Buckeye Amps: 2nd Generation Purifi EIGENTAKT 1ET9040BA Amplifier

[RickS]

@Buckeye Amps can correct me, but as I recall his buffer board is based on the Purifi 9040 reference design. Gains are a little different to allow matching between his other offerings, but now that he has an AP, can be pretty much assured that the performance is very close (if not identical) to the reference design.

That said even once Amir gets these amps, given his backlog, may be quite a while before you see his measurements. I like having Amir's verification too, but the risk is pretty low in comparison to risk of long lead times on the amp parts. Of course, is all a matter of one's own need to get a new amplifier. I wanted a subwoofer amp with auto on/off to replace the previous NCx500 monoblock that lacked it. I have not heard any difference but am pleased that I can turn on the system remotely again.

 

[Buckeye Amps]

That's the thing with analyzers. You can be lucky and get one way better than spec, or not. But at the price of the AP, I wouldn't be happy.

Is it down to one channel being better than the other, the A/Ds itself/themselves or the high performance sine gen with the H2 issue? If it's the gen, you may be able to better it from a modern SOTA D/A converter.

Sorry for the late reply. We wanted to do some further testing on the rig.

It certainly does appear to be related to the generator on the AP side of things...when using an E1DA 9039S DAC as the generator, the baseline distortion goes away.

 

[Buckeye Amps]

Ouch... it's a bit disturbing that a manufacture of expensive analyzers can't be bothered to verify compliance of product before selling it to important manufacturers like Buckeye Amps. Who tests the test gear? Without consistent results across analyzer units comparing reports from various sources will be difficult.

From what we were told by AP is it is within their tolerance limit.

From a sensible position, I get that we are talking about ultra low distortion/noise floor levels and nitpicking otherwise inaudible anomalies/variances...even at such a high price point of the AP.

However, now that amplifiers like the 9040 are pushing into the same ultra low floor as something like the AP, even a small anomaly in the test setup can be enough to make people question our designs (as evidence above a few posts). Which, again, is inaudible but in the new race of who can engineer the best design, it does have an effect on perception.

 

[tmtomh]

From what we were told by AP is it is within their tolerance limit.

From a sensible position, I get that we are talking about ultra low distortion/noise floor levels and nitpicking otherwise inaudible anomalies/variances...even at such a high price point of the AP.

However, now that amplifiers like the 9040 are pushing into the same ultra low floor as something like the AP, even a small anomaly in the test setup can be enough to make people question our designs (as evidence above a few posts). Which, again, is inaudible but in the new race of who can engineer the best design, it does have an effect on perception.

I'm all for the general principle of going for excellence in measured performance, and I'm all for continuing to pursue new levels of excellence to a point somewhat beyond human audibility - enough for a reasonable buffer or "wiggle room" to account for cumulative effects in a multi-component signal chain, unit-to-unit variation, demanding use cases (like nearfield listening and/or sensitive speaker or sensitive headphone listening, and so on).

But with that said, if two different models of a consumer component, and ones relatively affordable at that, are achieving measured performance so good that we can't reliably distinguish the two based on their measurements because the production tolerances of the gold-standard professional measurement gear itself are equal to or greater than the differences between the components, then IMHO that should make us take a step back and ask ourselves if it's worth focusing on those measurements beyond the level of excellence already achieved.

 

[DonH56]

From what we were told by AP is it is within their tolerance limit.

From a sensible position, I get that we are talking about ultra low distortion/noise floor levels and nitpicking otherwise inaudible anomalies/variances...even at such a high price point of the AP.

However, now that amplifiers like the 9040 are pushing into the same ultra low floor as something like the AP, even a small anomaly in the test setup can be enough to make people question our designs (as evidence above a few posts). Which, again, is inaudible but in the new race of who can engineer the best design, it does have an effect on perception.

We're back to the 1970's and 1980's when the race for lowest distortion numbers was king even though inaudible. Actually it's a lot worse today when -120 dBFS spurs are somehow unacceptable. I usually try to draw a mental line of loudspeaker or headphone distortion and the dynamic range of the source material to remind myself how far we've come -- and how little some of this actually matters in the real world.

I (and others) killed myself to achieve 160 dB dynamic range from a radar system, but I just cannot see needing that in my audio system.

 

[shion_ca]

Sorry for the late reply. We wanted to do some further testing on the rig.

It certainly does appear to be related to the generator on the AP side of things...when using an E1DA 9039S DAC as the generator, the baseline distortion goes away.

new measurement post then with improved source!

 

[Buckeye Amps]

new measurement post then with improved source!

Not enough voltage with that source to fully drive all gain stages. Still working with AP and will go from there.

 

[Buckeye Amps]

For a 5 watt SINAD plot? You only need about 240mV at the 25.5dB gain setting.

Correct but I said all gain stages. In regards to it wouldn't be a full proof fix in the face of the AP generator being lacking.

 

[Buckeye Amps]

Do you mean gain settings? For a 5 watt 4 ohm SINAD test input voltages are about
15.5dB = 750mV, 20.5dB = 422mV, 25.5dB = 240mV
Are you saying the E1DA is not capable of outputting 750mV?

Sorry for the confusion:
Yes we can do the 5w SINAD tests for all gain stages. But we wouldn't be able to do all the power sweeps.

What I was trying to convey is that this particular DAC workaround wouldn't be a full proof solution to the underlying problem going forward, in that the AP generator itself has a distortion issue. Ideally would like to get that fixed/AP to rectify it.

 

[Buckeye Amps]

Understood. I dug up one of Amirs loopbacks. The harmonics are lower, but I wouldn't say yours has an H2 problem. There is little difference in SINAD. Yours still looks in spec.

Is the E1DA really better than this?
Just noticed the blue channel is on par with Amirs. Can you use that?

There is a 10dB difference for H2 between his loopback and ours.

Again, I know in the overall grand scheme of consumer usage and audibility, it is a moot point really. But because of the nature of how each new iteration of Hypex and Purifi designs are trying to achieve engineering excellence at ultra low levels of distortion/noise floor, every little bit matters.

The mere fact we have been having this discussion around our 9040 measurements is evidence to that...

 

[Matias]

Plus if you pay 30k on an APx555, it better have harmonics below 200 dB!

 

[DonH56]

There is a 10dB difference for H2 between his loopback and ours.

Again, I know in the overall grand scheme of consumer usage and audibility, it is a moot point really. But because of the nature of how each new iteration of Hypex and Purifi designs are trying to achieve engineering excellence at ultra low levels of distortion/noise floor, every little bit matters.

The mere fact we have been having this discussion around our 9040 measurements is evidence to that...

The problem as I see it is two-fold:

1. We have two measurements systems both of which may measure the same device and yet exhibit a 10 dB (or so) difference in distortion. That will inevitably lead to questioning the measurements, particularly the worse one, and people pointing fingers at one or the other result.

2. The facts that the analyzer's distortion is within spec, and well below being inaudible, are lost on those to whom graphs and numbers determine the best product.

It would be great if AP's unit-to-unit variance were less. Given it is within their spec, then there is little incentive for them to "fix" Dylan's unit, although perhaps showing Amir vs. Dylan results might help. They might be interested in knowing (if they do not already) how such a relatively large discrepancy happens, and in the long run producing analyzers with improved performance should help their sales.

 

[Salt]

Plus if you pay 30k on an APx555, it better have harmonics below 200 dB!

To be defined by what measurement gear ...?

 

[slm0x]

amirm:
I listen loud enough that my 1000 watt amplifiers completely run out of power! Remember, the first few watts incrementally get quite a bit louder. But doubling from 500 to 1000 watts is nothing. The curve flattens quickly.....

Doubling in power is an extra 3db, correct. Each 3db increase takes double the power and equates to double the volume. If you know the sensitivity of your speakers (for instance 87db speakers put out 87db with 1 watt of power), with a simple decibel meter (which can be downloaded on your phone - don't know how accurate the phone version is), you can estimate the number of watts you're listening to.

When my 87db RBH bookshelf speakers are playing at 102db, I'm utilizing 32 watts.

BTW, I purchased a Hypex NCx500 stereo amp from Dylan at Buckeye, it's incredible. I've had it about 3 weeks, and I like it more every day. I'm not sure it isn't better than my $8,000 Mcintosh amp, and I love my Mac.

 

[pogo]

you can estimate the number of watts you're listening to.

You can't!
The sensitivity is only determined over a limited frequency range. In addition, the impedance below this range is not taken into account.

 

[slm0x]

YMMV.

A lot of variables come into play within someone's setup that could determine if and why they hear a difference between two amps that both have excellent performance metrics.

This is why customers are usually surprised at my recommendations based off their setups and normal usage when they ask me for advice between all the different amps.

I was one of these guys. I wanted a Purifi setup, 1ET9040BA monoblocks. Dylan talked me into a Hypex NCx500 stereo setup, even though it's approximately 1/2 price of the Purifi monoblocks. I'm thrilled with the NCx500 stereo setup. Dylan's recommendations are with the customer in mind, even if what is best for the customer isn't best for Dylan's pocketbook.

While I'm sure the Purifi offerings are great, I'm amazed by the sound of the NCx500 amp. It's an incredible piece of audio gear.

 

[Julf]

While I'm sure the Purifi offerings are great, I'm amazed by the sound of the NCx500 amp. It's an incredible piece of audio gear.

They are both at the level where your speakers, listening room and above all source material is the limiting factor, not the amp.

 

[DonH56]

Doubling in power is an extra 3db, correct. Each 3db increase takes double the power and equates to double the volume. If you know the sensitivity of your speakers (for instance 87db speakers put out 87db with 1 watt of power), with a simple decibel meter (which can be downloaded on your phone - don't know how accurate the phone version is), you can estimate the number of watts you're listening to.

When my 87db RBH bookshelf speakers are playing at 102db, I'm utilizing 32 watts.

BTW, I purchased a Hypex NCx500 stereo amp from Dylan at Buckeye, it's incredible. I've had it about 3 weeks, and I like it more every day. I'm not sure it isn't better than my $8,000 Mcintosh amp, and I love my Mac.

That is incorrect. Or rather, 3 dB is a doubling of power, but 3 dB change in the midrange is not a large change in volume. It takes 10 dB to double (or halve) the perceived loudness (volume). The curves compress somewhat for low frequencies (bass) meaning <10 dB for doubling the perceived loudness, but our hearing is less sensitive down there, so you need even more power anyway.

This has been discussed many, many times on ASR (and elsewhere). Here is my normal post in the subject, from a few decades ago:

Power Needs:

For many years I have used 17 dB peak-to-average power based on an old AES article I can no longer find. Anecdotally various audio fora report 20 dB or more for movies. IME most people use much less average power than they think, but require more peak power. An online calculator can help you estimate your average power; note 80 dB is very loud to me (YMMV). Here is one:

Peak SPL Calculator

Most people think of volume in dB and most modern AVR/AVP/etc. units list dB on the volume knob. Power in dB goes as 10log10(Power_ratio) so the change in power is 10^(dB/10). Here are some reference numbers in dB and power:

1 dB is barely noticeable and requires 1.26x the power
3 dB is what most people hear as "a little louder" and requires 2x the power
6 dB is significantly louder and requires 4x the power
10 dB sounds twice as loud and requires 10x the power
17 dB is the headroom for music and requires 50x the average power
20 dB for movies requires 100x the power

If you listen at around 1 W average, then you need 50~100 W to avoid clipping on most source material. You can figure out your estimated average power from the calculator knowing your speaker's sensitivity and distance from them. Note music may be more compressed and thus require less headroom, and the loudest sounds in movies tend to be things like gun shots and explosions where a little clipping is likely unnoticeable.

HTH - Don

 

[slm0x]

That is 2.3 dB. Not a huge audible difference.

They are both at the level where your speakers, listening room and above all source material is the limiting factor, not the amp.

The NCx500 certainly sounds great, that's for sure. Class D amps have come a LOOOOOOOONG ways.

 

[tmtomh]

So when does the first Hypex or Purifi-designed and produced precision analyzer come out?