Check out this post for test setup and limitations.

ITU-R IMD is one way to measure intermodulation distortion. I am feeding the circuit with two test tones at -6 dBV, which will sum to 0 dBV. The test tones at 19 kHz and 20 kHz will produce a modulated signal at 1 kHz. The difference between the test tone and the modulated signal is the ITU-R IMD.

The ITU-R IMD for Burson V6 Classic is -118.1 dB / 0.000124%.

Click here for a high resolution image of the measurement.

How does it fare in comparison to the audio analyzer itself and, equally interesting, our OPA2134 reference IC op amp?

Well, the audio analyzer can measure down to -122.2 dB / 0.000078%, so the Burson measurements are well within the analyzer's capabilities.

Click here for a high resolution image of the measurement.

The Burson V6 Classic loses out to the OPA2134 once again. The OPA2134 measures -121.0 dB / 0.000089%.

Click here for a high resolution image of the measurement.

Following up on the THD measurement of the Burson V6 Classic (which also shows the test setup and limitations), here's its frequency response.

These measurements are showing up now because the QA401 is not linear. The measurements had to be compensated for that.

The frequency response is measured with an output of 0 dBV / 1V rms and a gain of 1.36 dB / 1.17X.

The Burson V6 Classic's gain is 0.1 dB lower than expected. It performs great otherwise, showing -0.2/-0.1 dB all the way from 10 Hz to 80 kHz. This is as high as I can go with 192 kHz sampling frequency. The graph uses a relative scale since it would otherwise disappear under the 0 dB horizontal line.

Click here for a high resolution image of the measurements.

But how high does it go? Time to whip out the Rigol DG1022 function generator and the Rigol DS1054Z oscilloscope. ~1.5 MHz is the answer (with the noise starting at about -20 dB below the signal). Go beyond that and the amplification still holds for a while, but the signal is no longer a sinus.

The signal is down -3 dB at approx. 3.7 MHz. By then it's a pure saw tooth signal.

Background

To see if the Burson V6 operational amplifiers are viable in the Muffsy Phono Preamp, I went ahead and bought the dual version of both the V6 Classic and V6 Vivid (*). They don't come cheap, about 20x the price of an OPA2134. Getting two of these instead of 40 OPAs for my kit stock was quite a difficult decision...

The op amps from Burson were delivered with cable ties. I have not been able to figure out why... ;)

I have both feet firmly planted in the objective side of things, so I decided to get the Burson's to do some measurements. I did listen to them in the gain stage of my headphone amp, comparing them with both NJM2068 and LM4562. They sounded pretty damn good (as in more or less indistinguishable from the others) in that scenario (both 2.5 and 6.5X gain), so I really don't feel fooled or anything. This seems to be a proper product.

JFET inputs are best suited for moving magnet cartridges. The Bursons do have JFET inputs. I would like my op amps to be able to run on +/-15 volts. The Bursons handle up to +/- 16.5 volts. The basics are covered, let's see what they're capable of.

(*) Yes, I did buy them myself. They were not given to me by Burson.

Preparations

In my quest to measure the Burson V6 discrete operational amplifiers, I've established the measurement baseline and used the OPA2134 as reference.

I'll start with the Burson V6 Classic, and I'll have to visit the Burson V6 Vivid later. The V6 Vivid that I received appears to be defective. Burson kindly agreed to send me a new one.

I have shown that I can measure THD down to -118.7 dB / 0.00012%, Here's how the Burson V6 Classic fares.

The op amp is placed in a test rig with 1.36 dB gain. There is no filtering circuitry in the test rig. The output signal is -10.33 dBV, which is the level at which my audio analyzer performs the best.

NOTE: I have reduced the input impedance from 33k to 511 ohms, and updated the measurements. A man's gotta do something with those Tesla TR161 precision resistors...

The plastic around the Bursons gets comfortably warm (not hot, but they would be perfect for warming my hands now that it's winter), and they need a minute or so to warm up before the measurements stabilize. This is taken into account when performing the measurement below.

THD Measurements

Click here for a high resolution image of the measurement.

As it turns out, the V6 Classic is within the measurement limits of my QA401 with about 1.5 dB to spare. The THD shows up at -117.1 dB / 0.00014 %. The noise below 1 kHz shows up in this graph as well, showing that the Burson doesn't have the same noise rejection as the OPA2134.

While the OPA2134 performed beyond the abilities of the audio analyzer, the Burson didn't. This shows that the V6 Classic can't compete with the OPA2134 when it comes to THD at low gain. Still, this is pretty respectable.

Based on the THD alone, they should be virtually indistinguishable from most "normal" op amps in the gain stage of a DAC, headphone amp, CD player and similar line level equipment. I won't be testing their output capabilities, but they're probably not suited for output directly to a headphone (and neither are more than 99% of the IC op amps).

Measurements on their own are no good unless they can be seen in relation to something. Having established the measurements baseline, I want to get an idea on how the preferred OPA2134 performs.

The conditions are the same as they were with the baseline setup, and the op amp is placed in a test rig without any filter circuitry. The gain is 1.36 dB.

The OPA2134 shows a THD of -118.6 dB / 0.00012%, which is the same as I got on the loopback test. Notice that a lot of the noise below 1 kHz has vanished. This must be because of the OPA2134's noise rejection.

This is obviously beyond what the QA401 can measure, and Texas Instruments claims that the OPA2134 has a THD of 0.00008%. It does show what the audio analyzer is capable of, which is very useful for the next measurements.

Click here for a high resolution image of the measurement.

I've been asked about how the Burson V6 discrete operational amplifiers would perform in the Muffsy Phono Preamp quite a few times. They don't seem to have many published measurements, so I thought I'd give it a go.

To perform the measurements, I have a QuantAsylum QA401 audio analyzer. This will obviously not be measurements of datasheet quality, but I still hope I can find out if these op amps will work for the low signal levels from an MM cartridge. They have JFET inputs and work with up to +/- 16.5V power, which are two of the basic requirements.

To get the very best results from the QA401, I followed the tips in this blog article. Using the R- output and the R+ input gives the best measurements. I tested signal levels to find what gave the cleanest input signal and the very best THD measurements, which was -10.33 dBV.

A loopback test on the QA401 shows that I get a THD of -118.7 dB / 0.00012%. Operational amplifiers have great noise rejection, so I might get somewhat better measurement results. It shouldn't be by much though, this is basically as good as it gets.

The noise graphs at the bottom of this post shows that using a BNC cable (I tried several of them of varying brand and length) adds about 0.5 dB noise. I can't do anything about that...

Click here to see a high resolution image of the measurement.

For reference, here's the noise floor of the QA401, inputs shorted with BNC cables (hires image):

And finally the QA401's noise floor with the inputs shorted with BNC terminators (hires image):

Wondering how the Muffsy MC Head Amp measures? Look below for the numbers.

• Topology: Dual mono
• Gain: 24.6 dB
• 0.3 mV input = 5 mV output
• 0.5 mV input = 8.5 mV output 
• Input impedance: 47 - 100 ohms, in four steps
• THD: < 0.1 % *)
• THD+N: < 0.6 % *)
• SNR:  71.8 dB, referenced to 0.5 mV input @ 1 kHz
• Frequency Response: +0.0 / -0.1 dB (20-20.000 Hz)
• IMD (ITU-R): 0.001 %
• DC Offset: < 0.1 mV
• Dynamic range: > 110 dB, referenced to 0 dBV output
• Power requirements: +/- 15V at < 50 mA
• PCB Size: 84.0 x 59.4 mm / 3.3 x 2.3 in

The signal to noise ratio (SNR) compares the level of the audio signal to the background noise. An SNR of at least 65 dB is what's needed to have a completely silent phono stage. The Muffsy Phono Preamp's SNR, with OPA2134, comes in at 79.46 dB as shown below.

The measurement of signal to noise ratio can be a rather arbitrary affair. Some measure it as the difference between the noise and when the signal has 1% total harmonic distortion, while others compare it to a 0 dBV output. I've seen one manufacturer use 3 volts output as the reference point. With its 36 dB gain, it barely outputs a tenth of that with a normal MM cartridge connected.

For the Muffsy Phono Preamp, you'll get the signal to noise ratio as it is under normal use. This means that it will be referenced to an input of 5 mV, which is normal for a standard moving magnet cartridge. The output, at maximum gain, is then -2.54 dBV.

The measurement above shows the noise profile of the Muffsy PP-3 with OPA2134 operational amplifiers. The inputs are shorted in order to avoid picking up environmental noise, and the measurements are A-weighted. The graph is an average of 20 samples, instead of just picking "the best measurement".  

The red square shows the total noise power, from 20 to 20.000 Hz, which is -82.0 dBV.  The difference between the signal and the noise is then 82.0 dB - 2.54 dB. This gives a signal to noise ratio of 79.46 dB.

This is comparable to how commercial vendors  measure the signal to noise ratio, but here you can be sure that it is under normal operating conditions. That said, your cartridge will also create some noise of its own. When you connect your turntable to the Muffsy Phono Preamp, the actual signal to noise ratio will probably be around 77 dB.

It's worth noting that these measurements were done with the Muffsy Phono Stage powered by the Muffsy Power Supply. Both were housed in a B0905 enclosure with the Muffsy Back Panel, as described in the build instructions. The components were taken straight out of a Muffsy kit without any matching of any kind.

We've seen how the RIAA filter behaves with low and very low capacitor values. (See all simulation articles here) Now let's have a look at what happens if the capacitor values are higher than the recommended ones.

This is the Muffsy Phono Preamp with capacitor values that are 2.5% higher than specified.

As with the simulation with lower than recommended capacitors, this shows that we're still following the RIAA curve quite nicely even though the capacitor values are slightly wrong.

Now what happens if we bump the caps up to 10% higher than the recommended values?

The lowest level is -0.535 dB, and the highest is +0.186 dB. The average deviation is still only +/- 0.36 dB. There will be a slight attenuation of mid bass and the higher frequencies though, something that will become apparent if we really try to exaggerate by changing the capacitor values to +25%:

All the mids are attenuated and so is the treble, while the low bass is amplified a bit. +0.44/-1.16 dB is not something you'd want in your phono stage under normal circumstances.

A Conclusion...

The best scenario is to use the specified capacitor values, or within +/-10% of these. 

Deviations that are extremely low will give you a slight "loudness" effect, while extremely high capacitor values will attenuate the mids and the treble. Luckily, these extreme values won't enter your phono stage unless you do it on purpose.

These are also all good examples of what hides between the RIAA compliance numbers, that vendors seldom show in detail.

I wanted to know what happened to the Muffsy RIAA compliance if the capacitors were out of spec, and started with a plausible scenario where the capacitors were 3% below their stated values.

This time, lets see what happens if the capacitor values are extremely low. Let's say they're 10% below what it says on the capacitors. Like last time, the simulation goes from 20-20.000 Hz.

Clearly, the graph is a little more wiggly. But how bad is it? As it turns out, not too shabby at all.

The lowest deviation from the RIAA curve is -0.16 dB, which is all the way down at 20 Hz. The highest deviation is +0.55 dB, at around 19-20 kHz. Even though this could be read as if though the bass is lacking, notice that it is spot on again at 30 Hz.

What this simulation tells us is that you'll get a small boost in the bass and treble regions, an ever so slightly, very small loudness effect. Although this scenario isn't that probable, it could still happen as most of my measured capacitors are on the low side in the tolerance range.

So what happens if we take this to extreme levels? Pretty much more of the same. Below is a picture of the RIAA equalization circuit that is used in the Muffsy Phono Preamp.

These are the correct values:

•  C1: 136nF (2x 68nF)
  
• C2: 47nF

The results of the -10% capacitor values are expressed even more. You will get a huge boost in both the bass and treble frequencies.

The level is -1 dB at 20 Hz and gets compliant at 45 Hz. It increases to +1.85 dB at 200 Hz, down to -0.2 dB at 1.5 kHz, before it goes up to +1.55 dB at 20 kHz. 

You may very well find that you like this "loudness-RIAA", but it shouldn't be done in the phono stage. Nevertheless, this really illustrates what happens with extremely low capacitor values. Although it is way out of compliance, the result will be a fairly pleasant one.

For the next installment of these simulation, we'll look at what happens when the values are higher than what it says on the capacitors.

The RIAA equalization added to vinyl records can and do vary wildly, and you might rightfully say that a phono stage's RIAA compliance doesn't matter too much.

It's still a standard, and it should be one of the considerations when choosing a phono stage.

The Muffsy Phono Preamp uses Wima MKS2 capacitors with 10% tolerance. I have measured a couple of hundred of them, and find that they actually have much better tolerance. They tend to measure about 1 nF below the stated value, so I did some simulations to see what effects that deviation actually has.

This first graph shows a quite plausible scenario, where the values are off by -3%:

The lowest deviation from the RIAA curve is -0.01 dB, and the highest is +0.16 dB. 

This is realistically what you can expect from the Muffsy Phono Preamp kit in terms of RIAA compliance. Here's the measured RIAA compliance of a Muffsy Phono Preamp with no matched components:

These measurements show +0.21/-0.16 dB deviance. Just a smidge more than what we simulated above.

More simulations with really bad capacitor values will come, so stay tuned.