IGG: Hyperbolic Audio Amplifier

[EEVblog]

Yes, it's as LOL worthy as you can imagine from the title!
The guy actually emailed me and asked what I thought of it 

https://www.indiegogo.com/projects/introducing-hyperbolic-audio-amplifier-technology#/

[Muttley Snickers]

This IGG campaign is silly in more ways than one, it states that if you contribute $50 bucks then you will receive 50% off the retail value, how much is 50% off an unknown amount exactly ?, ridiculous.   

I just did the maths and it comes out to half price so it must be a good deal then.   

[Kean]

Those rewards! 

[FrankBuss]

"The frequency spectrum of an output of a hyperbolic amplifier circuit is shown in Figure 1 for an input of 1 kHz sine wave signal. The frequency spectrum graph confirms that the hyperbolic amplifier circuit creates a rich set of low order harmonics "



Yeah, -10 dB at 3x the input frequency for a 5 dB sine wave. This thing has a terrible THD and will sound horrible

[coppice]

Is it good or bad to put prank projects like that on IGG? I can't quite decide.

[X]

At least they're being honest about the presence of hyperbole.

[grumpydoc]

"The frequency spectrum of an output of a hyperbolic amplifier circuit is shown in Figure 1 for an input of 1 kHz sine wave signal. The frequency spectrum graph confirms that the hyperbolic amplifier circuit creates a rich set of low order harmonics "



Yeah, -10 dB at 3x the input frequency for a 5 dB sine wave. This thing has a terrible THD and will sound horrible

To be fair, he achieved his goal which is actually to have high levels of low order harmonic distortion.

However, I thought that the "valve sound" was just a question of the correct DSP algorithm these days

[Zbig]

However, I thought that the "valve sound" was just a question of the correct DSP algorithm these days

But that's not holistic.

[ebclr]

I beat those same people who pay 1000USD for a cable will make a bid there

[Cerebus]

"The frequency spectrum of an output of a hyperbolic amplifier circuit is shown in Figure 1 for an input of 1 kHz sine wave signal. The frequency spectrum graph confirms that the hyperbolic amplifier circuit creates a rich set of low order harmonics "



Yeah, -10 dB at 3x the input frequency for a 5 dB sine wave. This thing has a terrible THD and will sound horrible

Moreover, he hasn't figured out how to get a decent FFT display out of LTSpice.

[cat87]

  My god, that guy is really serious. At first I thought it was just another prank, but reading through this guy's claims, he really believes he has something there.

"I have invented, engineered and built a working prototype(1W) of a new class of analog solid state audio amplifiers that will called as Hyperbolic Amplifiers"
 - Ok, so building a 1 Watt audio amp and testing it isn't really that relevant. The distortion behavior is going to be way off, compared to a 20W or 100W model (here I'm referring to the same "hyperbolic Amplifier" sh... stuff)

"Hyperbolic Amplifier's open loop transfer function approximates a linear function and therefore negative feedback is not required. Also no hard clipping occurs when overdriven. This means spectral profile of the amplified signal consists primarily of low order harmonics."
 - So... no negative feedback? Ok, that sounds (pun intended) like a recipe for disaster. Harold Black would be turning in his grave right about now. I wonder how long will one of these amps take to fry someone's 1000$ speaker set?

"The hyperbolic amplifier is extremely power efficient. For the prototype, quiescent current is 20 mA. The power efficiency (ratio of output power to drawn power from power supply as a percentage) at maximum load is 95 percent.This is comparable to that of a class D amplifier!"
 - 20mA Iq for a 1 Watt amp? really? Just for giggles, a traditional 3 stage 100W audio amp using a complementary feedback pair output can have anywhere from 15mA to 30 mA Iq

"The goal is to mimic the spectral profile of a vacuum tube audio amplifier using only solid state transistors  - rich in lower order harmonics but negligible to none of the higher order harmonics."
 - So, after more that 40 years of research, development and innovation, where one can easily build a "blameless" amp, this guy is trying to increase the THD on purpose, just because it sounds nicer? 

Ok, I want to get off this planet.

[raj_houston]

Campaigner here. Thanks for all your inputs. Appreciate your comments. Some points to note :

The idea is not to compete with established companies who have mutli-million dollar budgets and an army of engineers who know very well what they do, but to complement the state of art.

- Ok, so building a 1 Watt audio amp and testing it isn't really that relevant. The distortion behavior is going to be way off, compared to a 20W or 100W model (here I'm referring to the same "hyperbolic Amplifier" sh... stuff)

The 1W was a just a reference point for proof of concept implementation. The circuit architecture is scalable. I am more interested in trying the circuit for headphone or desktop amplifiers which are in vicinity of 1W power mentioned above. Any perceived degradation in distortion for a 20W or 100W would be from that of deviation from ideal behavior of high power transistors (true for any amplifier circuit not just the hyperbolic amp. ckt).

So... no negative feedback? Ok, that sounds (pun intended) like a recipe for disaster. Harold Black would be turning in his grave right about now. I wonder how long will one of these amps take to fry someone's 1000$ speaker set?


Electronic systems including amplifiers predate Harold Black's invention and popularization of negative feedback technique. AT&T had signal amplifiers (without negative feedback of course) on their long distance telephone lines working in the field long before Black's invention. Not sure on what basis you are making that claim. Just to let you know, the theoretical equation for the output that I derived for the hyperbolic circuit matches well with hundreds of spice simulation experiments that I subjected the circuit to.

- 20mA Iq for a 1 Watt amp? really? Just for giggles, a traditional 3 stage 100W audio amp using a complementary feedback pair output can have anywhere from 15mA to 30 mA Iq


The Iq of the hyperbolic circuit is very much independent of output power rating. Over a wide range of  output power it remains constant. Like I mentioned my present interest is in the vicinity of a few watts. The more important point is the power efficiency. Analog audio amplifiers have power efficiencies reach at most of 80 % (in  many cases even less). Hyperbolic amp at full load reaches 95% efficiency.

[FrankBuss]

Hyperbolic amp at full load reaches 95% efficiency.

Is this another Spice simulation, or did you actually measure this efficiency with a real circuit?

[cat87]

Electronic systems including amplifiers predate Harold Black's invention and popularization of negative feedback technique. AT&T had signal amplifiers (without negative feedback of course) on their long distance telephone lines working in the field long before Black's invention. Not sure on what basis you are making that claim. Just to let you know, the theoretical equation for the output that I derived for the hyperbolic circuit matches well with hundreds of spice simulation experiments that I subjected the circuit to.

You are absolutely right. But, so am I  Implementing negative feedback lets you  increase you usable bandwidth by sacrificing some of the gain, lowers the THD and most importantly keeps things in check when the output stage starts getting hot. Not the case for the 1W, but if the architecture is scalable, as you say, then thermal runaway should be an issue to consider.

The 1W was a just a reference point for proof of concept implementation. The circuit architecture is scalable. I am more interested in trying the circuit for headphone or desktop amplifiers which are in vicinity of 1W power mentioned above. Any perceived degradation in distortion for a 20W or 100W would be from that of deviation from ideal behavior of high power transistors (true for any amplifier circuit not just the hyperbolic amp. ckt).

As a side note, you can basically take any circuit and scale it up to whatever lever of power you wish. How it will behave, now that's a whole different kettle of fish. To have the same behavior, i.e. the same level of low THD (of course, not the case with this design) at higher power, there are a lot of things to take into account. For example, the base-collector capacitance may not matter at 1W, but things start counting at 30W and up. Also, the thermal characteristic of the devices changes once the power goes up (device power dissipation actually goes down with increasing output levels), therefore so does the behavior of the amp as a whole.

Hyperbolic amp at full load reaches 95% efficiency.

It would be very interesting if you could do (you can do that even in a simulation) a graph of efficiency vs. output level.  I'm not calling you out on this, or something, it's just typical engineer's curiosity.

Now, just to throw some ideas into the wind.... To replicate the "tube sound" with BJTs, you could go to a lot less trouble. Just get a pair of BJTs with low f_T and low Beta. That should get you in the ballpark. To have  the right kind of harmonics, you have to take care when designing the amp, so that you have symmetrical currents going in and out of you input stage and VAS stage.

And, oh yeah, I almost forgot. Could you release a schematic of this thing? or at least some of the math behind designing this. Again, enginerd engineer's curiosity
 

[alexanderbrevig]

The exaggerating amp! https://en.wikipedia.org/wiki/Hyperbole heh

My guess would be that the hyperbolic part of this name comes from the soft clipping mapping from a sinusoidal to a hyperbolic sinusoidal.

Increasing harmonic content is all good fun. As a guitarist I do it all the time!