Here’s the old man again
A bit of mystory
Spoken as an EE engineer, it’s really not that difficult to make an amplifier to yield <-120dB THD+N @ 50W or even up to 100W. Get a good op-amp and apply as much as NFB in a big close loop from the a high power buffer back to the input, hell yeah!!! That’ll be a tougher game if we are going to apply PFFB for class D, but that’s definitely doable, especially that we have the high speed GanFet today that we can increase the audio bandwidth much much wider than before.
I started this hobby for more than 15 years ago, and once I was just like other engineers, which is very similar to the guys in A*R who just always talk about measurements, and thought that the lower the THD = the higher accuracy of audio reproduction. If one doesn’t like the sound coming from a 0.00000001% THD+N device, that’ll be his very own problem as he can’t accept what a real tone should sound like, legit right?
One day, I was so proud of myself looking at the FFT spectrum captured from the QA400, my amplifiers yield a result that beyond the capability of the QA400 could capture, and definitely score much better than those amplifiers I have cloned from.
…But wait, “I am using an AUDIO POWER AMPLIFIER, terminated with a 4ohm dummy resistor and using my eyes LISTENNING to a spectrum on the LCD screen?” Ain’t these devices supposed to be built and used as an entertaining materials? An enjoyable, relaxing and a luxury one? …My real AUDIO electronics engineering journey was started since then….
**Disclaimer: I don’t deny that the AA is useful and important in developing a product, especially when debugging an analog circuit.
Ever Heard of Push-Pull Topology?
The signal swinging from lower transistor to the upper transistor, even if it’s a 100% identical matched device, measured superb well and linear from lower to upper part without readable glitch. We don’t do that, we will not allow a signal to swing from one device conduct to another back and forth, this is not how a physical way of audio waves should be structured. Now you might ask, why there are push-pull topologies in the market??? The reason being it does provide a lot of extra power headroom, and most customers prefer larger power rating and think its more worth to have these number.
A controlled sound
Again, yielding low THD + low noise is not difficult, but does low THD really equal to good sounding and enjoyable? Are you sure? Have A*R forgot that the real target sectors of these products are someone who spent money to enjoy and have fun but not for doing any signal processing job?
For sure yes there are some good reasons when we should look into the FFT, eg when we debugging or developing a product. But don’t forget that some $100k ultimate machines does hiss when you put your ears next to the tweeter, and some $200 amp can be actually dead silent. Sometimes there are many sacrifices behind just to make the device sound nicer. Think about tube amp with poor PSRR.
And don’t forget that not all harmonics are bad, some engineers did it purposely because it does make the sound more enjoyable when they know how to do it.
Heat & Bias
In a push-pull design, signal swinging from the upper to the lower transistors, when the signal is crossing the transistors, one of it will conduct and the other will turned off, hence the signal glitching everytime crossing the center node. To avoid the ON/OFF switching glitches, we can force an amount of bias into both of the transistors and prevent it from turning off. There’s where the heat dissipated from, and we call it as Class-A operation.
Up to this point, you should understand that we can force an amount of bias current to turn the transistors ON all the time to prevent it from turning off, that’s what we call it as Class-A operation.
Unlike other Single Ended Amplifiers available in the market, you can try to imagine that there is a core transistor in the SS300B that is always ON, all the time. And the transistor is conducting like a copper wire even during power-off state. Yes, you hear me right, that’s a depletion mode transistor, a rare one. This is the best way to deliver the smoothest signal right into your speakers! The whole circuits in the amplifier are used to tame this always ON device into a stable state. A DC servo supervisory, a unique self-biasing circuit, a bit of current feedback, and a purely single ended topology from input to output. How does it sing? Smoothest ever, PERIOD.
Fully discrete amplification, no OPAMP, no Bl*ckg*te, no gimmick components, no gold wires, no audiophile cap, no W*T connector, no titanium made fuses, no 100mm thickness CNC faceplate, no FMPassKrellAccualiexpress1:1clone, no bullshit, just pure audio circuit engineering, we don’t rely on other audiophilebrandedstuffs to make our device shines. That’s what we call this an Ultimate Machine. The only thing taxing the selling cost is the design topology of this power amplifier.
Ok, everything sounds great now, next to do is to wait a CCY clone from Aliexpress? We are sorry to say this is not going to work as there are custom manufactured component inside our amp which there is no drop in part in the market. Btw, who’ll be so free cloning a sub USD5k price tag power amplifiers?
Lastly, if you are looking for a good measuring machine, we are sorry to say that the SS300B is definitely not for you because it carries a high level of 2nd harmonics, as what you will see in those tubes measurements.
Or the worst case that some of you will prefer a Gainclone LM3886 with WBT connectors, pure silver soldering, drop in some blackgates /mundorfs, cloth on a shining 100mm CNC chassis, and put a price tag of USD20k, which I saw that there are people buying amplifiers by looking on these components, then we are sorry to say that our amplifiers are not for you.
Why should you consider the SS300B? Or I should ask, why you should consider a real great Single Ended Class-A amplifier for your high end gear? Because that’s the only proper way to reconstruct the audio waves.
We are using a completely different design route than what most other engineers trying to do. Try think of the old Aleph, but designed in a much more elegant way, pure single ended from input to output, self biasing, DC servo, depletion mode Class A, a bit of current feedback, and a lot more implementation to make this amplifier out match other.