DSP-Based Phono Preamp/ADC with Digital Out?

[jdubs]

What doesn't?

Sorry, the Parks Audio Puffin.

 

[Soniclife]

Sorry, the Parks Audio Puffin.

48/24 is that much of a limitation?

 

[levimax]

I just completed a project that may be a solution for you if you don't mind DIY. I did it because I had computer/ Wifi noise pick-up issues with my TT as well as just wanting to experiment with DSP RIAA.

1. I built this "balanced transmitter" but changed the input resistor to 47K ohm to match cart, I bought the PCB as it is much easier that way. https://sound-au.com/project87.htm
2. I located the transmitter very close to TT with very short interconnects.
3. From transmitter I run balanced cables to my ADC (Steinberg UR 22 Mk2) connected to a PC.
4. I used rephase to make a RIAA convolution filter.

Much to my surprise this worked perfectly first try. The ADC has plenty of gain (no need for phono pre-amp with MM cart, I use SUT's first for MC), the RIAA convolution filter was trivial to make and works perfectly, and I have greatly reduced noise both the "hiss" and the "picked up computer noise". The ADC can feed the computer up to 192/32 if you wanted but I use 48/24 which is more than enough.

 

[Frank Dernie]

48/24 is that much of a limitation?

It is more than enough. In fact it results in a bigger than necessary stream.
The dynamic range of LPs is at very best 12 bit and whilst one can often see output above 24kHz it is mainly distortion artefacts from the cartridge and, in any case, inaudible.

 

[jdubs]

I'm going to try a TC Konnekt Live....an older firewire based interface, but it does RIAA correction, digitally, onboard.....and can output it at 24/96 via s/pdif. Also comes with the appropriate cabling to hook up to a turntable directly.

-Jim

 

[jdubs]

I would verify the RIAA conformance the description in the manual is ambiguous.

Yep, I've got an email into TC to check on this very thing.

 

[scott wurcer]

Yep, I've got an email into TC to check on this very thing.

Sorry I removed my comment because I wanted to read the rest of the manual but there was no further discussion. Their description of the "1964 RIAA" is not totally clear, I have seen simple approximations here and there that can be several dB off.

 

[jdubs]

Sorry I removed my comment because I wanted to read the rest of the manual but there was no further discussion. Their description of the "1964 RIAA" is not totally clear, I have seen simple approximations here and there that can be several dB off.

Lol, yes, very general descriptions for the two available:

"RIAA filter 1964 The RIAA filter 1964 gently enhances the low-end frequencies with 6 dB per descending octave. The high end frequencies are attenuated also with 6 dB per octave. Around 1 kHz the level is left untouched.

RIAA filter 1987 The RIAA 1987 is a variation of the original 1964 filter. In this type a soft cut in the low-end frequencies has been added to reduce rumble and acoustic feedback. Depending on you specific application you may prefer one or the other."

 

[jdubs]

Does anyone have the biquad coefficients for RIAA correction processing at 96khz? For input into a MiniDSP SHD.

Thanks
Jim

 

[levimax]

Does anyone have the biquad coefficients for RIAA correction processing at 96khz? For input into a MiniDSP SHD.

Thanks
Jim

Not sure if this helps but in rephase the following 3 filters work as RIAA correction.... I think you may be able to work backwards from this to get biquad coefficients or use directly with MiniDSP (I don't have one so not sure).

Normal Low pass first order 2122 Hz
Compensate High Pass first order 500 Hz
Normal High pass first order 50 Hz

 

[scott wurcer]

Does anyone have the biquad coefficients for RIAA correction processing at 96khz? For input into a MiniDSP SHD.

Thanks
Jim

Yes I derived them all. You need to check on the convention for a and b being numerator or denominator.

a coefficients 1. -1.866632 .86705829
b coefficients 1. -.85352776 -.11046426

gain at 1kHz 18.625dB

 

[jdubs]

Yes I derived them all. You need to check on the convention for a and b being numerator or denominator.

a coefficients 1. -1.866632 .86705829
b coefficients 1. -.85352776 -.11046426

gain at 1kHz 18.625dB

Thanks Scott! Is the gain at 1khz something that needs to be input? This is the minidsp interface:

 

[scott wurcer]

Thanks Scott! Is the gain at 1khz something that needs to be input? This is the minidsp interface:

View attachment 57852

I forgot any quirks in the minidsp interface you need to read the docs I remember some sign convention issue. The gain is what you get without scaling the coefficients, you can scale the output by multiplying them appropriately.

 

[jdubs]

Scott, does that mean that I multiply the coefficients each by 18.625?

-Jim

 

[scott wurcer]

Scott, does that mean that I multiply the coefficients each by 18.625?

-Jim

That's dB not 18.625X, just get the basic response to work and then scale it to fit for your device.

 

[audiopile]

miniDSP direct in is very interesting ! BTW-Parks Audio Puffin can now be ordered with either a 24/96 SPDIF (RCA) or optical TOSLINK as a extra cost option. Have now run this (SPDIF and analog at different times) with several MM carts and a handful of moving coils - very impressed and even more importantly - pleased. The latest software update adds a bucket of really useful test/analysis functions -these alone would be worth the price.

 

[watchnerd]

miniDSP direct in is very interesting ! BTW-Parks Audio Puffin can now be ordered with either a 24/96 SPDIF (RCA) or optical TOSLINK as a extra cost option. Have now run this (SPDIF and analog at different times) with several MM carts and a handful of moving coils - very impressed and even more importantly - pleased. The latest software update adds a bucket of really useful test/analysis functions -these alone would be worth the price.

Thanks for sharing.

What are the new software test/analysis features?

I'm not seeing anything on the website that wasn't there back in May.

 

[audiopile]

What I think is really nice about Shannon/Parks Audio in terms of a Puffin specifically - is he offers the first adopters with the earliest units that (8/3/2020) V1.21 download for free . Way better than the one year -maybe 2year product cycle that most of the audio world seems to run on nowadays. I'm going off of what is on his website in the last month or so - the Puffin was slightly below my horizon up until then -adding the SPDIF out got it to pop into view for me as a consumer. But this black box is more than just a phono pre-amp -it's a tool set if you choose to use it that way.

 

[metaleonid]

I do have PS Audio NuWave Phono Converter that I use to transfer vinyl. Additionally I use TASCAM US-366 USB Audio interface with microphone pre-amps to transfer vinyl playing it at speed 22.5 RPM and then applying RIAA curve digitally after correcting speed in the sound editor. I usually use this method on vinyl that is hard to track at normal speed (with higher recording volume). TASCAM USB interface has enough GAIN even for very low output MC carts. It also has very low s/n ratio. While devices such as Apogee Element 24 have higher gain for mic preamps than TASCAM US-366, their s/n is a lot worse.

 

[scott wurcer]

I do have PS Audio NuWave Phono Converter that I use to transfer vinyl. Additionally I use TASCAM US-366 USB Audio interface with microphone pre-amps to transfer vinyl playing it at speed 22.5 RPM and then applying RIAA curve digitally after correcting speed in the sound editor. I usually use this method on vinyl that is hard to track at normal speed (with higher recording volume). TASCAM USB interface has enough GAIN even for very low output MC carts. It also has very low s/n ratio. While devices such as Apogee Element 24 have higher gain for mic preamps than TASCAM US-366, their s/n is a lot worse.

Just curious the input referred noise on the Tascam is -120dB (no weighting mentioned) and the Apogee is -129dB with no weighting per the data sheets. I assume these are at 200 Ohm or so typical mic impedances.