What price power?

[Head_Unit]

This thread (now closed due to obnoxiousness)
https://www.audiosciencereview.com/...no-block-power-amplifier-review.39678/page-17
shows you can get a [email protected] load of essentially perfectly measuring stereo power for $3000, and similar Stereophile's review of the Schiit Tyr monoblock
https://www.stereophile.com/content/schiit-audio-tyr-monoblock-power-amplifier
"Measuring" underlined because we can't listen online, so for this thread forget about how the amp sounds or doesn't sound please.

This makes me wonder, what is the CHEAPEST we can get such? Then we have do define "essentially perfect measuring" I suppose-feel free to post about that. I'm thinking
- Distortions generally below I dunno 0.1%
- Noise more than 100 dB below 2V. Or is 90 already overkill?
- Response basically load independent.
- Crosstalk...what's the overkill level here?
And whatever else I'm forgetting.

Power output I don't care, just wondering when there are state-of-the-art measuring DACs for what $500? $100? (I haven't even kept up) how much are corresponding power amplifiers (or integrated for that matter, a taller hurdle perhaps).

 

[jae]

Amplifier SINAD list ASR + 3rd parties

List Name,THD+N (5W, 4 ohms),SINAD (dB) (5W, 4 ohms),Power (W) (1% THD+N, 4 ohms),Price (USD),Source Topping LA90 (stereo, bypass, low gain),0.00009%,121,90,800,<a href="https://www.l7audiolab.com/f/topping-la90/">L7 Audio Lab</a>,Suggestions, corrections and more info here: Topping LA90 Discre...

docs.google.com

Simple answer is Hypex or Purifi based for SOTA. Topping has a few good performing products at lower power (although I think there was some design complains or issues with PA5, DYOR, giant thread on that here). Best bang for your dollar in terms of power/price and performance are probably the Hypex mains-powered ncore boards or products using those which are about $500-800 but they are less perfect than the new Purifi or Hypex stuff. Beyond that, $1200-1500 for cheaper/DIY Purifi or Hypex based (400W+ at 4 ohm), and they go up to $3000+ depending on manufacturer or if you want a nicer one. For example the new Hypex kit: https://www.diyclassd.com/product/nilai500diy-250w-stereo-kit/185

 

[sweetchaos]

Amplifier spreadsheet posted above is my go-to-source of an amplifier rankings.

 

[mcdn]

I think your distortion number is a bit generous. 80dB SINAD (so 0.01% THD+N) seems to be a reasonable point to say "good enough for all uses". For power it's very speaker and listening volume dependent, but I don't think anyone could accuse an amp which did 200W per channel into 4 ohms of being underpowered for nearly all uses. Both SINAD and power would need to be in spec from 20-20000Hz.

Add to that list "no other obvious technical flaws", and "comes with decent warranty and service", and "costs less than USD500/channel", and it's hard to look past something like one of the various Hypex NC252MP options out there, such as https://www.buckeyeamp.com/shop/p/hypex-nc252mp-amplifier-2-channel

 

[Head_Unit]

I think your distortion number is a bit generous. 80dB SINAD (so 0.01% THD+N) seems to be a reasonable point to say "good enough for all uses".

Yeah that's an area that I don't know what the magic number is. Actually I dislike SINAD, I'd rather see a raw noise figure (can I hear hiss out of a horn?) and distortion as a separate thing. Somehow if we want to pass 16 bits of audio, then it seems noise should be less than -96 dB. Referred to what, that's a good question. 1 watt output? At maybe 90 dB at a meter for some speakers, that's likely swamped out by ambient noise, maybe only Badwater in Death Valley is that quiet* (with no cars). And yet it gives a bad taste in my mouth to get "only" 80 dB S/N...

Since DACs have been getting excellent noise and distortion performance for pretty cheap, and at least one preamp, I was wondering the power amp status.
--> Seems we are not to anything crazy amazing and cheap yet

*A friend does noise monitoring work in national parks and that was one of the places we went. You could hear the blood in your ears and the pump of your heart.

 

[levimax]

While specs are not SOTA it is doubtful they are an audible issue.... the XTi 2002 has 800 watts per channel and sells for $965 and has DSP

 

[NTK]

There is a practical limit to the maximum output power of an amplifier, which is what the typical wall socket can provide (minus some losses). In the US, it is about 2 kW.

The Hypex NC2000 module reaches that limit (1.6 kW @ 8 ohm). It's idle noise is spec'd at 20 μV, which means, with a speaker of 89 dB SPL @ 1 m @ 2.83 V sensitivity, the noise at 10 cm (4 in) from the speaker is 6 dB SPL — well below the noise floor of most dwellings.

Therefore, a competently executed amp using the Hypex NC2000 module, for almost all practical consumer use cases, has reached perfection (i.e. you can't get more power without custom house wiring, and any higher performance will be unnoticeable). The IcePower 2000AS2 HV is pretty close to that level too.

 

[mcdn]

@Head_Unit, how does this look to you for "good enough for nearly everyone" specs? I've added total output noise in microvolts, clarified the THD spec, and added a few more items:

Must have:

• 80dB SINAD @ 2.83V/4Ω, 10Hz-20kHz (AES17 if digital)
• 200W/channel into 4Ω, 20Hz-20KHz
• Total output noise, perhaps <100uV is good enough? High sensitivity or near-field speakers would want less than that
• Idle power, perhaps <10W total? Or would <5% of continuous rated power be better to allow the high powered amps a chance?
• Output impedance < 10mΩ 20Hz-20kHz
• Input impedance > 20kΩ 20Hz-20kHz
• Crosstalk < 60dB 20Hz-20kHz
• Self protection from output shorting

Nice to have:

• 12V trigger input
• Music signal detection

This is different from @NTK's "literally perfect" spec, which is a bit higher

 

[Head_Unit]

how does this look to you

Well hmm, or should I say HUMMMMMMM if there's too much mains leakage?
80dB SINAD @ 2.83V/4Ω, 10Hz-20kHz (AES17 if digital)
@ I'm not really too fond of this spec, I get it, but I prefer the raw noise separated out from distortion performance. Though to use the charts here it's needful spec.
Total output noise, perhaps <100uV is good enough? High sensitivity or near-field speakers would want less than that
@ That's -86 dB down from 2V = 1 watt into 4 ohms. Should be inaudible really...though I read people complaining about hiss through their horns from amps whose S/N spec seems better. It's a puzzlement.
@ The DISTORTION, really <0.1% might be OK, if the higher distortion comes from low feedback schemes. Those may have transient sonic benefits which are not captured in any testing, especially as testing is all steady-state into resistive loads, a very incomplete picture compared to music into actual loudspeakers.
200W/channel into 4Ω, 20Hz-20KHz
@ Here I personally don't care (am I contradicting myself from a previous post? ). Mostly looking for what does state of the art in clean amplification cost. Because if you did have sensitive horns, 50 or even 20 watts might be enough. I used to drive such to 117 dB in my dorm room with a little 40 watt NAD.
@ To begin to have significantly more power than the biggest AVRs I'd think means more like 300W/4Ω and ideally the amp would significantly increase power into 2 ohms as well.
@ And even more ideally, measure flat on The Power Cube
Idle power, perhaps <10W total? Or would <5% of continuous rated power be better to allow the high powered amps a chance?
@ Good idea. I have no idea about idle power of current Class D designs, and whether it scales with power output.
Output impedance < 10mΩ 20Hz-20kHz
@ 0.010 ohms? I don't know if it is practical to even measure that accurately!
@ I've always wanted this to be low, just the engineer part of my brain I guess. How much it matters below say 0.1 ohms who's to say?
Input impedance > 20kΩ 20Hz-20kHz
Crosstalk < 60dB 20Hz-20kHz
@ That's kind of what I was thinking, though it is nicer if it is super high.
Self protection from output shorting
@ That definitely is a requirement. 12V and music trigger not, though nice I agree.

So on this revised scheme

 

[Head_Unit]

There is a practical limit to the maximum output power of an amplifier, which is what the typical wall socket can provide (minus some losses). In the US, it is about 2 kW.

Yeah, 120Vx15A=1800W, without any derating or allowing for other stuff on the circuit. Like heaters or hair dryers! Also this does not count the efficiency of the amplifier itself (output stage AND power supply*), so you wouldn't even get 1800W continuous.
BUT someone could design an amp with big peak power, like NAD/Proton back in the day, or that ginormous Rockford Fosgate car amp. Why nobody does this, I don't know. Perhaps because really most folks can play loud enough with normal amps or even AVRs. Still you'd think somebody would jump into this niche, 100W continuous but 800W peak or something.

*in automotive the DC/DC conversion is a significant loss.