AVS Forum banner

1881 - 1900 of 2731 Posts

·
Registered
Joined
·
1,046 Posts
I completely agree that the Arcam unit should meet its 100 THD+N spec. It did not, at least with the unit tested and the way it was tested.
I disagree that 4V is something consumer electronics should shoot for or that it has anything to do with engineering excellence. Putting out 4 volts is essentially useless as you would completely overdrive any power amp you have.

Just so I understand the measurement graphic


I am using Monolith 7X amplifier which has 1.6V input sensitivity to full power.

So the SINAD I can hope to get before clipping my power amp is about 80. Am I reading it right?


Also AV40 has a nominal output of 2V RMS using XLR - so the SINAD in nominal use is about 80-85 - is that right?

Sorry if these are dumb questions just trying to interpret the test results and my equipment

Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
5,172 Posts

·
Registered
Joined
·
2,285 Posts
I completely agree that the Arcam unit should meet its 100 THD+N spec. It did not, at least with the unit tested and the way it was tested.
I disagree that 4V is something consumer electronics should shoot for or that it has anything to do with engineering excellence. (1) Putting out 4 volts is essentially useless as you would completely overdrive any power amp you have.

It's never clear why people defend mediocre performance. This isn't a psychological forum, but it is an interesting question.

I sent a piece of gear to Amir to test. I felt it would have mediocre measurements when I sent the piece of equipment and it did. I still use that piece of equipment, but within its limitations, which I understood before I purchase the item.

(1) I use Benchmark AHB2 power amplifiers. Measurements of these outstanding amplifiers are widely available. The AHB2's have selectable input sensitivities, but perform best with inputs over 4V. High quality preamps normally perform best at higher outputs as well.

It will be interesting to see if Amir's tests have an effect on manufacturers such that 4V (XLR) outputs improve in performance. It would be very easy to improve the performance of most of the gear that Amir tests that shows mediocre performance, usually at low cost, if the manufacturers were interested.

AKM DAC IC's have a voltage output that is very clean at 2V single ended. ESS DAC IC's, after the I/V converters (hopefully JBL didn't cheap out so much that the DAC IC is used in voltage output mode, which costs 6 to 8dB in S/N) are typically the same. Use good board layouts with a lot of layers, a reasonable power supply design, low distortion opamps in the output filters, Cirrus Logic CS3318 volume controls, and SE to balanced conversion to get 4V from the XLR's at 0dB gain on the volume control and you're done.

The fact that better performance would be easy to obtain is frustrating to Amir and others who have insight into these designs. In these higher priced components buyers too often aren't getting what they are paying for. That these buyers at times defend the equipment is curious, unfortunate, … what is the word?
 

·
Registered
Joined
·
64 Posts
Just so I understand the measurement graphic


I am using Monolith 7X amplifier which has 1.6V input sensitivity to full power.

So the SINAD I can hope to get before clipping my power amp is about 80. Am I reading it right?


Also AV40 has a nominal output of 2V RMS using XLR - so the SINAD in nominal use is about 80-85 - is that right?

Sorry if these are dumb questions just trying to interpret the test results and my equipment

Sent from my iPhone using Tapatalk
It looks more like about 85 to 86dB or so to me, eyeballing the chart. But remember, the dominating factor for this unit is the 2nd harmonic distortion. The noise floor is way below this. Based on his first measurements, the noise is maybe 20dB below that, so the SNR (as opposed to the THD+N) would likely be more like 105-110 dB. I'm interpreting (guessing) since he did not show the full spectrum graph at 1.6 volts.
 

·
Registered
Joined
·
12,058 Posts
Why would distortion increase with duration?
NOAH..
Because if the duration is very short, like a few microseconds the distortion will likely not be audible. Also depending upon circuit topology a component may do just fine handling a short signal burst but not extreme. Think about just like a power amplifier, typically the amplifier or preamplifier can handle a quick dynamic peak but not continuous... Dynamic peak power was not a problem in the analog, tube days as dynamic range was limited to 100dB are commonplace. Prior to the introduction of CDs in the early 70s/80s loudspeakers and amplifiers did just fine but quickly showed their weaknesses when higher dynamic range sources were available. They had to be redesigned accordingly for faster switching output devices for amplifiers, newer, improved, lighter speaker driver materials many from NASA, all supported along with better test methods and equipment. But now the discussion gets more complicated as slew rate, rise time become crucial factors..

Just my $0.02.. ;)
 

·
Oppo Beta Group
Joined
·
11,496 Posts
It's never clear why people defend mediocre performance. This isn't a psychological forum, but it is an interesting question.

I sent a piece of gear to Amir to test. I felt it would have mediocre measurements when I sent the piece of equipment and it did. I still use that piece of equipment, but within its limitations, which I understood before I purchase the item.

(1) I use Benchmark AHB2 power amplifiers. Measurements of these outstanding amplifiers are widely available. The AHB2's have selectable input sensitivities, but perform best with inputs over 4V. High quality preamps normally perform best at higher outputs as well.

It will be interesting to see if Amir's tests have an effect on manufacturers such that 4V (XLR) outputs improve in performance. It would be very easy to improve the performance of most of the gear that Amir tests that shows mediocre performance, usually at low cost, if the manufacturers were interested.

AKM DAC IC's have a voltage output that is very clean at 2V single ended. ESS DAC IC's, after the I/V converters (hopefully JBL didn't cheap out so much that the DAC IC is used in voltage output mode, which costs 6 to 8dB in S/N) are typically the same. Use good board layouts with a lot of layers, a reasonable power supply design, low distortion opamps in the output filters, Cirrus Logic CS3318 volume controls, and SE to balanced conversion to get 4V from the XLR's at 0dB gain on the volume control and you're done.

The fact that better performance would be easy to obtain is frustrating to Amir and others who have insight into these designs. In these higher priced components buyers too often aren't getting what they are paying for. That these buyers at times defend the equipment is curious, unfortunate, … what is the word?
Apparently, it did occur to manufacturers of some high-end processors and AVRs that someone would actually measure the preamp section since that has not happened in years.
At least now, those of us with adjustable gain amps can select the optimal setting.

- Rich
 

·
Registered
Joined
·
13,143 Posts
At least now, those of us with adjustable gain amps can select the optimal setting.
Too bad virtually no consumer amp offers such an option.
 

·
Registered
Joined
·
1,046 Posts
It looks more like about 85 to 86dB or so to me, eyeballing the chart. But remember, the dominating factor for this unit is the 2nd harmonic distortion. The noise floor is way below this. Based on his first measurements, the noise is maybe 20dB below that, so the SNR (as opposed to the THD+N) would likely be more like 105-110 dB. I'm interpreting (guessing) since he did not show the full spectrum graph at 1.6 volts.

Thanks and pardon my naïveté - is it good or bad that the SNR is 105-110 and SINAD is 80-85.

What real world implications are there?


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
2,663 Posts
So I think it was alluded to earlier, but if I use Dante amps with this, that supposedly bypasses this "issue"?
Depends if you are listening to audio with HDCP protection or not.
 

·
Registered
Joined
·
1,187 Posts
So if Dante can't pass a 24-bit signal and has to downgrade to 16-bit (As mentioned previously), is that a better trade off to get less distortion (assuming that the amp is better than the processor)?



I would really like to see Amir measure a SDP-55 to really see if there is any difference to the AV-40.
Can't really see Dante + L16 + a slightly better Dac being the whole cost difference, but those who actually know haven't really detailed the differences that I have seen.
 

·
Registered
Joined
·
1,046 Posts
So if Dante can't pass a 24-bit signal and has to downgrade to 16-bit (As mentioned previously), is that a better trade off to get less distortion (assuming that the amp is better than the processor)?



I would really like to see Amir measure a SDP-55 to really see if there is any difference to the AV-40.
Can't really see Dante + L16 + a slightly better Dac being the whole cost difference, but those who actually know haven't really detailed the differences that I have seen.

Don’t forget all the Dirac upgrades are included with JBL


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
3,661 Posts
So if Dante can't pass a 24-bit signal and has to downgrade to 16-bit (As mentioned previously), is that a better trade off to get less distortion (assuming that the amp is better than the processor)?



I would really like to see Amir measure a SDP-55 to really see if there is any difference to the AV-40.
Can't really see Dante + L16 + a slightly better Dac being the whole cost difference, but those who actually know haven't really detailed the differences that I have seen.
I would like to see this too and the Audiocontrol X9 to see how the biggest of the family performs. It seems the 9028 DAC is a different pinout so the DAC PCB at least on the SDP-55 must be different to the AV40; the JBL specs are nearly identical but maybe the JBL specs are more honest or Arcam accidentally used the JBL specs :D

BTW: Amir has posted some follow-up info in the AV40 review thread.
 

·
Registered
Joined
·
64 Posts
Thanks and pardon my naïveté - is it good or bad that the SNR is 105-110 and SINAD is 80-85.

What real world implications are there?


Sent from my iPhone using Tapatalk
Well this is going to be a little subjective and others are likely to have a different opinion. An SNR of 105 to 110 is excellent. You are unlikely to hear noise in real world applications.
The 80 to 85dB SINAD , for this unit is caused (mostly) by the 2nd harmonic distortion spike.
While not a great measurement, a 2nd harmonic distortion that is that far down is unlikely to be audible with music.
And remember, speakers have on the order of 10-100 times this amount and no one complains about hearing distortion in speakers nor has it shown itself to be audible in double blind listening tests with music (at least that I'm aware of).
Let me quote Floyd Toole again (He has decades of research to back up his opinion): "Traditional harmonic or intermodulation distortion measurements are almost meaningless".
IMHO the subjective sound quality is not going to be impaired in the least by these particular numbers.
 

·
Registered
Joined
·
1,187 Posts
Don’t forget all the Dirac upgrades are included with JBL
Sent from my iPhone using Tapatalk

Oh yeah great point that I forgot, and realistically that could be the price difference along with the others + additional JBL brand markup!


Also a question I have:

Is distortion cumulative?
For example:

I have an analog source with x distortion going to an AVR doing AD/DA going to an amp that has a DSP so another AD/DA.
Does each step add/amplify the distortion at the speaker or is distortion lower than the worst link just not impactful at all?
If so is it cumulative or multiplicative?


Thanks!
 

·
Registered
Joined
·
1,187 Posts
BTW: Amir has posted some follow-up info in the AV40 review thread.

Do you have a link to it or a name posted under?

Just searched for Amir and didn't come back with anything.


Thanks good sir!
 
1881 - 1900 of 2731 Posts
Top