Can Good Amplifier Design Mitigate Sonic Differences Between Cables?

[gene]

We uncovered an audio myth on our YouTube channel I didn’t even know was being perpetuated by some esoteric amplifier companies. Apparently there is a belief that designing an amplifier to be too stable and protected to prevent damage or fire during misuse or short circuiting will result in sound degradation. Really? And how will designing an amp to allow RF ingress or oscillation not do the same or worse? Read on to find out.

Read: Can Good Amplifier Design Mitigate Sonic Differences Between Cables?

 

[Leroy Jenkins]

Really good read Gene,

It's really amazing how gullible the audiophile can be sometimes.

I can recall a few amp manufacturers that stressed you use a certain brand of cable. Even going so far as to say if you did not use the correct one it would void the warranty. I would say these are not properly designed in the first place and would not buy a product like that. Also things like DC offset protection should not have any influence over the sound of an amp, I wouldn't trust amp designers that say things like that and steer clear of their products.

 

[slipperybidness]

Good work Gene!

Rooting out more AudioFoolery!

First point-- Your speakers SHOULD be the most expensive part of your audio system chain. I will do whatever it takes to protect my speaker investment! Only a fool would risk expensive speakers. Even if a protection network did degrade the sound slightly (which a well designed protection network won't), big deal. Sound quality is a moot point when you fry your $3000 speakers!

On my current project, a parallel config LM4780 chipamp, I have done something very close to the "Figure #2" provided by Classe. Yet, I have taken care of other RFI concerns in a manner as described by the Emotiva engineers. That tells me that my research and design decisions are very closely aligned with the guys that get paid to do this kind of work. Awesome.

For me, there are 3 books that are the "gospels of amp design". If you aren't following the advice from these books, then I would not even consider purchasing your piece of junk!

Douglas Self
http://www.amazon.com/Audio-Power-Amplifier-Design-Douglas/dp/0240526139/ref=sr_1_2?ie=UTF8&qid=1433512789&sr=8-2&keywords=douglas+self

Bob Cordell (We actually studied some lessons out of this book at the local community college).
http://www.amazon.com/Designing-Audio-Power-Amplifiers-Cordell/dp/007164024X/ref=sr_1_1?ie=UTF8&qid=1433512819&sr=8-1&keywords=cordell+amp+book

G Randy Sloan
http://www.amazon.com/High-Power-Audio-Amplifier-Construction-Manual/dp/0071341196/ref=sr_1_1?ie=UTF8&qid=1433512859&sr=8-1&keywords=building+power+amps

 

[slipperybidness]

I wanted to add a few more comments on this topic and concerning this article.

This kind of poor design has also worked its way into a segment of the DIY builders too. There are DIY designs out there that are extremely minimalist. These designs have no Zoebel networks, DC coupling on the input, and severely under-spec'd filter caps!!! These guys claim that this design sounds incredible (and it likely does sound pretty dang good), but the tradeoff is poor stability and risk to your speakers! I'll pass on that!

Next, I wanted to call attention to Douglas Self's "Load Invariant Amplifier".
http://www.signaltransfer.freeuk.com/invarint.htm
I have been considering this as a DIY build. But, dang it isn't cheap!

Finally, I wanted to make a comment about the info provided by the Classe engineer in the article. From the article:
The problem of interfacing a power amplifier with the outside world in the presence of a wealth of wireless devices, dimmers and SMPS powered office equipment placed in close proximity to our music systems has become an increasingly difficult challenge.
So, Classe engineers readily admit to the challenges and problems associated with SMPS designs for audio applications. Yet, the Classe marketing department will shout from the rooftops about the new SMPS design that is in the latest Classe products! Oh, the irony

 

[highfigh]

I built my speakers into the amplifier, so the wires could be as short as possible. There's no measuring MY speaker wires in feet. When I press my forehead into the top of the case, the separation is incredible! Linear crystal, oxygen-free copper wire, cold-welded at each end.

It's really.......something.

 

[jinjuku]

The problem of interfacing a power amplifier with the outside world in the presence of a wealth of wireless devices, dimmers and SMPS powered office equipment placed in close proximity to our music systems has become an increasingly difficult challenge.

So, Classe engineers readily admit to the challenges and problems associated with SMPS designs for audio applications. Yet, the Classe marketing department will shout from the rooftops about the new SMPS design that is in the latest Classe products! Oh, the irony

I think you missed the context. Classe is referring to the junk SMPS's out there where the only guiding mantra was budget powering other devices.

There are a lot of well done SMPS amps.

 

[wshuff]

Really good read Gene,

It's really amazing how gullible the audiophile can be sometimes.

I can recall a few amp manufacturers that stressed you use a certain brand of cable. Even going so far as to say if you did not use the correct one it would void the warranty. I would say these are not properly designed in the first place and would not buy a product like that. Also things like DC offset protection should not have any influence over the sound of an amp, I wouldn't trust amp designers that say things like that and steer clear of their products.

A friend of mine once won a pair of "high-end" speaker cables. They were pretty, with a braided fabric covering, terminated in either spades or bananas, I don't know. What I do know is that he sliced open the fabric and underneath was the same bulk Belkin cable that you can buy for cheap on your own. But I'm sure that braided fabric was worth the price that some paid for it, and surely would prevent an amp's warranty from being voided.

 

[highfigh]

A friend of mine once won a pair of "high-end" speaker cables. They were pretty, with a braided fabric covering, terminated in either spades or bananas, I don't know. What I do know is that he sliced open the fabric and underneath was the same bulk Belkin cable that you can buy for cheap on your own. But I'm sure that braided fabric was worth the price that some paid for it, and surely would prevent an amp's warranty from being voided.

Belkin, or Belden? Belden is a manufacturer of cable, Belkin sells cable.

 

[slipperybidness]

I think you missed the context. Classe is referring to the junk SMPS's out there where the only guiding mantra was budget powering other devices.

There are a lot of well done SMPS amps.

Ok. I will give you that one.

I guess they did choose their words very carefully on that one.

To be clear, the entire concept of SMPS have inherent problems. Those problems then must be engineered out in other ways, and you simply don't have similar problems when using a linear supply for audio applications. Thus the SMPS becomes a much more complex beast!

Yes, there are some SMPS out there that have been well engineered to work for audio applications.

Personally, I doubt that I would ever have one in my critical listening rig. On the other hand, there are many advantages to using SMPS and Class D topology in HT applications.

 

[wshuff]

Belkin, or Belden? Belden is a manufacturer of cable, Belkin sells cable.

It would have been the manufacturer, Belden.

 

[3db]

Gene....Thanks for the chuckle. You really are preaching to the choir in here. Its out "there" where you have to take your light sabre and clean house .

 

[Chu Gai]

We uncovered an audio myth on our YouTube channel I didn’t even know was being perpetuated by some esoteric amplifier companies. Apparently there is a belief that designing an amplifier to be too stable and protected to prevent damage or fire during misuse or short circuiting will result in sound degradation. Really? And how will designing an amp to allow RF ingress or oscillation not do the same or worse? Read on to find out.

Read: Can Good Amplifier Design Mitigate Sonic Differences Between Cables?

A number of years ago, Kal Rubinson was reviewing a multichannel amplifier from Adcom. As part of his review, he hooks up the amp to his speakers using one of the scores of cables he has been given by various companies. Lo and behold he found the amp was unable to drive his speakers as it kept shutting down. Bad amp? Well, yes and no. Switching out the cables for something else took care of the problem. Using the problematic cables on some other amp did not result in a shutdown.

When he contacted the cable manufacturer, he was informed that this problem, while rare, is not unknown. To rectify the phenomena, they had a contraption consisting of a resistor and capacitor that one could attach to the amp along with their cable, a Zobel circuit if you will.

The cable manufacturer was Alpha Goertz. They make and sell a line of flat cables made from copper or silver.What is unique about their cables is the two conducters are separated by a thin dielectric. The result is that the inductance is vanishingly small but the capacitance is quite large.

The stated advantage of such a design is a negligible effect on FR because of the trivial inductance.

Not so much nowadays, but I've heard about similar amp problems where they shut down with certain speakers. Invariably those speakers have an impedance profile that have ever decreasing phase angles at higher frequencies indicating they're presenting a more capacitive load.

A number of years ago, I was reading through Frank Van Alstine's writings from days gone by. He had written that when he tested a Bedini tube amp by adding a small amount of capacitance to the speaker outputs it wet into cardiac arrest. Given these are the fols who make the Bedini Clarifier, one might expect that while clever, they're pretty stupid when it comes to electronics.

What I'm getting at here, Gene, is that it would be my strongest recommendation for you to purchase or obtain by other means something like 4 meters of Alpha Goertz flat cable and let that be a litmus test for when you do an amplifier review. It would provide an indication of the amp's stability into a capacitive load.

Thoughts?

 

[gene]

A number of years ago, Kal Rubinson was reviewing a multichannel amplifier from Adcom. As part of his review, he hooks up the amp to his speakers using one of the scores of cables he has been given by various companies. Lo and behold he found the amp was unable to drive his speakers as it kept shutting down. Bad amp? Well, yes and no. Switching out the cables for something else took care of the problem. Using the problematic cables on some other amp did not result in a shutdown.

When he contacted the cable manufacturer, he was informed that this problem, while rare, is not unknown. To rectify the phenomena, they had a contraption consisting of a resistor and capacitor that one could attach to the amp along with their cable, a Zobel circuit if you will.

The cable manufacturer was Alpha Goertz. They make and sell a line of flat cables made from copper or silver.What is unique about their cables is the two conducters are separated by a thin dielectric. The result is that the inductance is vanishingly small but the capacitance is quite large.

The stated advantage of such a design is a negligible effect on FR because of the trivial inductance.

Not so much nowadays, but I've heard about similar amp problems where they shut down with certain speakers. Invariably those speakers have an impedance profile that have ever decreasing phase angles at higher frequencies indicating they're presenting a more capacitive load.

A number of years ago, I was reading through Frank Van Alstine's writings from days gone by. He had written that when he tested a Bedini tube amp by adding a small amount of capacitance to the speaker outputs it wet into cardiac arrest. Given these are the fols who make the Bedini Clarifier, one might expect that while clever, they're pretty stupid when it comes to electronics.

What I'm getting at here, Gene, is that it would be my strongest recommendation for you to purchase or obtain by other means something like 4 meters of Alpha Goertz flat cable and let that be a litmus test for when you do an amplifier review. It would provide an indication of the amp's stability into a capacitive load.

Thoughts?

As soon as you mentioned a cable causing an amp to go unstable, I thought of Goertz. Sandwiching flat conductors results in very low inductance at the expense of very high capacitance. Not a good idea.

I did measure and write about this cable over a decade ago:
http://www.audioholics.com/gadget-reviews/speaker-cable-reviews-faceoff-2

I probably still have that cable in a storage bin somewhere but you'd have to take the zobel network off to see what it will do to a marginally stable amp that has a very high unity gain crossing.

 

[jinjuku]

Would like to see how a pro-audio amp would drive a highly capacitive load vs a high end 'audiophile' amp.

 

[Chu Gai]

As soon as you mentioned a cable causing an amp to go unstable, I thought of Goertz. Sandwiching flat conductors results in very low inductance at the expense of very high capacitance. Not a good idea.

I did measure and write about this cable over a decade ago:
http://www.audioholics.com/gadget-reviews/speaker-cable-reviews-faceoff-2

I probably still have that cable in a storage bin somewhere but you'd have to take the zobel network off to see what it will do to a marginally stable amp that has a very high unity gain crossing.

Sure, take the Zobel off. But if you had it, I think it would be a pretty cool test to run. If an amp had no issues, it might give some suggestion that the amp would be unlikely to have issues with most speakers as far as stability.

 

[Paul Scarpelli]

Twenty years ago I was part of the product development and sales team for the reintroduction of the new Harman Kardon Citation products, which were really Fosgate Audionics gear under the Harman name. Particularly troublesome were the amplifier projects. There were two 4-channel amps, designed by the pragmatic and talented Stephen Mantz, who I contracted to design the original Fosgate Audionics amps before Harman bought us. Harman insisted the amp design get the blessing of a "favorite engineer," and he decided to make some "changes" to make the amps "better." Actually, I think he wanted to make them "betterer." One design change was running the amp's frequency response from DC to 600 kHz, an invitation to disaster. Although you can make an argument that an amp that isn't high-passed may sound better, DC does tend to obliterate speakers. In addition, the amps oscillated and burnt up until we dropped the HF response down from Channel 13 to 200 kHz. And on top of it all, this designer had the amps biased so hard that at idle, you could use them for a hibachi. You couldn't touch the top of either amp without burning yourself.

All these things were done to "make the amps sound better" when in fact, not only did they not sound better, they became hand grenades. They succumbed to reason, and the Citation 5.1 and 7.1 amplifiers turned out to be good sounding and reliable...but not without a fight with the engineers who had to futz with everything.

 

[scattershot]

Hey guys:
I have always wondered why instead of using RCA pre-outs and XLR jacks that manufacturer's don't create a digital interconnect standard between amplifiers and pre's/pro's (or audio video receivers) so that the signal is passed digitally all the way to the final stage - the amplifier?
For any input that is a digital signal (ideally DTS HD, DTS Master HD, or some other type of lossless audio) wouldn't it be best if the signal passed digitally all the way to the amplifier?

I know that isn't ideal for analog aficionados, but I'm taking for most people who only have digital sources...

I also know that some people use monoblocks, some people 5 channel amps, etc., but they could also create, for example, one (space-saving) micro-HDMI output per each audio channel.

I mean a maximum quality (in terms of through-output) HDMI cable is under $10, so it's not just convenient and result in 0 loss in signal quality, but it also can save consumer money on cabling between the amplifier and pre/pro/avr.

Thoughts?

 

[slipperybidness]

Hey guys:
I have always wondered why instead of using RCA pre-outs and XLR jacks that manufacturer's don't create a digital interconnect standard between amplifiers and pre's/pro's (or audio video receivers) so that the signal is passed digitally all the way to the final stage - the amplifier?
For any input that is a digital signal (ideally DTS HD, DTS Master HD, or some other type of lossless audio) wouldn't it be best if the signal passed digitally all the way to the amplifier?

I know that isn't ideal for analog aficionados, but I'm taking for most people who only have digital sources...

I also know that some people use monoblocks, some people 5 channel amps, etc., but they could also create, for example, one (space-saving) micro-HDMI output per each audio channel.

I mean a maximum quality (in terms of through-output) HDMI cable is under $10, so it's not just convenient and result in 0 loss in signal quality, but it also can save consumer money on cabling between the amplifier and pre/pro/avr.

Thoughts?

Signal must get converted form digi to analog eventually, the sound that we hear is analog.

For something similar to what you are suggesting, go read up on Nad's "Direct Digital Amplifiers"

 

[scattershot]

Signal must get converted form digi to analog eventually, the sound that we hear is analog.

For something similar to what you are suggesting, go read up on Nad's "Direct Digital Amplifiers"

Yup I understand what you are saying and I definitely know all sound we hear is analog. I just thought that signal loss could be minimalized if the audio could remain digital until the very last stage (it hits the amplifier).

I don't know enough about the physics behind this method of using a lossless digital type connection to amplifier vs. RCA preout / XLR Jack... can anyone with knowledge in this area share their thoughts on this?
Thanks

 

[Swerd]

Yup I understand what you are saying and I definitely know all sound we hear is analog. I just thought that signal loss could be minimalized if the audio could remain digital until the very last stage (it hits the amplifier).

I don't know enough about the physics behind this method of using a lossless digital type connection to amplifier vs. RCA preout / XLR Jack... can anyone with knowledge in this area share their thoughts on this?

In most home audio systems, an analog audio signal is amplified in several stages, not all at once. A typical line level analog signal is about 0.1-0.3 volts. This is the level of the analog signal coming out of a typical audio source, such as a disc player + DAC. At the preamp level this signal is amplified to the range of about 2-8 volts. A typical audio amplifier has an input sensitivity as low as 1-1.5 volts, and amplifies the signal to as high as 100-200 volts depending on the amp power.

What you proposed would have the signal remain digital until going to the power amp. After conversion to analog it would still have to be boosted to preamp level before going to the power amp.

I believe these various voltage levels are conventions widely used by the audio industry to allow different components to be compatible. They aren't dictated by physics, however, there may be practical reasons why they are used.