:30
seanadams wrote:
> Hmmm looks like using 128x (instead of 64x) oversampling brings the
> out-of-band noise way down, mostly under -115dBu all the way up to
> 96KHz.

wonder what are the pros/cons of using 128x all the time. the noise is
low level and out of band, but could be interacting with signal, no? is
the level still that much lower when processing signal?

--rt

seanadams
2005-09-14, 15:58
seanadams wrote:
> Hmmm looks like using 128x (instead of 64x) oversampling brings the
> out-of-band noise way down, mostly under -115dBu all the way up to
> 96KHz.

wonder what are the pros/cons of using 128x all the time. the noise is
low level and out of band, but could be interacting with signal, no? is
the level still that much lower when processing signal?

--rt

well it does have a marginally higher noise level in the audio band. here they are overlaid:

http://www.seanadams.com/64_128.gif

These jive with the data sheet (page 28) - note that their scale goes all the way to 8 times fs and mine only goes to 2x.

fuzzyT
2005-09-14, 16:19
seanadams wrote:

> well it does have a marginally higher noise level in the audio band.
> here they are overlaid:

promising. for the few db in the audio band and for the potential
harmonic interactions of the slightly higher band stuff.

hard to say what the ears will make of it, but we'd sure love to find out.

--rt

rme
2005-09-15, 17:40
Any suggestions as I have very little idea how to optimally set this new function.

I have nuforce amps and the specs say: max input voltage=1.2V, Gain=26db, Input Impedance: 47K ohm, SNR = 120 db at 100W.

Any help would be appreciated.

Right now without applying the updated firmware and slimserver update, volume 1 is low, 2 is normal, 3 is loud and everything from 4-40 is way too loud.

seanadams
2005-09-15, 21:27
Any suggestions as I have very little idea how to optimally set this new function.

I have nuforce amps and the specs say: max input voltage=1.2V, Gain=26db, Input Impedance: 47K ohm, SNR = 120 db at 100W.

Any help would be appreciated.

Right now without applying the updated firmware and slimserver update, volume 1 is low, 2 is normal, 3 is loud and everything from 4-40 is way too loud.

The optimum depends on a) your listening preference b) amp gain and c) speaker sensitivity. i.e. experiment and then pick the value that gives you the preferred listenable range within the main volume control.

The latest nightly releases have this feature:

http://www.slimdevices.com/downloads/nightly/latest/6.2/

Just grab the right one (exe for Windows, dmg for Mac, etc) and give it a try.

Also note Vidur has recently made some improvements to the main volume control (higher resolution and replaygain support) and we are working to address issues with the volume curve, so things may change a little going forward.

cbemoore
2005-09-15, 22:24
Well the latest volume control settings have made a BIG improvement in low-volume resolution.

I've now set the SB2 preamp control to 15dB attentuation (along with 10dB at my power amp) and there's plenty of adjustment at low volumes.

This also means that my maximum volume is now 15dB louder!! Which is nice... :-)


Hi Sean,

Just downloaded the latest firmware, and it works perfectly. Thanks!

(BTW, I've set the SB2 preamp control to 30dB attenuation, *and* my power amp to 10dB attenuation, and the volume is just about perfect! I must have a particularly loud amp!!!)

Chris

cbemoore
2005-09-15, 22:36
Just to clarify any potential misunderstanding:

My maximum volume has only increased by 15dB because I've decreased my SB2 preamp attenuation setting from 30dB to 15dB.

I wouldn't want anyone to think that the new volume control changes will magically increase the SB2 maximum volume! It simply provides better volume resolution at low volumes.

Chris

hansg
2005-09-21, 04:10
I have two slimp3-players and one squeezebox2, and they are all connected directly to power amps.

With this fix the whole volume range for the squeezebox2 is useable, but for the slimp3 I can only use the first quarter without blowing my ears and/or speakers.

Is it possile to do this for the slimp3-players? Or do I have a great excuse to buy two more squeezebox2?

Regards,

Hans

pfarrell
2005-09-21, 07:53
On Wed, 2005-09-21 at 04:10 -0700, hansg wrote:
> Is it possile to do this for the slimp3-players? Or do I have a great
> excuse to buy two more squeezebox2?

I don't know officially, but I've only seen comments that it might be
possible on SB1s, nothing about Slimp3s.

Seems like a great excuse to buy some Squeezeboxen.


--
Pat Farrell
http://www.pfarrell.com

sleepysurf
2005-10-17, 06:06
Just wanted to thank the Slim guys for implementing this so fast. I'm no techie, but am now feeding my Sunfire Cinema Grand amp directly from SB2 analog outs, and my system has NEVER sounded so pure. With this setup, the native SB2 DAC sounds fantastic. Can't even imagine going with an outboard DAC.

Whatever future volume/output voltage tweaks are implemented will be icing on the cake!

rme
2005-10-20, 04:35
Well this is fascinating - the short answer is yes we can probably make it work.

Our noise floor in the audio band is down at -127 dBu. However delta-sigma DACs inherently generate impulses (high frequency content) of some couple hundred microvolts even when fed a DC level - i.e. you trade much higher performance in the audio band for some out-of-band noise. Even for all-zeroes the DAC is actively working to hold the output between two levels.

So what is tripping your amp is probably this component of the noise floor. Looking at the rms level of noise floor it is about -60dBu when the DAC is active. I tried setting the attentuation to -64dB (mute) but this apparently does not actually disable the DAC, it's the same as playing zeroes. Next I tried inhibiting the clocks completely, which worked - it goes down to -90dBu, which is the same as the analyzer measures with RCA cables not connected to anything. Next I tried disabling the DAC by setting its reset bit. This worked too. So with some smarts in the SB2 to shut off the DAC when appropriate we should be able to do this.

http://www.seanadams.com/sb2_noise_floor.gif
http://www.seanadams.com/dac_disabled.gif

Sean,

Any update on this as my powered sub-woofer which is directly connected to the sb2 along with my amp (thru a y rca splitter) stays on all the time due to the noise floor issue.

occam
2005-10-20, 09:25
Just a quick comment to say how impressed I am with the speed and sophistication that you've brought to bear in making the variable outputs flexible and resolving.

Kudos!

seanadams
2005-10-20, 09:51
Sean,

Any update on this as my powered sub-woofer which is directly connected to the sb2 along with my amp (thru a y rca splitter) stays on all the time due to the noise floor issue.

The feature is not in yet, but I did get as far as testing the disabling of the DAC to confirm that this will probably work for your situation

sbjaerum
2005-10-20, 10:17
The feature is not in yet, but I did get as far as testing the disabling of the DAC to confirm that this will probably work for your situation

Is there a bug registered for this feature?

Steinar

Robin Bowes
2005-10-20, 10:20
rme said the following on 20/10/2005 12:35:
> Any update on this as my powered sub-woofer which is directly connected
> to the sb2 along with my amp (thru a y rca splitter) stays on all the
> time due to the noise floor issue.

Does it turn off if you disconnect the SB2?

Does it turn off if you disconnect the amp?

R.
--
http://robinbowes.com

If a man speaks in a forest,
and his wife's not there,
is he still wrong?

seanadams
2005-10-20, 10:21
Is there a bug registered for this feature?

Steinar

Yes, we are going to add it along with #1397