Re-boxing SB2

[Triode]

Andrew,

Any views on the noise spectrum of Nicads/NMHi batteries? I am speculating about a battery psu for the HCU04. As I am only interested in digital out I am wondering whether this is a way to go. I'm Currently using a simple linear reg for this.

Adrian

[Yannzola]

Andrew,

Any views on the noise spectrum of Nicads/NMHi batteries? I am speculating about a battery psu for the HCU04. As I am only interested in digital out I am wondering whether this is a way to go. I'm Currently using a simple linear reg for this.

Adrian

You should get in touch with Vinnie at Red Wine Audio, SB2 battery mods are his specialty. http://www.redwineaudio.com/SB2_Mods.html

y.

[Triode]

Sean - any chance you could confirm the family of the Xilinx chip. I just want to check it doesn't have any problems with input voltages above 3.3V (should be fine if interfaces to 5V or 3.3 V logic which I think it does based on the datasheets I've looked at)

The thing I want to try is is 3x 1.2V cells to the HCU04....

[Andrew L. Weekes]

I'm not a big fan of batteries, under most real-world loads they are always worse (noisier, lower bandwidth, poorer transient performance) than a *well-designed* linear regulator BUT they are possibly easier for the average DIY'er to get reasonable results from, hence the appeal, I guess.

Added to the inconvenience of charging and maintenance, they'll never be my first choice, I've certainly never acheived state-of-the-art performance with them. I've never tried NiMH, but have no reason to suppose their chemistry makes them any better than NiCD for audio.

Andy.

[joshk]

Andrew,

What do you mean by lower bandwidth for batteries? I understand your other two points, but I am not grasping this one.

[Mike Hanson]

It doesn't maintain voltage (and/or supply current) for some frequencies as well as others. For example, it might handle the big, slow surges of LF, but not the fast, finesse of HF.

-=> Mike Hanson <=-

[Triode]

Well it was worth a play, but I agree with Andy's view - at least NiMH batteries aren't as good as a linear regulator and appropriate smoothing caps.

I tried 3 AA cells to power the HCU04 which is acting as spdif buffer and oscillator for the 11.29 Mhz clock [12 Mhz clock lifted]. On a scope I can clearly see noise at the clock frequency on the spdif output in this configuration - indicating supply impedance. With a linear regulator and a couple of Oscon caps this is much less noticable. Sound wise the linear reg has more focus too. [NB you can only see clock noise once the 12 Mhz crystal is removed so the clock and spdif signals are in phase]

Adrian

[Andrew L. Weekes]

What do you mean by lower bandwidth for batteries? I understand your other two points, but I am not grasping this one.

D.C. power supplies are often considered as simply DC, i.e. thay have no AC content, but of course the reality is this is never the case.

For a PSU that experiences a dynamically changing load, as most do, it's response to this is important.

In an active regulator, it is attempting to keep it's output constant, based upon the changes it measures through it's (internal or external) feedback connections.

The load placed upon any power supply will have an inherent bandwidth, i.e. it will be making current demands upon that power supply, in relation to it's own internal activity. In the case of something like an audio amp, the demands placed upon the PSU will be related to the audio frequencies it's dealing with, primarily, in the case of digital they will be related to clock speeds and edge transitions, which can generate very high frequency demands on a PSU.

The reality is one can never deal with very high frequency demands actively, only through passive means, like decoupling etc.

The bandwidth of a PSU therefore needs carefully defining, in order to ensure it is capable of meeting the demands of the loads placed upon it, and keeping the supply within the design criteria. For feedback based regulators one needs to define these bandwidths very precisely to ensure that the system is stable and performs as expected. The regulator can affect the system being powered, but equally the system can affect the performance of the regulator, in a critical manner - it's for this reason it's much harder to do stuff actively.

For batteries, the ability of the system to respond to transient (and by definition higher-frequency) demands, is related to the ability of the battery to maintain a constant DC potential, which is related to the internal impedance of the cells and the chemical reaction that goes on internally.

A chemical reaction is, by it's inherent nature, slow in comparison to the demands of either an audio, or particularly, a digital system. It also tends to be ill-defined in terms of response to a transient demand, which from my perspective makes batteries a poor choice unless care is taken to limit the demands placed upon them, which ususally requires active circuitry after them, to acheive this.

Andy.

[occam]

I'm not a big fan of batteries, under most real-world loads they are always worse (noisier, lower bandwidth, poorer transient performance) than a *well-designed* linear regulator BUT they are possibly easier for the average DIY'er to get reasonable results from, hence the appeal, I guess.

Added to the inconvenience of charging and maintenance, they'll never be my first choice, I've certainly never acheived state-of-the-art performance with them. I've never tried NiMH, but have no reason to suppose their chemistry makes them any better than NiCD for audio.

Andy.

Why in heaven's name, in the context of the SB2, would one compare a raw battery to a linear regulator? Isn't this why the good Lord, in Her infinite wisdom, gave us capacitors? Isn't the proper comparison between a 6v SLA and your toroidal transformer & rectifier bridge? What precludes us from adding those charming Pannasonic FC/FMs or Nichi HE after both? As our basic requirement is a 1 amp, 5Vdc supply, for batteries one presently has to choose a 6v 12a SLA and regulate down.