Make a good measurement system with REW and Raspberry

[Antonio Di Motta]

the development of the 4.0 version analyzer continues. Today I am attaching the block diagram of the analyzer as it was designed, suggestions are clearly accepted.
In the meantime, i am finishing making the sound card and the raspberry pi4 block and cooling system as the presence of the sound card very close to the CPU card causes overheating that is not very safe for the operation of both cards.

Thanks for your attention.

 

[Antonio Di Motta]

Here is the new sound card, it has been updated to have a further increase in distortion and S / N ratio performance, to do this I had to change part of the components and above all provide a different positioning of the card compared to the raspberry to minimize electromagnetic interference and improve the CPU heat dissipation of raspberry PI 4.
I can't yet know the level of improvement achieved because the board is physically being assembled, but I can show you how it will be built once it is built.
Thanks for your interest

Greetings
Antonio

 

[trobbins]

...

 

[Shonver]

Antonio, thanks again for sharing your project progress. I just re-read the thread and was wondering about the compatibility with or, otherwise, choice of calibrated microphones that will work with your sound card. The 48V phantom powered mics are common, while you have mentioned a specific model (lower phantom voltage?) that I believe you want to apply the Linkwitz modification to. Have you had any further thoughts on mic candidates?

 

[Antonio Di Motta]

Good evening, the very first versions of the instrument dating back over 20 years ago I used a Pioneer CM-75M microphone powered with 1.5V battery but the performances were rather limited especially in terms of dynamics, then from version 2.0 up to version 3.0 of the instrument I used the microphone built with a modified Panasonic Linkwitz microphone capsule. powered at -9V this microphone was calibrated by an Italian company Teleprodotti.
Currently in version 4.0 of the instrument I have redesigned the microphone preamp stage and it works in balanced mode with phantom voltage of + 48V.
To be precise, version 4.0 has the ability to work with both configurations with power supply at -9V and microphone Panasonic WM-61A and with standard + 48V power supply for condenser microphones with balanced output.
I use a Sonarworks XREF20 individually calibrated and fairly affordable cost of around 70 / 80E.

Where can I get the calibration profile for my XREF 20 microphone? | Sonarworks Support

SoundID Reference software calibrates your speakers and studio headphones, so you can trust that every mix will translate.

www.sonarworks.com

Greetings

Antonio

 

[Antonio Di Motta]

I am very late in the construction of the instrument, , but I want to show you roughly how it will be made by posting a figure that I created by combining two circuits and aborates with kicad. On the left side there is the whole power supply system consisting of the +/- 12V regulators for the + 12V opamps to power the i + 5V relays to power the Raspberry Pi 4 and the Arduino Nano microcontroller and + 48V to power the microphone.
Then there is the unregulated +/- 24V power supply to power the LM3886 power amplifier. The toroidal transformer is placed entirely on the left and will certainly be electromagnetically shielded. In the central part there is the analog signal switching system, the microphone preamplifiers, the line input and the line output.
On the right side there is the sound card which acts as a bridge with the GPIO bus of the raspberry which is positioned all the way to the right. This solution allowed me to free the raspberry from the sound card and to mount a passive heatsink on the whole size of the circuit in order to have a greater heat dissipation, moreover the sound card is further away from the digital circuits in order to reduce the electromagnetic interference.
Finally, the raspberry Pi 4 is positioned all on the right.
I will still have to work a lot because the motherboard is still to be completed, we hope to be able to finish everything by 2022.
Thanks for the attention.

Greetings

Antonio

 

[goginux]

sold Sonarworks XREF20 and bought Sonarworks XREF20R5, which is a little more sensitive. I have some measurements compared to umik-1 cross spectrum calibrated labs. see here: https://www.facebook.com/sorin.claudiu.16752/

 

[Antonio Di Motta]

Interesting your speaker, well built. I have seen some measurements made with REW and I must say they have a very linear response, great job.
As for the two microphones, they have very similar performance I use XREF20.

Greetings
Antonio

 

[goginux]

https://www.hifiberry.com/shop/accessories/mic/ Have you tried this microphone?

 

[Antonio Di Motta]

Sorry, I haven't tried any USB microphone solution.
Greetings Antonio

 

[WesB]

Hi Antonio:

I would be very interested in the results of the noise improvement from your sound card. As we've discussed earlier the majority of sound boards on the market for the raspberry pi are not as quiet as one would like. I have the hifiberry dac+ adc pro. but would like to know if yours is going to be quieter.

Thanks again for sharing all this hard work.

Wes Brenner

 

[Antonio Di Motta]

my sound card VT board is silent let's say that with RTA the noise level remains at around -140dbfs, but it was difficult to reach a particularly low level of distortion around 0.00037% at 1Khz with input signal at -1dbfs .
This value when compared to the HiFi Berry dac + adc card is much better in fact this card stops at about 0.034% distortion with a much lower input signal level about -5dbfs. For these reasons the VT Board is certainly much better than the HiFi Berry dac + adc while I don't know the performance of the dac + adc pro.

Greetings

Antonio

 

[Antonio Di Motta]

I recovered this graph where you can see the performance of the card in this case with an input signal level at -3dbfs

 

[Antonio Di Motta]

Unfortunately the latest version of the card with changes that should give some further improvements was not built due to lack of chips. I'm currently busy building the instrument's motherboard, I'm well underway and hope to build it soon.
Thanks for the attention.

Greetings
Antonio

 

[WesB]

Hi Antonio:

Glad to hear you are still working on the project. Wasn't sure since it had been a while since you posted.

Since the Raspberry Pi's operating system is not a real time OS the Hifiberry DAC+ADC PRO apparently has a time base built-in. Does your sound card include a time base or does it rely on the Pi's oscillator?

Too bad about the chip delays, but I'm glad and appreciative of your efforts.

Wes Brenner

 

[Antonio Di Motta]

I have been working on it for many years even if not very assiduously but I hope to be able to complete the entire project of the 4.0 version of the instrument between the end of 2022 and the beginning of 2023.
The big delay mainly depends on the realization of the sound card as the ones I tested they didn't perform well. Raspian OS is a great operating system, it doesn't have a kernel optimized for real time but it can be upgraded to be one.
However the VT board has the clock generator on board which can be a quartz or a high stability TCXO. Version 2.0 of the card can be configured in three different solutions:
Base edition : clock with 30ppm quartz and standard performance in / out buffer
Standard edition : clock with 30ppm quartz and high performance in / out buffer
Performance edition : Clock with TCXO 0.1-1ppm and high performace in / out buffer.

The measurements I made refer to the Base edition.

The VT board was created to be used in the audio measurement instrument that uses Raspberry and REW and if you ask me if it can be improved I will answer yes but it is tremendously difficult. I am now working on version 3.0 of the card, but maybe I won't be able to complete it any time soon and I'm not sure if it performs better than version 2.0, but I hope I can make it to find out.

Thanks for the attention

Greetings Antonio

 

[WesB]

Hi Antonio:

Thanks for all the hard work. When you complete this sound card will you make it available for sale as a separate board? My current need is for the sound card board to replace the Hifiberry board in my test setup with the best 192k/24 board designed for the Raspberry PI which I think may well be yours.

Thanks again,

Wes Brenner

 

[WesB]

Hi Antonio:

Its been awhile and I thought I would send a note to see how things are going on the project.

As I had indicated I am very interested in the sound card portion of you project and was wondering if you have achieved the performance you had hoped for. Particularly, noise levels at 192K 24 bit configuration.

 

[Antonio Di Motta]

Good evening Wes Brenner, I am very pleased that there is someone who appreciates my sound card. My intention was to create a kit for acoustic enthusiasts to build their own measuring instrument using what in my opinion is the best free software available (REW).

Unfortunately the pandemic and the war in Ukraine have made the availability of chips almost non-existent for us private individuals and considering that the Raspberry costs at least four times the original price, I remained firm with the project. The intention was to resume the project as soon as the chips are available again, in the meantime I try all the versions of REW on the raspberry to see if they work correctly.

a greeting

Antonio DiMotta

 

[WesB]

Good to hear from you again, Antonio.

I understand the situation with chips and the currently outragious prices for the Raspberry Pi's that have been caused by the pandemic and the Ukraine war. I'm sure we all hope both of these situations are solved soon.

I remember you had some experience with the audioinjector and hifiberry products and I would like to see any comparisons for 192khz/24 bit measurements whenever they become available. Particularly noise levels.

Thanks for all the hard work on this project and I'm sure others are appreciative of the effort on the main device as well as the sound card portion.

Wes

 

[koenbro]

Any news about the card, Antonio? I am following with great interest.

 

[Antonio Di Motta]

Hi, I designed two other versions of the card trying to improve some aspects but at the moment I couldn't make them due to lack of time and above all of chips. I hope the semiconductor market recovers to have the chips I need to make this board.
Greetings from Italy

Antonio

 

[WesB]

Hi, I designed two other versions of the card trying to improve some aspects but at the moment I couldn't make them due to lack of time and above all of chips. I hope the semiconductor market recovers to have the chips I need to make this board.
Greetings from Italy

Antonio

Hi Antonio:

Using amazon as an indicator it seems the Pi's are getting closer to normal (specifically the Pi 3's and Pi 4's). I'm still interested in your project and in particular your sound card portion since it is hard to find a good one that has a direct connection the Pi. Hopefully the chips will come down and you can make progress on the project. Any updates would be appreciated. Thanks again for all the hard work.

Wes Brenner

 

[goginux]

The last one is raspberry pi5, very fast. https://www.raspberrypi.com/products/raspberry-pi-5/

 

[Antonio Di Motta]

I have the Raspberry 5, but I haven't had a way to test compatibility with my sound card. Unfortunately, due to the lack of low-noise LDO regulators, I was unable to update the sound card and I am concentrating on creating the new motherboard of the analyzer ver. 4.0.
At the moment I'm very late because I can dedicate little time to the creation due to work commitments, but I promise to post something as soon as I can make progress with the work.

Greetings from Italy.

Antonio Di Motta