Is SINAD around 5-15 KHz relevant?

[gags11]

I’m trying to build some active 3-way bookshelves. At this point, I need to decide for amp selection for my high frequency >2.5 KHz. I love Hypex products and will be powering the low end and midrange with hypex modules.

I’m a bit hesitant about the tweeter amplification. Most newer class D amps from Hypex and Purifi are great with power and low distortion, until they get to >10 KHz, where they suck. There are a few clas A/B amps out there that have better SINAD than the new class D amps above 10KHz. Would like some insight, am I overthinking it? Is it possible that some of the high frequency fatigue can be related to distortion from the amp at this frequencies?

Any recs for tweeter amp ( 8 ohm, 93db sensitivity)? My setup will include MiniDsp flex 8 for processing.

Thanks in advance!

 

[MaxwellsEq]

I’m a bit hesitant about the tweeter amplification. Most newer class D amps from Hypex and Purifi are great with power and low distortion, until they get to >10 KHz, where they suck

Do they? Please demonstrate how you have worked this out.

 

[gags11]

Do they? Please demonstrate how you have worked this out.

Here you can see the frequency dependent drop in SINAD.

 

[gags11]

I’m just wondering if there are any relatively affordable amps out there that keep the SINAD above 90db at 5-15 KHz region. The Hypex NCx500 slomst fits the bill, but is costly, and still not state of art for HF amplification.

 

[kemmler3D]

1) 90dB SINAD is a LOT. You will be hard-pressed to hear anything better than -60dB distortion in most cases.

2) Look at the wattages where the SINAD drops and consider that this is going through the tweeter. The 93dB sensitivity tweeter. If you can sit there and listen to >5khz at 113dB SPL for more than 2 seconds, your ears are made of steel. 100w is also the power handling limit of a lot of tweeters, so they will burn up or the VC will jump out of the gap before the distortion gets really bad. To put things in perspective.

Although the graphs don't look very encouraging in those regions, in reality it's not a serious problem because you rarely pump that much juice through a tweeter for any reason.

The energy spectrum in music almost always drops off exponentially with frequency, so at 10-20khz you are typically already 20-50dB below where the bass is... meaning if you're pulling 200w for bass while really cranking the volume, the tweeter will only be drawing something like 1-3 watts. So in practice the tweeter will generally be in the comfortable, low distortion region for both of these amps.

The counter-argument to this would be "what about loud peaks" and it's a valid point, but the energy is still disproportionately concentrated in the bass.

I'm not saying you're wrong for wanting lower distortion across the entire power envelope, but more an explanation as to why this is actually not as bad as it looks. It's a good question.

 

[staticV3]

Here you can see the frequency dependent drop in SINAD.

What this graph doesan't take into account is the spectral distribution of the content that you'll actually listen to:


So when maxing out the NCx500 with 400W into 4Ω, then the typical output power in that moment is 20W at 5kHz, 10W at 10kHz, and 4W at 15kHz:


100+W at those frequencies is a highly unlikely, synthetic test load, making that graph more misleading than educational IMO.

 

[olieb]

I’m just wondering if there are any relatively affordable amps out there that keep the SINAD above 90db at 5-15 KHz region.

Why 90dB? What makes 90 so special, compared to 80dB or 100dB?
And what power are you talking? Up to 5W the Buckeye is more or less noise dominated anyways. Higher power in very high frequency will be quite rare for hifi purposes, maybe it will never happen.
And harmonic distortion of 10kHz is certainly inaudible with 60dB SINAD and even with bat hearing as the tweeter will produce much higher levels of distortion.

 

[LSPhil]

Imagine you have an amplifier with only 25 watts, which corresponds to 14 dB. Now, add these two numbers 93+14, and you get 107 dB. I am certain that at 25 watts and 5 kHz, you will leave the room, even fly out of it from the other side.

Imagine that you have 93 dB; that's already loud. Subtract SINAD 93dB from itself, and you get zero dB. I don't know anyone who can hear zero decibels, not even one or three. Our normal rooms have basic noise levels around plus 30 dB. Add 90 dB to this number.

 

[Salt]

Short answer now: use a hypex plate amp. Won't regret ( we know at least one or two speaker designers that rely on them, too ).

 

[DanielT]

IF the physical size doesn't matter to you, buy some used class A/B amplifier and test. Check what the prices are around where you live, so you can buy and sell at roughly the same price.

To only power up a pair of 93 dB sensitive tweeters, then it is EXTREMELY difficult for me to believe that you can hear any difference between any top of the line SOTA (State-of-the-art) new amplifier (whatever amp,power and price class it is) with and a fully functional used, for example:

Review and Measurements of vintage Yamaha AX-396 integrated amplifier

Review and Measurements of vintage Yamaha AX-396 integrated amplifier This review shows measurements of Yamaha AX-396 integrated amplifier, ser. no. Y399871TV, made in Malaysia. I bought this amplifier in the year 2002 and have been using it shortly. Then it was sleeping in my stock and 5...

audiosciencereview.com

Used Yamaha AX-396 Integrated amplifiers for Sale | HifiShark.com

Used Yamaha AX-396 Integrated amplifiers for sale on 400+ second hand hifi sites & shops. Use Hifi Shark to monitor pricing and global availability

www.hifishark.com

Or:

Review and Measurements of NAD C 320BEE PWR Amplifier

This is a review and detailed measurements of the NAD C 320BEE power (speaker) amplifier. It is on kind loan. The C 320BEE is a vintage product, dating back to 2003 I think. The BEE designation is apparently a reference to Bjorn Erik Edvardsen who was one of their designers. Can't tell if he...

audiosciencereview.com

Used NAD C320BEE Integrated amplifiers for Sale | HifiShark.com

Used NAD C320BEE Integrated amplifiers for sale on 400+ second hand hifi sites & shops. Use Hifi Shark to monitor pricing and global availability

www.hifishark.com

Incidentally, with a 93 dB tweeter, a 30 watt amp would have been more than enough for me.

Tweeters don't need that much power. See here, a project with the Dayton ND25FW. That tweeter has:

Sensitivity

91 dB @ 1W/1m

Power Handling (RMS)	20 watts
Power Handling (max)	40 watts

Introducing Denotiva DIY speakers!

Well was planning to wait until a later date to announce this, but the budget Directiva project required that I exchange the Seas DXT tweeter and so it seemed the time had arrived. To date all Directiva speakers have used the DXT tweeter but as I worked on a budget version I realized that the...

audiosciencereview.com

Dayton Audio - ND25FW-4 1" Soft Dome Neodymium Tweeter with Waveguide 4 Ohm

ND25FW-4 1" Soft Dome Neodymium Tweeter with Waveguide 4 Ohm

www.daytonaudio.com

 

[DanielT]

I’m just wondering if there are any relatively affordable amps out there that keep the SINAD above 90db at 5-15 KHz region. The Hypex NCx500 slomst fits the bill, but is costly, and still not state of art for HF amplification.

Do you know how much distortion your tweeter has? What is the manufacturer and model of your tweeter?

Just so we can put this SINAD above 90db into perspective.

 

[Blumlein 88]

The Benchmark would give the OP what is wanted. I think the Multitone test gives a clearer picture in regards to tweeter usage myself. When not run to super high power class D amps are fine on that.

 

[restorer-john]

100+W at those frequencies is a highly unlikely, synthetic test load, making that graph more misleading than educational IMO.

Not misleading at all.

Amplifiers must be linear devices, which means they treat ALL frequencies with the same level of gain, phase and deliverable voltage and current (power) across their rated bandwidth and rated power.

No excuses.

 

[staticV3]

Not misleading at all.

OP's post proves the opposite.

A layman looks at that graph, sees rising distortion at higher frequencies, and concludes that these Class D Amps distort violins and cymbals more than they do cellos and basslines.

The nature of music spectra and how it relates to Amir's frequency sweep graphs, is important context left out of his reviews.

 

[restorer-john]

OP's post proves the opposite.

A layman looks at that graph, sees rising distortion at higher frequencies, and concludes that these Class D Amps distort violins and cymbals more than they do cellos and basslines.

The nature of music spectra and how it relates to Amir's frequency sweep graphs, is important context left out of his reviews.

OP's post "proves" nothing and you know it.

You have failed to acknowledge the contents of my post in your reply to me. The amplifiers exhibit high frequency non-linearities at high powers. In fact, they cannot deliver their rated power into high frequencies for more than a few seconds in some cases. Amir's graph is deceptive in that it doesn't show how bad the THD at high frequencies really is...

 

[Palladium]

Do you know how much distortion your tweeter has? What is the manufacturer and model of your tweeter?

Just so we can put this SINAD above 90db into perspective.

-50dB to -60dB THD for 5Khz and above @ 90dB SPL from Neumann's own data for the KH120 II.

So, purely academic.

 

[levimax]

For state of the art Class AB amps take a look at Neochrome https://neurochrome.com/collections/power-amplifiers. Mostly DIY but Tom is doing more complete modules these days I believe, you can contact him and he can let you know about complete assembly.

Not that it is going to make any audible difference but these AB amps are very well behaved at all frequencies and their measurements speak for themselves.

 

[staticV3]

You have failed to acknowledge the contents of my post in your reply to me.

True, I refuse to engage with your unfounded agenda of trashing everything that's not Class AB.

It's become a community meme at this point, not without reason.

An Amplifier needs to amplify the content that people want to amplify with low distortion and high stability.

Being able to output a continuous 400W 20kHz sine wave with low THD may be a vital criterium in some specialized lab environment, but it has absolutely nothing to do with HiFi consumer audio.

Times have changed.

Today's best HiFi Amps are designed for efficiency and actual music instead of outdated lab tests like continuous wideband output power.

 

[restorer-john]

True, I refuse to engage with your unfounded agenda of trashing everything that's not Class AB.

An Amplifier needs to amplify the content that people want to amplify with low distortion and high stability.

Being able to output a continuous 400W 20kHz sine wave with low THD may be a vital criterium in some specialized lab environment, but it has absolutely nothing to do with HiFi consumer audio.

The FTC has ruled on this and they side with me and 548 other people who bothered to comment or provide submissions to their requests.

The updated Rule requires any and all amplifiers being advertised for sale in the US from next month (August) who wish to advertise a power output of any value above 2W to rate them in accordance with the regulations or suffer the consequences.

Those regulations provide for THD+N from 250mW to rated power and across the entire audio bandwidth (20Hz-20kHz). So, whether you think HF THD+N is relevant or not doesn't matter one iota. Class D amplifiers will have to be advertised and rated the same way as all other amplifiers. They will have to be honest for once. If their amplifiers descend into noise at the low end- bad luck. If their THD+N skyrockets at high frequencies, bad luck- rate them properly.

And there's plenty of other classes, brands and designs of amplifiers that have been rorting consumers through false and misleading specifications. It's not just Class D, but a few major players have been egregious in this regard.

 

[DWPress]

I'm looking for the same thing for my 3 way XO. I was using a Topping PA5 for a few years but it recently died (surprise!). Currently I'm using a 4ch Hypex NC252MP for tweeters and midrange because it's conveniently in the rack but looking to use something more appropriate + then I can bridge the 252 and use it for another passive sub.

One thing to keep in mind is you'll probably be padding down your tweeter to match sensitivity to the mid and bass drivers so the 93dB spec won't be as important in the end, mine are 93dB sensitivity too but need to be reduced 9dB to match at LP. I figure 50 watts for my needs but my speakers are probably at least 5x bigger than what you're building and can reach deadly SPL.