Member
Joined 2006
To save some time on this discussion, here is the long thread documenting the develepment journey of the sdtrans384 sd card player and the 9018 syndac by Bunpei.
www.diyaudio.com/forums/digital-source/142562-microsd-memory-card-transport-project-50.html
Its one exceptionally fine digital files playback system, really quite amazing. I had the pleasure of auditioning it at the MJ Audio Festival in 2017 that Bunpei mentioned and reported the event on my webpage.
www.diyaudio.com/forums/digital-source/142562-microsd-memory-card-transport-project-50.html
Its one exceptionally fine digital files playback system, really quite amazing. I had the pleasure of auditioning it at the MJ Audio Festival in 2017 that Bunpei mentioned and reported the event on my webpage.
The primary purpose of my initial post is not to argue “CD-player versus PC” but to express our comment on the superiority of ES9038PRO and our gratitude to designers and manufacturers of the DAC chip.
I can just say, the more you make efforts on eliminating a various kind of noises, the better music the chip plays.
I can just say, the more you make efforts on eliminating a various kind of noises, the better music the chip plays.
Impressive work! Your 9038 output is current mode, driving active speaker amps without transformers?
Regarding noise, I am certainly not a purist. However, the sound I hear via I2S coming through opto-couplers is no different whether the I2S source is complex and noisy or simple and presumably quiet. Do you include isolators or have you tested optocoupling of I2S/DSD in your system?
My recent configuration includes no transformers at the output side of the DAC. I’m not sure there is no transformer in the analog signal path because ADAM A7X may have a line transformer at its input for isolation.
Neither I’m not sure that the configuration is “current mode” because the A7X must have input impedance 300 ohm for XLR input.
The interface for I2S/DSD-raw between SDTrans384 and ES9038PRO Dual Mono DAC adopts LVDS, incorporating isolators.
Neither I’m not sure that the configuration is “current mode” because the A7X must have input impedance 300 ohm for XLR input.
The interface for I2S/DSD-raw between SDTrans384 and ES9038PRO Dual Mono DAC adopts LVDS, incorporating isolators.
Bunpei
That is an amazing effort my friend.
I love your extreme approach to every detail of this DAC. People like you
make this forum worthwhile.
cheers
Terry
Bunpei,
I don't think the A7X has an input transformer. Generally speaking pro audio,
line level input gear for 'transparency' these days stay away from signal
transformers.
Terry
Bunpei,
WRT AES67, is this Ethernet protocol for your SS transport?
Do you know of any other (readily available) hardware that supports 384k using AES67?
Terry
Hi, Terry!
I appreciated your comments very much!
Yes, AES67 is a specification built on Ethernet technology. I have no devices that support the interface. I have just tried it a little bit by using Merging Pyramix/HAPI. It covers multichannel 384k and DSD256.
The Adam A7X user manual says the input impedance is 30kohms, and that it uses switch-mode amplifiers. Probably the best amplifiers of that type are those made by Bruno Putzeys.
Regarding input impedance for professional audio balanced inputs, usually they are 'bridging' which means high impedance. That is so multiple loads could be driven by one one line amp. If desired, a 600-ohm terminator could be added to the last device in the chain. Some amplifiers used to have a switch on the input panel to add a 600 ohm termination resistor if desired, but that practice seems to have become very rare in modern times.
Therefore, most likely the DACs are operating in voltage mode. Doesn't matter I guess if it sounds good then it sounds good, right?
Regarding input impedance for professional audio balanced inputs, usually they are 'bridging' which means high impedance. That is so multiple loads could be driven by one one line amp. If desired, a 600-ohm terminator could be added to the last device in the chain. Some amplifiers used to have a switch on the input panel to add a 600 ohm termination resistor if desired, but that practice seems to have become very rare in modern times.
Therefore, most likely the DACs are operating in voltage mode. Doesn't matter I guess if it sounds good then it sounds good, right?
I appreciated your indication very much!
Yes, the Adam A7X user manual shows input impedance 30 kOhms.
The active speaker has both unbalanced input of RCA connector and balanced input of XLR connector. I can't believe both have such an identical high impedance.
Moreover, an usual impedance value representation for XLR input might show an impedance between Hot and GND (and Cold and GND). In my case, an impedance value should be one between Hot and Cold. I speculate the value might be different. (However, I'm not sure.)
I expect your opinions!
Hi Bunpei,
No technical reason why they could not make the unbalanced and the balanced inputs both have 30k input impedance if they want to. Since they only give one number for input impedance in the manual, 30k ohms, it would appear it applies to both types of inputs. For the balanced input I would expect the input impedance to most likely be 30k ohms from hot to hot, and 15k ohms from hot to ground.
One could of course measure it. For example, if 30k ohms were put in series with an input then the voltage at the input should be divided in half, due to voltage divider effect. Easy enough to try it with a 30k resistor and an AC voltage multimeter (select a meter which is accurate at the input frequency to be used for the test). Many modern multimeters are still accurate at 1kHz, although the same test should also work at 60Hz.
Or, one could merely trust the speaker owner's manual. Most likely it is correct.
Regards,
Mark
No technical reason why they could not make the unbalanced and the balanced inputs both have 30k input impedance if they want to. Since they only give one number for input impedance in the manual, 30k ohms, it would appear it applies to both types of inputs. For the balanced input I would expect the input impedance to most likely be 30k ohms from hot to hot, and 15k ohms from hot to ground.
One could of course measure it. For example, if 30k ohms were put in series with an input then the voltage at the input should be divided in half, due to voltage divider effect. Easy enough to try it with a 30k resistor and an AC voltage multimeter (select a meter which is accurate at the input frequency to be used for the test). Many modern multimeters are still accurate at 1kHz, although the same test should also work at 60Hz.
Or, one could merely trust the speaker owner's manual. Most likely it is correct.
Regards,
Mark
Last edited:
Dear Mark,
I have realized that my initial understanding was incorrect. I had believed that XLR was of impedance 600 ohms for both output and input.
Now, I know that a modern XLR input impedance is 10k ohms - 50k ohms.
I appreciated your indications very much!
In spite of the high input impedance, I perceive good SQ in my present configuration.
Best regards,
Bunpei
I have realized that my initial understanding was incorrect. I had believed that XLR was of impedance 600 ohms for both output and input.
Now, I know that a modern XLR input impedance is 10k ohms - 50k ohms.
I appreciated your indications very much!
In spite of the high input impedance, I perceive good SQ in my present configuration.
Best regards,
Bunpei
Hi,
I hope nobody minds that I jump in here with my question.
I got a populated diyinhk es9018 (the 8 channel not k2m) board without es9018 from my collegue at work. He has fried the chip. I got the chip and had it soldered in in an
university electronic lab by people who do this kind of soldering on a regular basis. They have also replaced the quartz.
I have connected the usb xmos interface, psu and my MUTE control is always on and no music.
The voltages seem to be ok everywhere.
I have checked with the oscilloscope that the data is flowing from the xmos (diyinhk) board (apart from data pins, I have seen nice 44.1khz square wave on one of the pins!) Obviously I cannot check with the oscilloscope whether the data makes sense ,
but I managed to get sound out of diyinhk AK4399 board using the same usb xmos board.
Also I cannot check whether the quartz is working properly, but I do get some waves on the scope (my scope Hantek6022BE cannot handle 75Mhz quartz freq, but I presume I observe some interference, so I do not conclude that the quartz is dead).
I have seen some people could not get music out unless they pressed reset on the board. I tried that with no luck.
I am running out of ideas what and how to test. Do you think the es chip is dead?
Best regards
Greg
I hope nobody minds that I jump in here with my question.
I got a populated diyinhk es9018 (the 8 channel not k2m) board without es9018 from my collegue at work. He has fried the chip. I got the chip and had it soldered in in an
university electronic lab by people who do this kind of soldering on a regular basis. They have also replaced the quartz.
I have connected the usb xmos interface, psu and my MUTE control is always on and no music.
The voltages seem to be ok everywhere.
I have checked with the oscilloscope that the data is flowing from the xmos (diyinhk) board (apart from data pins, I have seen nice 44.1khz square wave on one of the pins!) Obviously I cannot check with the oscilloscope whether the data makes sense ,
but I managed to get sound out of diyinhk AK4399 board using the same usb xmos board.
Also I cannot check whether the quartz is working properly, but I do get some waves on the scope (my scope Hantek6022BE cannot handle 75Mhz quartz freq, but I presume I observe some interference, so I do not conclude that the quartz is dead).
I have seen some people could not get music out unless they pressed reset on the board. I tried that with no luck.
I am running out of ideas what and how to test. Do you think the es chip is dead?
Best regards
Greg
There are no many other active components (in the digital part) on the board.
Just 3.3V->1.2V converter which seems to produce 1.2V as required.
The chip documentation is here: https://myl8test.files.wordpress.com/2013/05/download-es9018-datasheet-here.pdf
I do not have the schematics of the board.
The thing that puzzels me is the reset pin on the chip.
in the Buffalo III docs it says reset pin is active low.
Somewhere here somebody has posted some pics from an oscilloscope showing that the reset pin is high for a moment on power up and power down.
As far as I can tell on the diyinhk board the reset pin is connected through 3k3 resistor to the ground and the reset switch pulls it high to 3V3.
If I read it correctly in ES9018 | Dimdim's Blog dimdim says his buffalo board came alive when he shorted the reset pin:
My eye fell on the section pertaining to the Reset pin. It stated that it was an active-low pin and that a system reset could be performed either by pulling the pin low or by a software command. The Buffalo III design called for this pin to be pulled low by default so I hadn’t paid any real attention to it. It turned out that I should have. I soldered a 2 pin header and put a jumper on it. The board immediately came alive and was detected by my I2C scanner. 😀 So, the Reset pin should be pulled up.
In my case the reset pin is low.
Just 3.3V->1.2V converter which seems to produce 1.2V as required.
The chip documentation is here: https://myl8test.files.wordpress.com/2013/05/download-es9018-datasheet-here.pdf
I do not have the schematics of the board.
The thing that puzzels me is the reset pin on the chip.
in the Buffalo III docs it says reset pin is active low.
Somewhere here somebody has posted some pics from an oscilloscope showing that the reset pin is high for a moment on power up and power down.
As far as I can tell on the diyinhk board the reset pin is connected through 3k3 resistor to the ground and the reset switch pulls it high to 3V3.
If I read it correctly in ES9018 | Dimdim's Blog dimdim says his buffalo board came alive when he shorted the reset pin:
My eye fell on the section pertaining to the Reset pin. It stated that it was an active-low pin and that a system reset could be performed either by pulling the pin low or by a software command. The Buffalo III design called for this pin to be pulled low by default so I hadn’t paid any real attention to it. It turned out that I should have. I soldered a 2 pin header and put a jumper on it. The board immediately came alive and was detected by my I2C scanner. 😀 So, the Reset pin should be pulled up.
In my case the reset pin is low.
Looks like holding reset low keeps the chip in reset mode. There is an RC circuit that holds the pin low when power is first applied until the cap charges up enough (assuming the jumper if needed is in place to allow pullup).
Also, there is not just a single 3.3v supply. A schematic for an earlier ESS board for an ES9008 chip can be found here: http://www.esstech.com/files/5114/4095/4310/Sabre_8_2Channel_64PIN_V3_SCH.pdf
Of all the supplies, AVCC (AVCC_R and AVCC_L) is the most critical for sound quality.
Do you have an I2C sniffer? They are very cheap like around $5 or $10 on ebay. Maybe a little more on Amazon. New USB Logic Analyzer Device Set USB Cable 24MHz 8CH 24MHz for ARM FPGA M100 7000000492822 | eBay
First, it would be good to make sure every single power supply is working. Then there is normally a microcontroller that configures the dac chip over I2C bus. If all the power is up (and reset pin is up) maybe the next thing would be to check what the microcontroller is doing.
If you know how to use an Arduino, there are I2C libraries that can allow you to access the dac chip control registers. You would need to know a little more about how to do that though. I could point you to some info is there is any interest.
Also, there is not just a single 3.3v supply. A schematic for an earlier ESS board for an ES9008 chip can be found here: http://www.esstech.com/files/5114/4095/4310/Sabre_8_2Channel_64PIN_V3_SCH.pdf
Of all the supplies, AVCC (AVCC_R and AVCC_L) is the most critical for sound quality.
Do you have an I2C sniffer? They are very cheap like around $5 or $10 on ebay. Maybe a little more on Amazon. New USB Logic Analyzer Device Set USB Cable 24MHz 8CH 24MHz for ARM FPGA M100 7000000492822 | eBay
First, it would be good to make sure every single power supply is working. Then there is normally a microcontroller that configures the dac chip over I2C bus. If all the power is up (and reset pin is up) maybe the next thing would be to check what the microcontroller is doing.
If you know how to use an Arduino, there are I2C libraries that can allow you to access the dac chip control registers. You would need to know a little more about how to do that though. I could point you to some info is there is any interest.
Last edited:
Hi Mark,
This is very useful info. I will study it (I am a slow diyer). I am aware that there are more power supplies inputs on the board. As far as I can tell all of them are working properly.
I do not understand this part:
>Then there is normally a microcontroller that configures the dac chip over I2C bus.
I understand that you can control the board with an arduino. That is my next goal. At this stage I would like to get it running as is. As far as I remember (I had this board - well maybe not the one that has been fried my my collegue, but the same model - on a loan some time ago) it worked as I have it now ie without an arduino.
I will have the bus pirat tomorrow. Shall I check the data sent from xmos with it? Is it what you meant?
I have arduino lying around for this project. So if you can direct me what to do to read the dac registers I certainly would try this!
Many thanks
Greg
This is very useful info. I will study it (I am a slow diyer). I am aware that there are more power supplies inputs on the board. As far as I can tell all of them are working properly.
I do not understand this part:
>Then there is normally a microcontroller that configures the dac chip over I2C bus.
I understand that you can control the board with an arduino. That is my next goal. At this stage I would like to get it running as is. As far as I remember (I had this board - well maybe not the one that has been fried my my collegue, but the same model - on a loan some time ago) it worked as I have it now ie without an arduino.
I will have the bus pirat tomorrow. Shall I check the data sent from xmos with it? Is it what you meant?
I have arduino lying around for this project. So if you can direct me what to do to read the dac registers I certainly would try this!
Many thanks
Greg
No, not XMOS. Either there would be some microcontroller chip on the board or an I2C bus connector for one. The dac chip does have some default behavior if no microcontroller configures the registers, but you would have to look at the data sheet to see what all the register defaults would do in that case. Boards that don't have a microcontroller on the main board usually come with another small microcontroller board to that plugs into the I2C bus socket, maybe with a display and some buttons and knobs, whatever, to operate the dac.
Regarding using an Arduino, I use one to take over control of an ES9038Q2M dac board. Some pictures and info can be found here: http://www.diyaudio.com/forums/digital-line-level/314935-es9038q2m-board-142.html#post5452170
http://www.diyaudio.com/forums/digital-line-level/314935-es9038q2m-board-78.html#post5412858
Regarding using an Arduino, I use one to take over control of an ES9038Q2M dac board. Some pictures and info can be found here: http://www.diyaudio.com/forums/digital-line-level/314935-es9038q2m-board-142.html#post5452170
http://www.diyaudio.com/forums/digital-line-level/314935-es9038q2m-board-78.html#post5412858
Last edited:
Mark,
Before I hook up the arduino (the board does not have a ucontroller), I still would like to run the dac in the hardware mode. It seems that many people at least test like this.
The reset pin still bothers me: not many people even install the reset jumper (between DVCC and the RC network 3k3 and 0.1uF - here my board looks like the ES9008 schematics you kindly pointed me to). How on earth would it ever go high?
Before I hook up the arduino (the board does not have a ucontroller), I still would like to run the dac in the hardware mode. It seems that many people at least test like this.
The reset pin still bothers me: not many people even install the reset jumper (between DVCC and the RC network 3k3 and 0.1uF - here my board looks like the ES9008 schematics you kindly pointed me to). How on earth would it ever go high?