[BradJudy]

I saw this on another forum.  It looks like the next generation of BC products is all about massive capacitance being used essentially as a battery. 

Product page:

http://www.bluecircle.com/index.php?page_id=8721

An article:

http://www.ultraaudio.com/features/2008_07_01.htm

That's a lot of large caps (the one pictured above has something like 2500 farads on it).

[JoshK]

It'd be interesting to model such a PSU with PSUDII and see what affects so much capacitance has on the results. 

Generally the rule of thumb is to design with enough capacitance to knock down the ripple to reasonable levels and gaurd against overshoot and ringing, but also not so much that rise time is unduely long.  But I would hazard a guess that at some point, there is so much capacitance that there is no change from a change in current.      I don't know the answer, but its an interesting question.  Obviously any diy'er could do something similar as there isn't a lot of clever design in this approach. 

[poseidonsvoice]

Obviously any diy'er could do something similar as there isn't a lot of clever design in this approach. 

Right you are. It's all been done before. YAWN. Thank god he's working with solid state, with low voltages. With tubes, you stick you fingers in that chassis, and bye, bye...

Anand.

[JoshK]

Your right that a low impedance high voltage supply would be particularly lethal.  However, the idea of parralleling a box full of motor run caps does have its appeal.   

[BradJudy]

Right you are. It's all been done before. YAWN. Thank god he's working with solid state, with low voltages. With tubes, you stick you fingers in that chassis, and bye, bye...

Even lower voltages than I expected.  One of his comments about the production design was that it would use 220F 2.3V supercaps. 

[aerius]

However, the idea of parralleling a box full of motor run caps does have its appeal.   

I did that once.  And deliberately shorted it with a sacrificial steel bolt.  I'm never trying that stunt again.

[Occam]

Even lower voltages than I expected.  One of his comments about the production design was that it would use 220F 2.3V supercaps.

That's not particularly indicative of the rail voltages. 10 220F 2.3v caps in series give you a 23v cap of 22 farads, that's 22,000,000 uf. Yeah, it takes some clever equalization of voltages between those caps, but Gilbert is an incredibly clever guy.
Others might yawn, but even if you go to batteries, an SLA is pretty dang noisy below 10kHz, and you need caps across them anyway. (or regulators, capacitance multipliers, etc....)
Gilbert is simply taking advantage of the economics brought to us from the digital world, and very cleverly utilizing them.

FWIW,
Paul

[BradJudy]

That's not particularly indicative of the rail voltages. 10 220F 2.3v caps in series give you a 23v cap of 22 farads, that's 22,000,000 uf. Yeah, it takes some clever equalization of voltages between those caps, but Gilbert is an incredibly clever guy.
Others might yawn, but even if you go to batteries, an SLA is pretty dang noisy below 10kHz, and you need caps across them anyway. (or regulators, capacitance multipliers, etc....)
Gilbert is simply taking advantage of the economics brought to us from the digital world, and very cleverly utilizing them.

Yeah, I figured he must be doing something like that - there aren't a lot of components in the audio world designed for 2v rails.  It certainly isn't a cheap way to store power though. 

[lonewolfny42]

Gilbert just loves that silicone.....



Photo from the March 2006 Blue Circle Rave with his prototype speakers....here...

[Papajin]

Gilbert has a few more pics on flickr as well as a thread on the Blue Circle forums concerning "Pinkie" if you're interested in a bit more info.

[kyrill]

Silicon has a terrible dielectric constant of >10

[Frihed91]

These are storage caps not filtering caps.  You can't really see the preamp circuit through the maze of storage caps.