February 2006


Bolder Cable Modified Slim Devices Squeezebox 2
Network Music Player
Part 2 Of Several
(and no, I don't get paid by the word)
See Part 1 By Clicking Here

Review By Scott Faller
Click here to e-mail reviewer

  As you can tell from my last article introducing the Slim Devices Squeezebox 3, I've fallen off the computer based digital audio wagon. In fact, I got run over by a whole wagon train headed westward in their search for gold. No kidding, the SB3 is the best thing to hit the audio market in a long, long time.

Last go round I talked about the stock unit and how it sounded both unmodified and with a cheap linear wall wart power supply and an old CAL DAC. This issue we're going to explore the Bolder Cable modified version of the Squeezebox 2.

I'm sure you just picked up on the fact that the SB 2 and 3 look different.
Internally, there are differences between the two (as I'm told). The main circuit board has changed to accommodate modifications to the internal circuitry and architectural layout of the unit (the unit is now vertical rather than horizontal). As I'm told, the SB 3 has added a separate board for the input and output connectors. A couple of the major differences between the SB 2 and 3  are the SB 3's internal power supply rail (used for the digital buffer chips) now uses a linear regulator in lieu of a switching regulator. Another differentiator between the two units is the SB3 now sports dual, internal wireless antennas. I've not had the pleasure to compare a stock SB 2 to the SB 3 but the common consensus is that the stock units sound effectively the same.

If you're not familiar with Bolder Cable, Wayne Waananen has been doing business for a number of years selling his own designs of cables, interconnects and power cords. Wayne has also built a faithful following performing mods to commercial equipment. He has been providing mods to many of the new digital switching amplifiers, the Art DI/O external DAC's and now the Squeezebox. Wayne also carries a line of audiophile accessories and power conditioners including the infamous Bybee filters.

Wayne offers several different levels of mods to the SB. You can do anything from a Digital Only mod, to Digital Output and Analog Output mod right up to his uber-mods that include Bybee filters and multiple types of SoniCap'sin the signal path. Wayne also has custom designed a couple of different power supplies to be used specifically with the SB. The difference between the two is the addition of the Bybee's and different brands of capacitors for filtering and reservoir. Specifically, Jensen Four Pole electrolytic caps used before and after the regulator that are supposed to give a slightly different sonic flavor to your Squeezebox. Just to be clear, the power supply I have here for review does not have the Bybee's or the Jensen caps.

So Why Mod?

That's a good question, especially for the newbies out there. Hopefully, the next few pages will help you understand the reasons to have mods done. Sometimes, modding gear only produces a ‘different' sound, not necessarily a ‘better' sound. As I'll try to explain, you'll see what mods are performed and the reasoning behind each. This article will be scattered with my subjective opinions on the improvements gained by each mod. So lets have a go at it, shall we?

Lets Talk Power Supplies

Since the topic of changing the power supply of the SB is such a bone of contention to some, I thought I'd explain the basic of differences between the typical switching power supply and a linear power supply. If you remember, I  mentioned back in the first article that there was a dramatic difference in sound between the two.

The following is a brief explanation why the linear power supplies sound better than the typical inexpensive switching mode power supplies when used in audio applications (don't worry, I'll keep this really simple and I'll use pictures and graphs for those that don't want to do any research). Notice I emphasized the word inexpensive. Not all switching power supplies are inferior to linear. It all depends on the design, application and their ability to shunt noise whether internal or that which rides the incoming mains, display low ripple and transient recovery time (which isn't the case of the stock switcher supplied with the SB).

Electrical Noise & Ripple

As you can see from the waveforms from the typical 5 volt switching and linear power supplies, the switch-mode power supply generates pulses of power due to its switching (or pulse width modulation PWM) design. The linear power supply generates a nice, smooth voltage output.

Pretty graphs courtesy of Lambda Americas

Quoting the Lambda technical papers, linear power supplies can be up to 10,000 times quieter than their switching counterparts. That's pretty significant especially when it comes to high resolution audio.

Transient Response

This is a term that is completely familiar to the hardcore audiophile. Transient response is how a piece of equipment, speakers or a power supply (in our case) reacts to a change in load. As we all know, music is made up of nothing but  transient responses. If it wasn't, we'd be listening to sine waves.

Again, quoting directly from the Lambda technical papers,

"If the output load quickly changes from say full load to half load, the output voltage of the power supply will rise (overshoot) before the internal control circuit has time to compensate, and then undershoot a little less as the circuit over compensates. The length of time is takes from the instant of the load change to the time the output voltage settles back into the load regulation limits can be critical to some loads. Here the linear again outperforms the switch-mode.

For a 50% change in load the switch-mode will often take 3000us* to recover. A linear supply will recover in 50us."

The linear power supply provides a pretty dramatic improvement in recovery times from heavy load conditions.

Next is a table I swiped from the National Semiconductor website comparing switch mode to linear power supplies.

* Please note, Lambda Americas and National Semiconductor state differing transient recovery times (likely) due to differing switching power supplies under test.

A little explanation of this table is probably in order here. First are the things we really aren't concerned with (necessarily). Efficiency, honestly we as audiophiles could care less how efficient the products are. Heck, just take a look at all the guys that are running true Class A amplifiers and tubed gear. To us, it's all about the sound regardless of how well (and how much) the power is converted. I can heat my house when I turn on all of my tubed gear.

Next up is Voltage Range. Your power utility (typically) does not vary the incoming line voltage by more than 5% to 8% one way or the other. If they exceed a 10% variance, the power company is (typically) required to notify you. The voltage range states pretty much the obvious, the power supply won't put out the 5 volts (in our case) if the incoming power dips below 90% of ratings. So this really doesn't effect us.

Power Density. God knows what this is but I don't remember this rating in any of the books I've read. Besides, I'm not sure I've ever seen power rated in cubic inches before (except big block Chevy's).

Hold Up Time. This is of a marginal concern for us. What Hold Up time represents is the power supplies ability to maintain its 5 volt output (in milliseconds) in case of an instantaneous interruption of power.  The hold up time can change based upon the design of the circuit and it's total energy storage. Honestly, any commercial design of a linear or switching power supply will be able to handle those momentary interruptions that we never know even happen. I've sat though a couple of quick flickers of lights here at my house and the linear power supply didn't falter. The Squeezebox continued to operate.

Line and Load Regulation are both functions of the power supplies ability to either regulate or provide constant power at the output. As you can see from the table, the linear power supply does a far better job at maintaining the output voltage than the switching power supply. Now that said, does a 1% variance of the 5 volt output translate into something audible? Who knows. I'm not sure I could hear it but there is something to the numbers.

Output Ripple and Transient Recovery. These are the two biggies as far as I'm concerned. The heavy output ripple of a switching power supply translates directly into noise, high frequency hash, and generally a hard, grainy sound. Throw in poor transient response recovery time and now you have power supply that likely won't be able to respond well to the heavy transients found everywhere in music. Those crisp leading edges of transients and the trailing decay of the sounds will be minimized as a result. Will it be horrible? No, but it sure could be better with a nice, properly sized linear power supply.

BTW, all of this information, including the pretty pictures can be found at the Lambda Americas and the National Semiconductor Website.

As an aside, the Squeezebox switching power supply (as you can tell from its small size and weight) doesn't contain the type or quality of transformer the Bolder linear power supply does. Due to the nature of the typical transformer, some of the noise that rides on the back of our incoming power is eliminated by the transformer and the subsequent filtering capacitors in a linear power supply design. Some transformers do an inherently better job of filtering than others. Such is the case of the split-bobbin transformer. Coupled with quality capacitors that have a low ESR (equivalent series resistance) and DF (dissipation factor), the noise floor of a linear power supply will be much lower than the inexpensive switching power supply that comes with the Squeezebox. Such is the case with the Bolder power supply.

Since I've just (essentially) slammed the switching power supply, I feel the need to point out the obvious. 99% of the populous won't be able to hear the difference between the switching and the linear power supplies because their systems don't have sufficient resolution for the switching power supplies artifacts to be heard. It's only when you get into the high end realm of audio do the differences become clearly audible. This is no different than when you get into the high end realm of anything, Wines, Furs, High Definition Home Theatre, Jewelry or any market where true ‘premium' goods can be found. The differences are there but you have to develop your skills to see, hear, smell or taste a discernable difference.

The Bolder Rev 1 Power Supply

Before I get started with the comparisons between the stock Squeezebox power supply and the Bolder , I need to make you aware of a couple of factors. First and foremost, in my descriptions of how the power supplies sound, I am not going to be using any incoming power filtration. I will be powering these directly from my wall sockets. I typically use a DeZorel line filter which does an excellent job of cleaning up the garbage that rides on the mains without giving me that transformer based coloration like the big power stations.

The Bolder Rev 1 Power Supply is rather inauspicious to look at. The Rev 1 is a standard NEMA gray 4" x 6" x 3" box with squishy feet on the bottom to dissipate vibrations. On the line side of the Rev 1 you'll find an Acme Audio silver plated IEC connector.

Downstream of the IEC is an Acme Audio silver plated fuse holder. All of the internal wiring is completed with Silver plated Copper wiring that has Teflon insulation. Delving deeper inside this little gray box you'll find a well chosen EI core (split bobbin) transformer. This feeds a quad of HexFRED's taking on the rectification duties. Following the rectifier bridge is a Panasonic FC series filter cap which has a SoniCap Gen 2 as a bypass. Then comes a carefully chosen Linear Technologies regulator with properly sized heat sinks. Following the regulation comes another filtering stage with another Panasonic FC cap and Sonicap Gen 2 bypass, then it's off to the Squeezebox via a Neutrik XLR. The cable feeding the Squeezebox is silver plated copper with Teflon insulation.

As you can see from the picture, the layout is quite simple. Much of the costs for this power supply are the parts. The design itself is nothing ‘new' or ‘innovative'. It's a plain old, textbook power supply with one clear distinction, Wayne listened to how it sounded. He carefully chose the parts he used not only on how they measured but on how they sound.

Just in case you were wondering how the Bolder Cable Rev 1 Power Supply actually measures compared to the stock switching power supply, here's a couple of pics to demonstrate (unverified and courtesy of Bolder Cable). This first picture is the 5 volt waveform produced by the stock switching power supply. The second is a picture of the Bolder Rev 1 Linear Power Supply.

For you uber geeks out there, the scope used was a Tektronics 475. The probe used was a TEK P6109. The scopes settings were .5 volts per division and the time base was .5us.

As you can clearly see the Bolder Rev 1 Power Supply displays far more linearity in its waveform. Sort of reminds you of what I tried to explain above, doesn't it?

Now, I'm going to catch all kinds of grief for publishing the Bolder findings. I realize Slim Devices has published their own contradictory information regarding their switching power supply. It has been suggested on one of the forums that Wayne improperly measured the power supply with the switching power supply in place. As I see it, as long as Wayne measured from the same point on the board (which he did), no matter where Wayne grabbed the power rail, noise (ripple) is still noise and the Bolder Rev1 power supply made the power rail(s) measure better.

The Bolder Rev 1 Power Supply checks in with .2 mV of ripple (unverified) which is considerably better that the theoretical power supply quoted by National Semiconductor above.

Listening

OK, so enough of the geek stuff. For this part of the test I chose a single song from Alison Krauss and Union Station, Gravity which is found on Lonely Runs Both Ways. I chose this song because it is extremely well recorded, very simple and uncomplicated. Although not to every ones tastes, I really enjoy Alison and the guys.

First Up is the stock switching power supply. The first thing I notice is that the music carries less weight and impact. Dan's guitar doesn't have quite the resonant characteristics that it normally has (I've photographed him as he played about 8 feet in front of me at Eric Claptons Crossroads Guitar Festival). As Alison begins to sing, I can't help but notice some pretty heavy grain and sibilance in her voice that shouldn't there.

Most all of the detail you would expect is there. The soundstage is a bit shallow not extending much more than 5 feet behind the speakers. Placement of the performers on the virtual stage is fairly consistent though the performers are slightly enlarged.

The dynamics are definitely constrained which minimizes the immediacy of this recording. The bass is decent but lacking impact and depth also. 

And now the Bolder with a standard power cord... The very first things you notice are the dynamics of Dan's guitar. The lovely harmonics I've come to love are now being reproduced in spades. Its' as if the music has taken on a darker and more lively presentation.

The bass has become far deeper and now has impact. I'm hearing detail that was lost with the stock switcher. Detail like the clarity at the end of a sentence in the first verse sung by Alison "that turned out to be questions in the end". When Alison sings the word ‘end' you can clearly hear enunciation "ennnd-uuuh" as she finished the word. Though audible with the stock switcher, it covered up all of the nuances that bring the music to life. These are the things we audiophiles live for.  Cords from the guitar literally begin to leap from my Lowthers.

All of the grain and sibilance that I noticed with the stock switcher is now gone. Alison's voice sounds quite natural. The soundstage has grown in depth and in width considerably also. The Bolder Rev 1 power supply has brought the music to life.

Just to Irritate the Non-Believers...

This go round I've installed one of Wayne's uber power cords to replace my nifty 12 gauge power cord. This is the Nitro Plus (and then some). The Nitro Plus sports Silver Plated Copper conductors with Teflon as a dielectric. The conductors are braided in a Star Quad geometry. At the IEC connector you will find a pair of big honkin Bybee purifiers. Finally, Wayne has terminated the power cord with Oyaide connectors. What power cord wouldn't be complete without being double Cryo dipped. This little number checks in right at a cool $1000. That's right, this little jewel will cost you one cool Cleveland and is (not quite) twice what the mods AND the power supply cost combined. There is no substitute for excess. 

Listening to the Nitro Plus brings more air and naturalness especially to vocals. The music now erupts from the speakers. The presentation gets bigger and the imagining of the performers get tighter at the same time. The bass is less flabby and has more definition. The sound is far cleaner top to bottom. The Nitro Plus seems just a tad thinner sounding than the stock power cord which is actually a good thing for the bass but if you like a fuller, more embellished midrange sound, this cord dries it up a bit. The Bolder Nitro Plus definitely brings you closer to what has been recorded. Bravo!

Power Supply Ratings

These ratings are a comparison between the different power supplies. What I've done is set the Stock Squeezebox power supply as a Zero or a baseline reference. The differences are noted on a scale of –5 to a +5 set against the 0 baseline reference.

Digital Modifications

Next up are the Bolder Cable digital modifications. The main purpose of the digital modifications is twofold. First is to reinforce the power supply energy storage capacity (and quality) and the second is to reduce jitter as far as possible (within reason).

Here is a brief summary of the Bolder Cable digital modifications.

Replace two main power supply capacitors with higher value, low ESR versions

Replace the digital buffer capacitor with higher value, low ESR versions

Remove the 12.28MHz crystal oscillator (optional)

Remove the headphone opamp circuit (optional)

Reroute the digital signal path to the S/PDIF

Replace the stock S/PDIF with a 75 ohm WBT NextGen, gold plated copper RCA

The first thing you should know is that removing the 12.28 mHz clock crystal limits some of the internet radio stations that can be received. Basically, any station that streams music below 56k (48k and under) can't be received anymore. For me, that works just fine. I'd prefer not to listen to low bandwidth feeds anyway since they sound pretty crappy IMO. The other benefit of removing the clock crystal is it lowers jitter (as I'm told). I haven't verified this on a scope but lowering jitter is a good thing in digital.

Listening

Under these set of listening tests I've decided to use a pair of matching aftermarket 5 volt, 1.5 amp linear power supplies that I picked up from my local electronics surplus store. This gave me consistency of power quality. The power supplies were plugged into my DeZorel line conditioner giving the cleanest mains voltage available to me. The DAC used for the digital output comparisons is the MHdt Labs Paradisea (look for the review in the coming months). This is an extremely good sounding tubed DAC. It rates right up there with the sounds of the uber DAC's I've played with. The digital cable used was a solid silver design measuring a true 75 ohms.

Doing the comparisons I synchronized the two units leaving the volume controls independent so I could level match SPL's with my Sencore SP-295c. Switching DAC's was a matter of hot swapping the two SB head units. Unfortunately, I know of no digital switchers that don't introduce excessive amounts of jitter into the signal, hence the hot swapping. Again, the music I used under test was Alison Krauss and Union Station.

As I start hot swapping the two units the first thing that comes to my attention is the stock digital output (SDO) adds a rounded quality to the bass region. This rounding also adds a bloom to the midbass. Unfortunately the bloom tends to cover up some detail contained on the source music.

The Bolder Modified digital output (BMDO) tightens up the bass and midbass region considerably. In addition the midrange and treble regions become considerably cleaner also. There is a considerable amount more definition and treble extension. In turn you get better separation of the instruments, better definition and depth in the soundstage. In addition, the BMDO displayed considerably better dynamics than the SDO top to bottom.

You can hear that the artifacts of jitter have been reduced. The audible hardness and slightly grainy texture of upper mids and highs have been eased considerably. I couldn't say what the total reduction of the jitter is since I don't have a scope to measure it but it is definitely audible.

Ultimately I would classify the differences between the stock Squeezebox digital output and the Bolder Modified digital not as subtle, but absolutely audible. Obviously the best way to demonstrate the differences is to do a direct a/b comparison as I've done.

Digital Modification Ratings

These ratings are a comparison between the stock S/PDIF and the Bolder Modified S/PDIF output fed into a high quality DAC. What I've done is set the Stock Squeezebox 3 as a Zero or a baseline reference. The differences are noted on a scale of –5 to a +5 set against the 0 baseline reference.

Analog Modifications

Finally are the Analog mods that Bolder Cable does to the Squeezebox. The unit I have here under test is the Squeezebox 2. The SB2 has slightly more room inside its case than the SB3. This allows Wayne to use slightly different parts in his mods.

Here is a brief summary of the mods Wayne does to the SB 2.

The internal power supply caps are replaced with high quality Panasonic capacitors with a low ESR.

The cheap RCA analog output jacks are replaced with Cardas Rhodium plated RCA jacks.

The internal output gain stage is bypassed and coupled directly from the output if the DAC to the Cardas RCA jack with a standard SoniCap. This lowers the output voltage to just over 1 volt but should not be an issue as many preamps have ample gain to compensate. The unit under review has a pair of SoniCap Gen 2 bypass caps paralleled to the standard SoniCap.

The existing DAC's power is derived from the internal switching power supply rail. The modified DAC's power is rerouted to draw from the linear power supply rail.

The headphone opamp is removed (optional)

The internal 12.28 mHz clock crystal is removed (optional)

The stock Burr-Brown PCM1748 DAC is removed and replaced with the higher spec'ed Burr Brown KE series.

Here are just a few comments regarding the mods Bolder Cable has done to the Squeezebox. Removing the output gain stage has advantages and disadvantages. Getting rid of the mediocre JRC opamp and its associated gain stage circuitry causes a dramatic increase in clarity and detail. On the down side, the output voltage of the unit drops to just over 1 volt rather than the 6 volts in the stock unit. This shouldn't be a problem for most as many preamps have ample gain to boost this lowered output signal. In my case, the highly modified Korato KVP 10 isn't phased in the least with the lower input voltage and it has only 15dB of gain.

The other issue with removing the output gain stage is that the DAC doesn't like to see a high capacitance load from the interconnects. You will need to keep you interconnect lengths as short as possible. Using high capacitance interconnects will likely result in some loss of the sonic qualities gained by the analog mods. A simple, short parallel (or twisted) run of wires would be best to use with this mod. Advanced braided or other high capacitance topologies should be avoided.

The other mods that Bolder does to the Squeezebox are obviously to improve the sound. You can draw your own conclusions about each of the mods. I won't go into details as this article is getting lengthy but they all make perfect sense to me.

Listening

Again, to set this listening comparison up I used the same power supply scheme as described in the digital section. For this comparison I have used identical pairs of my own design of MIL spec silver plated copper interconnects that have Teflon insulation, shielding and utilize low mass RCA jacks. These were plugged into an NEC AVX-910 switching unit that has a remote control.

 As I begin, I start with the same set of reference music that I've used for the previous listening tests. Pushing play on AKAUS and flipping back and forth between the stock Squeezebox (SSB) and the Bolder Modified Squeezebox (BMSB), the differences are rather dramatic.

 The first thing that draws my attention is the very narrow focus of the music when played on the SSB. The sounds are strictly confined between my Lowthers. There is little depth width and height to the soundstage. Placement of the instruments is fair but it is a rather ‘soft' focus rather than a sharpened image of the BMSB. In addition, the SSB doesn't allow the soundstage to open up much beyond the outer edges of the speakers. In contrast, the BMSB casts an extremely large and well focused soundstage. The sounds and ambiance of the recording now completely fill my listening room.

Comparing tonality and clarity of the two units provides even greater differences. The SSB and its analog output tend to veil the music. Though the veiling isn't as great as some entry level CD players I've listened to in the past, it still exists. The bass region tends to be rounded and soft. This bloom extends up into the midbass region masking some of the midbass detail. The treble of the SSB, though quite easy on the ears, is rather soft, undefined and not very well extended.

Switching over to the BMSB is like listening to an entirely different piece of gear. Wayne's mods bring the depth, clarity and detail that we audiophiles live for. The bass is now extremely well defined which reveals new levels of detail in the midbass region. This new level of detail extends all the way up the octave band into the treble. The highs are now crystal clear and well extended. 

The soundstage has opened up completely with the BMSB. My Lowthers are now playing those psychoacoustic tricks I've come to love so well. Things like instruments being projected from three feet outside of the speaker cabinets. I now have excellent placement of the performers on the virtual soundstage that remain rock solid.

Another cool feature of the Squeezebox is the fact that it has an integral volume control. This means for you minimalists out there, you don't need a preamplifier if you are using your computer as your sole source. I tried the Bolder directly in front of my amps using the analog outputs and it sounded wonderful. Something to keep in mind (as I mentioned previously), the Bolder mods drop the output voltage to a little over 1 volt. Providing you have decent gain from your amp, you can easily drive your system to ear splitting sound pressure levels. One item of note, if you decide to connect the Bolder modified SB directly to your amps be sure to use the lowest capacitance interconnects you can find. Since you have (nearly) a direct connection to the DAC (with only a cap separating the two), high capacitance cables will start to act as a filter rolling off the highs. For the average audiophile with a modest preamp, the direct connection will sound dramatically better. If you have an extremely ‘transparent' preamp, the differences will be less noticeable.

Analog Modification Ratings

These ratings are a comparison between the stock Analog Outputs and the Bolder Modified Analog Outputs. What I've done is set the Stock Squeezebox 3 as a Zero or a baseline reference. The differences are noted on a scale of –5 to a +5 set against the 0 baseline reference.

Summing It All Up

The stock Squeezebox utilizing a decent aftermarket linear power supply can sound quite nice. Far better than your typical entry level CD player. As I've repeated over and over, the Squeezebox is the best product to hit the market in a long time. You have the ability to rip your CD's in a lossless format on your hard drive, then have all of them available at a touch of a button.

For those of us who seek the ultimate involvement and significant strides in resolution we will need to turn to guys like Wayne at Bolder Cables to perform modifications to the Squeezebox. The modifications performed by Wayne move the ultra-affordable Squeezebox clearly into realm of the best of the best CD players.

When you start adding the costs of the Stock Squeezebox, the standard Bolder Cable Power Supply and the standard Analog Mods together, you are still in the $1000 range (shipping included). For many of us into high resolution audio, that falls right in our sweet spot price wise. From the basic mods, Bolder Cable offers numerous individual upgrades combining different features and parts to fine tune the sound to suit your individual needs.

The Bolder Modified Squeezebox simply provides so much more of everything contained on our favorite discs. Its ability to convey the music is stunning. My CD collection (now ripped to hard drive) has never sounded better.

If you are in the market for a Squeezebox, I highly recommend you consider sending it to Bolder Cable and have it modified by Wayne. At the same time, order one of his Rev 1 Power Supplies. The costs are very reasonable when you weigh the leaps in sound quality you gain by having it modded.

Specifications

Type: Network music player

Pricing as reviewed
Power Supply - $230
Analog Mods - $265 with the Sonicap Gen 2 bypass.
Digital Mods in addition to the Analog Mods - $50 (not stand alone pricing)
Nitro Plus w/Bybees Power Cord - $1,000

Company Information

The Bolder Cable Company
8047 W. Morraine Dr.
Littleton, CO 80128

Voice: (303) 904-2443
E-mail: BolderCable@comcast.net

Slim Devices, Inc.
958 San Leandro Avenue
Suite 400
Mountain View, CA 94043

Voice: (650) 210-9400
Email: sales@slimdevices.com
Website: www.slimdevices.com