Review and Measurements of Behringer A500 Amplifier

[Ron Texas]

I set the switch and remeasured. Not much changed since I was testing the outputs as stereo as opposed to bridged. I updated the dashboard, power and text to reflect the new findings ion the review.

Thank you. Could you explain the difference between the SINAD which is in the high 60's and signal to noise ratio which is in the 90's? What is typical for a class AB amp?

 

[amirm]

@amirm I think it's very important when measuring these amps to take a THD vs frequency sweep if possible?

I am truncating these measurements because of the poor performance. When we find better ones I will provide more measurements. For now, here is the THD vs frequency for this one:

Those sudden peaks with "T" symbols are due to mains interference and the analyzer giving up on trying to get a stable reading.

Looking to the right, we see the typical rise with frequency due to reduced feedback gain.

 

[andreasmaaan]

I am truncating these measurements because of the poor performance. When we find better ones I will provide more measurements. For now, here is the THD vs frequency for this one:

View attachment 17178

Those sudden peaks with "T" symbols are due to mains interference and the analyzer giving up on trying to get a stable reading.

Looking to the right, we see the typical rise with frequency due to reduced feedback gain.

Thanks

 

[amirm]

Thank you. Could you explain the difference between the SINAD which is in the high 60's and signal to noise ratio which is in the 90's? What is typical for a class AB amp?

SINAD includes distortion but SNR does not (both include noise). That is why I like SINAD better than SNR.

I don't have a typical number to offer. As I have been searching, it seems that the typical values have declined substantially in recent years.

As far as I am concerned, we should be able to get one more decimal place of performance as a minimum (0.003% versus 0.03% for example).

 

[amirm]

Is it also not important to measure the distortion with varying impedance?

That is a TODO project (building a load that mimics a real speaker). For now, I only have high-power dummy loads at 2 and 4 ohms (I am using two 2 ohm ones in series). I have other dummy loads but they are not as high power.

 

[Ron Texas]

I would love to see a high performing amplifier tested so there is a benchmark. It doesn't have to be the best in the world, but it needs to be excellent. After that a Crown XLS 1502.

 

[amirm]

I would love to see a high performing amplifier tested so there is a benchmark. It doesn't have to be the best in the world, but it needs to be excellent. After that a Crown XLS 1502.

I am on the hunt for some and one good candidate should be coming in a month or so.

 

[sergeauckland]

SINAD includes distortion but SNR does not (both include noise). That is why I like SINAD better than SNR.

I don't have a typical number to offer. As I have been searching, it seems that the typical values have declined substantially in recent years.

As far as I am concerned, we should be able to get one more decimal place of performance as a minimum (0.003% versus 0.03% for example).

I don't disagree that a figure that low is both achievable, and indicates generally good design. However, all things being equal, it's completely unnecessary for transparency, which is all we should, I think, be concerned with. The 0.02% THD+N I measured on mine (8ohms, 1kHz) I consider perfectly adequate, so no reason to look for better. The very slight droop at 20kHz might just be audible to an 18 year old, but certainly not to me. Mains derived buzz was just audible in my active system, into the 92dB/W mid-range driver, which I solved by putting a 10,000uF cap across each of the existing power supply reservoir capacitors. With passive crossover loudspeakers of more normal sensitivity, the noise isn't likely to be a problem. The A500 is far from state of the art, but even disregarding the low price, it doesn't need to be.

S

 

[amirm]

How hard can it be to make a power amplifier with a SINAD of > 80 dB?

It should not be hard. SINAD of 80 is just 0.01% distortion. Here is Doug Self on his "blameless amplifier" measurement:

At 1 kHz, his design is getting 0.001% or SIAND of 100 db.

 

[amirm]

The 0.02% THD+N I measured on mine (8ohms, 1kHz) I consider perfectly adequate, so no reason to look for better.

Correlation between THD and audibility is weak. As such, one cannot make statements about transparency unless we push the THD so low where its spectrum no longer matters. You simply are not there with such THDs.

And regardless, the purpose of these reviews is to find equipment that performs far better than others at similar prices. In other words, there should not be cost to better fidelity. At least not a lot. Once there, I don't see a good reason for settling for less.

 

[Killingbeans]

I am on the hunt for some and one good candidate should be coming in a month or so.

I'd just like to see a Benchmark AHB2 and a set of Hypex NC400 monoblock kits butt heads

 

[restorer-john]

At the end of the day, it is a very cheap amplifier. There is no point holding it up to high expectations and then lambasting for not being SOTA.

For $199, it's a steal.

At 1 kHz, his design is getting 0.001% or SIAND of 100 db.

Amplifiers are rated from 20-20KHz and from 250mW to rated power at 4 and 8 ohms. Notice his graph is just a spot power (30W@8ohms).

 

[amirm]

Notice his graph is just a spot power (30W@8ohms).

So is my dashboard SINAD measurement.

 

[Ron Texas]

Nothing disinfects like sunlight. Based on a quote by Louis Brandeis.

 

[restorer-john]

So is my dashboard SINAD measurement.

And at what random output level is it today, Amir?

You seem to now be testing at 5.228W (or is that the wrong symbol you picked) for SINAD/THD+N when the other day it was 5.0V (6.25W). What will it be tomorrow?

Self clearly has picked full power at the bottom of the THD curve, just before the knee. Every power below it and above it will be a hell of a lot worse won't it?

Unless you have a consistent number (1W@8ohms is the standard and full rated power) none of these 'tests' can be compared against each other. This is the same issue I keep raising and you keep ignoring when comparing SINADs of D/As with wildly differing output levels. You can't compare in a simple ranking based on the number.

Level the playing field.

 

[sergeauckland]

When I test amplifiers, I test first at full power, i.e. just below visible clipping on the 'scope, then at 1W and 100mW, and check that distortion is still low, and any increase in the number is purely a function of noise and not anything else like crossover distortion. One disadvantage of measuring distortion using a spectrum display is that one doesn't see what the residual looks like, which can show up crossover distortion very obviously. Equally, a THD meter won't show the individual harmonics or power supply artefacts like a spectrum, so both are really needed to characterise an amplifier fully. Depending on how deep I want to go, I also measure at different frequencies and loads, but if I'm just checking an amp after repair, then 1kHz & 10kHz 8ohm at full power will generally tell me if it's OK.

S

 

[March Audio]

I'm also surprised at trying to judge a power amplifier by DAC criteria. A power amplifier with a SINAD of -60dB at any frequency and any permitted load is transparent, and that's all that's required for fidelity. There's no benefit except specmanship to do more. In those terms, the A500 is adequate for the task. That other amplifiers can do better at higher cost is besides the point.

S

Whilst I sort of agree, this is very cheap, it is also very poor performance. You can acheive much, much better performance for not much more cost. I don't agree with 60 dB SINAD being good enough.

We aren't going to definitively answer the question of what's good enough performance, but this ain't it .

 

[March Audio]

When I test amplifiers, I test first at full power, i.e. just below visible clipping on the 'scope, then at 1W and 100mW, and check that distortion is still low, and any increase in the number is purely a function of noise and not anything else like crossover distortion. One disadvantage of measuring distortion using a spectrum display is that one doesn't see what the residual looks like, which can show up crossover distortion very obviously. Equally, a THD meter won't show the individual harmonics or power supply artefacts like a spectrum, so both are really needed to characterise an amplifier fully. Depending on how deep I want to go, I also measure at different frequencies and loads, but if I'm just checking an amp after repair, then 1kHz & 10kHz 8ohm at full power will generally tell me if it's OK.

S

Surely crossover distortion will show as harmonics? There will also be obvious distortion in an an Fft before you see it on a scope.

 

[restorer-john]

When I test amplifiers, I test first at full power, i.e. just below visible clipping on the 'scope, then at 1W and 100mW, and check that distortion is still low, and any increase in the number is purely a function of noise and not anything else like crossover distortion. One disadvantage of measuring distortion using a spectrum display is that one doesn't see what the residual looks like, which can show up crossover distortion very obviously. Equally, a THD meter won't show the individual harmonics or power supply artefacts like a spectrum, so both are really needed to characterise an amplifier fully. Depending on how deep I want to go, I also measure at different frequencies and loads, but if I'm just checking an amp after repair, then 1kHz & 10kHz 8ohm at full power will generally tell me if it's OK.

Exactly.

I want a residual noise number in uV with no signal as that is the resolution limiting factor in amplifiers.

The output of standalone THD meters will give you a nice representation of the distortion and with experience, that waveform tells a lot about the characteristics of said distortion.

Again, this has been the standard for many decades- a waveform plot with an (not to scale) aligned and below representation of the distortion component. Crossover distortion is very obvious in that situation whereas on an FFT, not so much.

 

[sergeauckland]

Surely crossover distortion will show as harmonics? There will also be obvious distortion in an an Fft before you see it on a scope.

The distortion residual output of a THD analyser visible on a 'scope triggered from the driving signal shows very clearly where the distortion occurs in the waveform, and how much of the total % is hum and noise. The FFT will show a slew of harmonics, but won't necessarily distinguish crossover distortion from just upper harmonics. I think it important to do both, and certainly FFT is far far easier than the old picking out individual harmonics with a tuned filter as I had to do in the late 1960s and early 70s. That really was tedious!
S.