Thank you OR, my question was not about carrying balanced output from your DAC through RCA... but if you think that there is a sonic difference, apart from RFI noise, between unbalanced or balanced connection from a balanced DAC (and apart from the dynamic range).
In other words, as I have unbalanced connections in my system I have RFI interference with or without a balanced DAC. Is it worth to have an ES9038 DAC instead of an ES9016 DAC ? ... but maybe, as you said, "it just wouldn't make sense to use PRO series Sabre DAC, which has differential outputs by itself and then convert the signal to single ended (vulnerable to RFI) to be transported out of the chassis." ...
Cheers
Marking a clear anodized aluminium with a CO2 laser is a tricky business (unlike colored anodized aluminium, which is very easy to mark). It's one of the processes we had to master while designing and making an enclosure for our upcoming DAC. To be unveiled very soon.
Kreisky:
Perception of sound quality is highly subjective, so I don't want to make any judgements, but my personal choice between the two would be properly designed DAC with ES9016 and balanced wiring over ES9038PRO with unbalanced wiring.
Switching to balanced signal paths is certainly worth it. Beware of equipment with improper shield/ground termination, suffering from so-called "Pin 1 problem".
Pavel, Okto Research
www.oktoresearch.com
info@oktoresearch.com
Kreisky:
Perception of sound quality is highly subjective, so I don't want to make any judgements, but my personal choice between the two would be properly designed DAC with ES9016 and balanced wiring over ES9038PRO with unbalanced wiring.
Switching to balanced signal paths is certainly worth it. Beware of equipment with improper shield/ground termination, suffering from so-called "Pin 1 problem".
Pavel, Okto Research
www.oktoresearch.com
info@oktoresearch.com
Hello Okto Research - your products look fantastic! Exactly my philosophy of few things done right.
Please pardon my lack of understanding, but how exactly does a setup with a DAC8 work in conjunction with a Raspberry Pi running Volumio, and a MiniDSP product? My understanding is that the MiniDSP products have a built-in DAC, though also have both I2S in and out, with the exception being the MiniSHARC, which is all I2S. So is the expected usage:
RPi (via I2S) --> MiniSHARC (via I2S) --> DAC8 --> amplification?
I have 3 way speakers with dedicated binding posts for each driver. While they have "crossovers", they are actually just filters for each driver, which I may bypass if I get the hang of digital crossovers right. Then the idea is for me to have individual amplification for each driver, with the amp modules (Neurochrome) having balanced inputs.
How would I accomplish this?
Also a couple of comments:
1) the link to the manual in Post #10 is no longer active:
2) there is a small typo in the original post:
Thanks!
Please pardon my lack of understanding, but how exactly does a setup with a DAC8 work in conjunction with a Raspberry Pi running Volumio, and a MiniDSP product? My understanding is that the MiniDSP products have a built-in DAC, though also have both I2S in and out, with the exception being the MiniSHARC, which is all I2S. So is the expected usage:
RPi (via I2S) --> MiniSHARC (via I2S) --> DAC8 --> amplification?
I have 3 way speakers with dedicated binding posts for each driver. While they have "crossovers", they are actually just filters for each driver, which I may bypass if I get the hang of digital crossovers right. Then the idea is for me to have individual amplification for each driver, with the amp modules (Neurochrome) having balanced inputs.
How would I accomplish this?
Also a couple of comments:
1) the link to the manual in Post #10 is no longer active:
2) there is a small typo in the original post:
Thanks!
Hi,
there is also NanoSHARC. They have similar processing capabilities, but while the MiniSHARC have AES3, SPDIF and TOSLINK inputs, the NanoSHARC have UAC2 USB with XMOS (and TOSLINK too). So you could feed the NanoSHARC from the RPi using the USB.
The setup you described seems like a great match.
Thanks for the typo check, I'll ask Franta to fix it once he's back from holiday. The link on the website is working:
https://www.oktoresearch.com/assets/okto-dac8-module-user-manual_v3.pdf
Pavel, Okto Research
www.oktoresearch.com
info@oktoresearch.com
there is also NanoSHARC. They have similar processing capabilities, but while the MiniSHARC have AES3, SPDIF and TOSLINK inputs, the NanoSHARC have UAC2 USB with XMOS (and TOSLINK too). So you could feed the NanoSHARC from the RPi using the USB.
The setup you described seems like a great match.
Thanks for the typo check, I'll ask Franta to fix it once he's back from holiday. The link on the website is working:
https://www.oktoresearch.com/assets/okto-dac8-module-user-manual_v3.pdf
Pavel, Okto Research
www.oktoresearch.com
info@oktoresearch.com
Hi Okto,
Cabinet design is indeed a difficult thing. Please keep in mind that nanoSHARC or miniSHARC do not have cabinets. Please make sure to accommodate them in your cabinet. This way you can offer one box that provides a complete digital crossover + DAC. If this means that you have to make a choice of Nano or Mini on behalf of the customer, that is perhaps fine. I2S does not have a known physical connectors standardized and so it is best for this interface to be within the same box.
Cabinet design is indeed a difficult thing. Please keep in mind that nanoSHARC or miniSHARC do not have cabinets. Please make sure to accommodate them in your cabinet. This way you can offer one box that provides a complete digital crossover + DAC. If this means that you have to make a choice of Nano or Mini on behalf of the customer, that is perhaps fine. I2S does not have a known physical connectors standardized and so it is best for this interface to be within the same box.
Last edited:
feature request
Very nice design!
My interest is for an 8-channel DSD DAC board (@5,644,800 MHz sampling rate), with individual, full-range volume control per channel. My preference is a Tascam DB25 connector for the analog outputs.
My questions:
Is it possible to control volume for each channel individually via i2c? If so, can you add instructions to your documentation?
The 26-pin connector supports PCM and DSD modes but it is not clear to me how to select each mode.
Would you consider adding DB25 output option.
Will the dedicated power supply unit drive 2 boards?
Thanks,
Simon
Very nice design!
My interest is for an 8-channel DSD DAC board (@5,644,800 MHz sampling rate), with individual, full-range volume control per channel. My preference is a Tascam DB25 connector for the analog outputs.
My questions:
Is it possible to control volume for each channel individually via i2c? If so, can you add instructions to your documentation?
The 26-pin connector supports PCM and DSD modes but it is not clear to me how to select each mode.
Would you consider adding DB25 output option.
Will the dedicated power supply unit drive 2 boards?
Thanks,
Simon
Hello Simon,
I suggest waiting for DAC8 as a complete unit in an enclosure (will be announced this year!). Separate volume control for each channel will be available in the menu. We could supply it without any input module (with bare I2S/DSD inputs) if that is of interest to you. It shouldn't be a problem for us to make a custom back panel with DB25 instead of XLR's too.
Incoming signal format (PCM/DSD) to the DAC8 module is detected automatically by the Sabre itself.
If multiple DAC8 modules were used, separate PSU would be needed for each one of them because of galvanic isolation.
Pavel, Okto Research
www.oktoresearch.com
info@oktoresearch.com
I suggest waiting for DAC8 as a complete unit in an enclosure (will be announced this year!). Separate volume control for each channel will be available in the menu. We could supply it without any input module (with bare I2S/DSD inputs) if that is of interest to you. It shouldn't be a problem for us to make a custom back panel with DB25 instead of XLR's too.
Incoming signal format (PCM/DSD) to the DAC8 module is detected automatically by the Sabre itself.
If multiple DAC8 modules were used, separate PSU would be needed for each one of them because of galvanic isolation.
Pavel, Okto Research
www.oktoresearch.com
info@oktoresearch.com
Beautiful!
Some questions for you:
1) Can we see pictures of the rear and the innards?
2) If not, is there any room to fit a RPi inside the chassis? If not, Is the NanoSHARC USB port exposed in the rear so one could connect a RPi externally?
3) Is there a mechanism to get i2s input to the MiniSHARC?
4) And this applies to your DAC8 board in general - how tedious is the process to convert from multi-channel to Stereo? Is it something that can be done in software? It would be real neat to have the ability to switch between the two (option1: 6 channel with digital x/over. option2: 2-channel full range - either through two XLR outs alone, or repeated through all xlr outputs). I understand this scenario is not very useful for many.
I see this beginning to nip at the heels of the MiniDSP SHD, which is on my list. The built-in streaming using Volumio and the free ADC and input select makes this super versatile. Its DAC is what is holding me back.
Either way, great work on this and happy new year!
Some questions for you:
1) Can we see pictures of the rear and the innards?
2) If not, is there any room to fit a RPi inside the chassis? If not, Is the NanoSHARC USB port exposed in the rear so one could connect a RPi externally?
3) Is there a mechanism to get i2s input to the MiniSHARC?
4) And this applies to your DAC8 board in general - how tedious is the process to convert from multi-channel to Stereo? Is it something that can be done in software? It would be real neat to have the ability to switch between the two (option1: 6 channel with digital x/over. option2: 2-channel full range - either through two XLR outs alone, or repeated through all xlr outputs). I understand this scenario is not very useful for many.
I see this beginning to nip at the heels of the MiniDSP SHD, which is on my list. The built-in streaming using Volumio and the free ADC and input select makes this super versatile. Its DAC is what is holding me back.
Either way, great work on this and happy new year!
Thread title changed at user request.Thank you, I am going to reply in a thread dedicated to our complete products. Please keep this thread for the discussion about our modules only.
Happy new year,
Pavel, Okto Research
www.oktoresearch.com
info@oktoresearch.com
Hi Pavel,
I m considering your module as a possible multichannel DAC solution for a DSP module which I m prototyping, I have 2 questions:
- is it possible to access the I2C bus without too much conflict with the onboard MCU ? may be I could remove the MCU ?
- in the manual you say 3.4Vrms for the output, I assume this is the differential voltage on the XLR pins 2-3, it looks quite small compared to the possibilities of the OP1632 powered with the 2x15V transformers... I m asking because I d like to drive a NC500 (or other hypex but without their grain stage) and I d need ideally 8 to 12V rms.
- is there a possibility to configure the board so that it could be master with a lrclk at 96khz, bclk at 64fs (don't need the mclk) just by leveraging the onboard crystal (hopefully 49152k?) so my DSP could stream the original flow without going trough any asrc ?
many thanks
I m considering your module as a possible multichannel DAC solution for a DSP module which I m prototyping, I have 2 questions:
- is it possible to access the I2C bus without too much conflict with the onboard MCU ? may be I could remove the MCU ?
- in the manual you say 3.4Vrms for the output, I assume this is the differential voltage on the XLR pins 2-3, it looks quite small compared to the possibilities of the OP1632 powered with the 2x15V transformers... I m asking because I d like to drive a NC500 (or other hypex but without their grain stage) and I d need ideally 8 to 12V rms.
- is there a possibility to configure the board so that it could be master with a lrclk at 96khz, bclk at 64fs (don't need the mclk) just by leveraging the onboard crystal (hopefully 49152k?) so my DSP could stream the original flow without going trough any asrc ?
many thanks
Hi maxidcx,
- the on-board MCU should be removed from the socket in case of custom I2C access. We have a PDF for our customers regarding the Sabre register settings
- output voltage isn't limited by OPA1632 output swing or power rail voltage (+12V and -12V after on-board LT3045 super-regulators), it is just our choice of feedback resistors. It can be set to up to 7.9Vrms on request
- the MCLK oscillator is indeed 49152k. Looking into the Sabre datasheet, in the master mode, BCLK can be set from MCLK/2 to MCLK/16, so with MCLK/8 you should be able to achieve your goal. If you email us, I could send you more info regarding this for you to evaluate.
Pavel
- the on-board MCU should be removed from the socket in case of custom I2C access. We have a PDF for our customers regarding the Sabre register settings
- output voltage isn't limited by OPA1632 output swing or power rail voltage (+12V and -12V after on-board LT3045 super-regulators), it is just our choice of feedback resistors. It can be set to up to 7.9Vrms on request
- the MCLK oscillator is indeed 49152k. Looking into the Sabre datasheet, in the master mode, BCLK can be set from MCLK/2 to MCLK/16, so with MCLK/8 you should be able to achieve your goal. If you email us, I could send you more info regarding this for you to evaluate.
Pavel
Thank you for your answer Pavel; can you tell us if increasing the OPA1632 max rms would change THD and SNR figures in same order of magnitude (meaning it might be useless and we can tune the gain chain later on) or if this brings a tangible benefit?
guessing the OPA as a current bias and some ideal sweet spot, same for noise level compared to resistor value and dac.
thank you
guessing the OPA as a current bias and some ideal sweet spot, same for noise level compared to resistor value and dac.
thank you
This looks like a very nice solution. In the manual it says that other clock solutions are available, would it be possible to fit a "Pulsar clock" ? (I allready have it). Is it possible to connect the signal directly @ the sabre dac bypassing the onboard output sections with alternative ones ? (I also already have). How far of would you estimate that the 9038pro is ? Are the 15v supplys for opamps only ?
Last edited:
No problem in using Pulsar clock with DAC8 module. The on-board oscillator can be turned off with a solder jumper and the "MCLK I/O" pin on the main connector then serves as a clock input.
Regarding replacing the I/V sections - not sure what would be the point. The choice of components, very tight layout on 4-layer PCB and the use of fully-differential amplifiers is what makes the DAC8 module perform so well. There are no connectors, but if you wanted to, we could make you a module without the analog sections at a discounted price and you could solder wires to the pads of the original components.
The 15V voltages are being fed into two LT3045s which provide power to the opamps and a second set of LT3045s in a role of reference voltage (AVCC) supplies.
Our statement on ES9038PRO:
While it has potential for 6dB higher dynamic range, available measurements at Audio Science Reviews show that converters equipped with it suffer from so called "Sabre hump", a peak in THD+N/IMD vs. amplitude measurement that might suggest some issues, while ES9028PRO doesn't suffer from it, at least in our configuration. Since we don't develop our DACs for good full-scale numbers, but rather for the best possible performance over the whole audio range, we wouldn't be satisfied with that. Additionally, ES9038PRO has no advantage in THD and IMD over ES9028PRO. They are the same chips, the only difference being that ES9038PRO has 4x the output current.
Regarding replacing the I/V sections - not sure what would be the point. The choice of components, very tight layout on 4-layer PCB and the use of fully-differential amplifiers is what makes the DAC8 module perform so well. There are no connectors, but if you wanted to, we could make you a module without the analog sections at a discounted price and you could solder wires to the pads of the original components.
The 15V voltages are being fed into two LT3045s which provide power to the opamps and a second set of LT3045s in a role of reference voltage (AVCC) supplies.
Our statement on ES9038PRO:
While it has potential for 6dB higher dynamic range, available measurements at Audio Science Reviews show that converters equipped with it suffer from so called "Sabre hump", a peak in THD+N/IMD vs. amplitude measurement that might suggest some issues, while ES9028PRO doesn't suffer from it, at least in our configuration. Since we don't develop our DACs for good full-scale numbers, but rather for the best possible performance over the whole audio range, we wouldn't be satisfied with that. Additionally, ES9038PRO has no advantage in THD and IMD over ES9028PRO. They are the same chips, the only difference being that ES9038PRO has 4x the output current.
- Status
- Not open for further replies.