A "Dummy" Cart For Phono Preamp Noise Testing

[mediatechnology]

I did a quick comparison of the 2i2 Gen 3/Gen 4 here: viewtopic.php?p=17346#p17346

For me the extra 11 dB DR on the A/D makes it worth it.
They also lowered the output impedance from 300Ω/leg to 100Ω leg.
I think they were able to lower the output impedance by using external power rather than USB power.

I haven't used a E1DA Cosmac yet but I have a number of clients that use it for LP ripping and Test and Measurement.

[AnalogJoe]

Sounds like a great improvement. Amazon has it on stock and it is at a relatively low price.

To be honest, I think I would rather go with the Scarlett than the COSMAC. Focusrite is a very good company and their products are very robust.

Not too long ago I needed a 4 in/out interface for a project. I was going to buy the Scarlett 4i4 but I thought it was a bit too expensive. Someone recommended me the Behringer U-Phoria, they told me that it was reviewed by a famous audio enthusiast/reviewer or whatever and by several webpages, and that they all said it was amazing, that it had really good preamps, low noise, etc... and that the price was considerably lower than the rest of the sound interfaces of the same category. All of that is true, at least on paper.

I went ahead and bought one, knowing very well that my experience with Behringer products has not been the best in the past, and, as expected, it turned out to be the same old Behringer crap, functional, but crap. The drivers are mediocre and there is the occasional audio glitch which appears unexpectedly from time to time. The preamps are not that quiet after all and I get the feeling (haven't looked at a schematic yet) that they use the low-budget open-loop discrete diff amp front end followed by an op-amp architecture; the impedance at the outputs is higher than I would like, and everything feels "plasticky" and wobbly. As of now, it is being used as my personal computer's sound card whenever I want to listen to music on headphones. In essence, it is being used as a very expensive plug to miniplug converter for my headphones.

On the other hand, my Scarlett 2i2 3rd Gen has zero problems and it is solid as a rock, with excellent performance. So I am hesitant to try something different again; the devil you know....

[mediatechnology]

For me the Cosmac's fixed input gain and lower input noise would allow it to replace the PCM4222 EVM I use for critical measurements.

The 2i2 Gen 3 however is my daily driver for test and measurement. I also have a second one on a desktop PC.
Both have an iFi USB ground isolator with external power injection.
They're subjectively quieter with them.

[AnalogJoe]

For me the Cosmac's fixed input gain and lower input noise would allow it to replace the PCM4222 EVM I use for critical measurements.

The 2i2 Gen 3 however is my daily driver for test and measurement. I also have a second one on a desktop PC.
Both have an iFi USB ground isolator with external power injection.
They're subjectively quieter with them.

Yikes! that iFi isolator can cost more than the interface, and the cheapest ones are around $60.

How are you using the PCM4222EVM? i.e., in what arrangement/situations?

I just noticed what you said about the Cosmac. It has dip switches beneath it to change the input sensitivity, seems a bit of a hassle, which is why I ask how are you currently using your PCM4222

[mediatechnology]

For me the Cosmac's fixed input gain and lower input noise would allow it to replace the PCM4222 EVM I use for critical measurements.

The 2i2 Gen 3 however is my daily driver for test and measurement. I also have a second one on a desktop PC.
Both have an iFi USB ground isolator with external power injection.
They're subjectively quieter with them.

Yikes! that iFi isolator can cost more than the interface, and the cheapest ones are around $60.

How are you using the PCM4222EVM? i.e., in what arrangement/situations?

You can make a USB isolator/injector with a resistor in the ground line and DC power supply. "Thor" over at GDIY showed a simple sch of the iFi.

The PCM4222EVM has AES and SPDIF outputs which require it to be used with the Emu 0404 or Roland Quad Capture.
I use the PCM4222 EVM whenever I need a lower noise measurement. It has another 20 dB of DR.
It has "naked" low input impedance inputs without gain control and a big drag is that it leaks Vcm out of the inputs because it has no input coupling caps.

[AnalogJoe]

For me the Cosmac's fixed input gain and lower input noise would allow it to replace the PCM4222 EVM I use for critical measurements.

The 2i2 Gen 3 however is my daily driver for test and measurement. I also have a second one on a desktop PC.
Both have an iFi USB ground isolator with external power injection.
They're subjectively quieter with them.

Yikes! that iFi isolator can cost more than the interface, and the cheapest ones are around $60.

How are you using the PCM4222EVM? i.e., in what arrangement/situations?

You can make a USB isolator/injector with a resistor in the ground line and DC power supply. "Thor" over at GDIY showed a simple sch of the iFi.

The PCM4222EVM has AES and SPDIF outputs which require it to be used with the Emu 0404 or Roland Quad Capture.
I use the PCM4222 EVM whenever I need a lower noise measurement. It has another 20 dB of DR.
It has "naked" low input impedance inputs without gain control and a big drag is that it leaks Vcm out of the inputs because it has no input coupling caps.

I see, so you are going in through the SPDIF inputs of the sound interface and using the PCM4222 for its AD converter. Sounds like a good setup. I just checked and the Cosmos doesn't seem to be available anywhere for purchase, even in their own webpage the "add to cart" button is disabled.

Regarding the isolator, I will check Thor's thread, but by isolation I was expecting more like galvanic isolation, and what you describe seems more of attenuation rather than isolation. A transformer at such frequencies required for USB's fast data transfer shouldn't be easy to find though, maybe one of TI's isolation amplifiers could do the job, I am starting to understand why those iFi solutions are not cheap.

[mediatechnology]

The iFi doesn't offer full galvanic isolation. It's expensive because it has a high mark-up.
The added series R in the ground, which sometimes has back-to-back diode clamps in some people's builds, is usually enough to quieten them.

Here's a link to E1DA's store: https://www.aliexpress.com/store/1101346048
And his site: https://e1dashz.wixsite.com/index

He posts frequently at DIYAudio - I forget his handle but I see him often in Equipment and Parts. I wanna say inxs

[AnalogJoe]

The iFi doesn't offer full galvanic isolation. It's expensive because it has a high mark-up.
The added series R in the ground, which sometimes has back-to-back diode clamps in some people's builds, is usually enough to quieten them.

Here's a link to E1DA's store: https://www.aliexpress.com/store/1101346048
And his site: https://e1dashz.wixsite.com/index

He posts frequently at DIYAudio - I forget his handle but I see him often in Equipment and Parts. I wanna say inxs

Im very tempted to buy one of those Cosmos interfaces. I have a couple of reservations though. First, it is only AD, and that would be ok if it had SPDIF outputs, however, it is a full-fledged interface; I don't know if all software will be comfortable by doing a loopback or by playing/recording through different interfaces. Have you experienced with that? having one audio interface for reproduction and another for recording? I also use MATLAB, and I don't think MATLAB will be very happy to do that

[mediatechnology]

I haven't found a good work-around in AudioTester for ASIO devices that are not full duplex.
The TI PCM4222 EVM shows up as the Emu0404 when its SPDIF output is internally routed to the 0404's output.
I can use the 0404's D/A and the 4222 A/D simultaneously using the 0404's console.
The D/A locks to the 4222's sample rate.
I don't use that configuration often: The 2i2 is good enough for most measurements.

ASIO4ALL can build aggregate devices but they have to share a common clock. The 2i2 is internal only.

I would use the COSMOS for noise measurements, vinyl capture, and THD measurement using ULDO-Nacho.
For vinyl capture it's overkill.

[AnalogJoe]

I haven't found a good work-around in AudioTester for ASIO devices that are not full duplex.
The TI PCM4222 EVM shows up as the Emu0404 when its SPDIF output is internally routed to the 0404's output.
I can use the 0404's D/A and the 4222 A/D simultaneously using the 0404's console.
The D/A locks to the 4222's sample rate.
I don't use that configuration often: The 2i2 is good enough for most measurements.

ASIO4ALL can build aggregate devices but they have to share a common clock. The 2i2 is internal only.

I would use the COSMOS for noise measurements, vinyl capture, and THD measurement using ULDO-Nacho.
For vinyl capture it's overkill.

I see that the COSMOS also has stellar THD, so I guess I would like to use it for THD measurements, although that also will depend on the purity of the tone being reproduced, I have a Tek SG505 , but that works at only one frequency, for a THD sweep I would need the computer to do it and reproduce the swept tone, so we are back again where we started.

I think ASIO4ALL is amazing, I've used it many times with MATLAB and it works like a charm.

Regarding vinyl capture. I still can get in mind around why would anyone need an a/d with such a low noise, the SNR of a vinyl record is what, like 60 dB? this thing has 129 dB of SNR. What is the logic behind it?

PS That ULDO-Nacho looks amazing.... the distortion spec is even lower than the SG505.