I think worrying about cables inducing lots of jitter is not too productive. They can add some but not from capacitance or generic EMI. The capacitance would need to change in a random sense which won't happen if the cable is static and made from quality materials, e.g. PE or PTFE. If its shield is decently implemented for RF it won't be susceptible to EMI and more likely would be a source of EMI. The terminations and the circuitry the cable connects to will be more sensitive to EMI AND more likely to be a source of significant EMI if not addressed properly.
Cables can contribute jitter if they are subjected to mechanical vibration. https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=901491 Simple answer- use good stiff cables that will resist vibration. Semirigid coax terminated in SMA's properly matched on the PCB's would be the best option. An example of doing this right is here: ADCs and DACs for audio instrumentation applications The cables are flexible but rigid can be substituted once the rest of the design is final.
Simplest- don't put your DAC anywhere near the speaker.. Think of it as a turntable.
Cables can contribute jitter if they are subjected to mechanical vibration. https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=901491 Simple answer- use good stiff cables that will resist vibration. Semirigid coax terminated in SMA's properly matched on the PCB's would be the best option. An example of doing this right is here: ADCs and DACs for audio instrumentation applications The cables are flexible but rigid can be substituted once the rest of the design is final.
Simplest- don't put your DAC anywhere near the speaker.. Think of it as a turntable.
Hi Demian,
your explanation is useful for many members but not for yunyun, you are losing time with him.
He don't understand that a coaxial cable has static capacitance, so when the transmission line is properly terminated the added phase noise is practically zero.
But unfortunately he don't know what does mean line termination.
He usually throws tons of nonsense statements because he lacks the minimal skill.
Indeed the coaxial cable used to connect the oscillators to other devices and the oscillator itself are EMI source (they are antennas), that's the reason we have suggested to keep the oscillators far from the FIFO/DAC using semi-rigid RG400 coaxial cable SMA terminated to avoid EMI/RFI interferences.
We have also tested the immunity from vibrations measuring the phase noise of the oscillator placed at 20 cm from the bass driver.
Using the aluminum box we have indicated and decoupling the crystal from the board by neoprene sheet and thin copper wires, the oscillator is practically insensitive to the vibrations induced by the speaker driver since the phase noise plot did not change.
The really curious thing is that yunyun states about Ian's limited ability while Ian has developed a lot of stuff FPGA based.
And the even more curious fact is that this statement comes from those who lack the minimum technical knowledge.
But on the other hand this forum is open to everyone, I just hope that no one takes seriously the foolish claims of yunyun.
your explanation is useful for many members but not for yunyun, you are losing time with him.
He don't understand that a coaxial cable has static capacitance, so when the transmission line is properly terminated the added phase noise is practically zero.
But unfortunately he don't know what does mean line termination.
He usually throws tons of nonsense statements because he lacks the minimal skill.
Indeed the coaxial cable used to connect the oscillators to other devices and the oscillator itself are EMI source (they are antennas), that's the reason we have suggested to keep the oscillators far from the FIFO/DAC using semi-rigid RG400 coaxial cable SMA terminated to avoid EMI/RFI interferences.
We have also tested the immunity from vibrations measuring the phase noise of the oscillator placed at 20 cm from the bass driver.
Using the aluminum box we have indicated and decoupling the crystal from the board by neoprene sheet and thin copper wires, the oscillator is practically insensitive to the vibrations induced by the speaker driver since the phase noise plot did not change.
The really curious thing is that yunyun states about Ian's limited ability while Ian has developed a lot of stuff FPGA based.
And the even more curious fact is that this statement comes from those who lack the minimum technical knowledge.
But on the other hand this forum is open to everyone, I just hope that no one takes seriously the foolish claims of yunyun.
Hello Andrea,
My friend who just like me wanna use your '' clock boards '' in the near future just finished a more standard clock board using some '' tricks '' to isolate it.
So we can consider the board as a kind of membrane that will pick up vibrations and transfer them to the clocks. SO neoprene and thin wire is a good solution.
Most cables connected to the board will pick up/ transfer vibrations too so tackle the problem at the clock itself will be the most rewarding.
I still have some small sheets of copper, and also have copper foil. I think we will be able to make an almost '' airtight '' box just with a few opening for the cables. If i am right a smaller and a more closed box will stop most interference and keep temperature inside at the same level.
Where can i find the prices of the parts you pictured in the drawing with the DDDAC added that you posted some time ago?
Greetings, Eduard
My friend who just like me wanna use your '' clock boards '' in the near future just finished a more standard clock board using some '' tricks '' to isolate it.
So we can consider the board as a kind of membrane that will pick up vibrations and transfer them to the clocks. SO neoprene and thin wire is a good solution.
Most cables connected to the board will pick up/ transfer vibrations too so tackle the problem at the clock itself will be the most rewarding.
I still have some small sheets of copper, and also have copper foil. I think we will be able to make an almost '' airtight '' box just with a few opening for the cables. If i am right a smaller and a more closed box will stop most interference and keep temperature inside at the same level.
Where can i find the prices of the parts you pictured in the drawing with the DDDAC added that you posted some time ago?
Greetings, Eduard
I think that the 3rd version of the Fifopi doesn’t have this 3.3v regulator? The reason that I ask this is because I do have 5v crystals and a high quality 5v psu. If I upgrade to Fifopi version 3 it would safe the costs of a new 3.3v psu
what power supply can do 0,0005 Ohm output (0,5 mOhm... which is a typical PCB trace of 1 mm wide and 1mm "long")?
and what is the bandwidth for this spec ?
Last edited:
As I reported earlier, I found with the earlier gen Driscol that isolating from vibration was the most rewarding thing you can do that results in audible improvement. I like the approach with the floating crystal using the rubber and fine wires together with the Hammond case. I plan to take one more step. I am going to put that entire assembly into a larger case containing the power supply and encase the Hammond case in sand. I have found sand to be very effective at isolating from vibration in the room. Probably overkill as Andrea points out his measurements were taken with the case specified. Still will be an interesting experiment.
@dennusch
It won't be any damage but not recommended. It would be better to use external power supply for XOs if they run at 5V.
Ian
MAXWELL 3000F ultracapacitor has 0.29 mΩ ESR
Ian
Hello,
Take a look at those '' high end '' nice aluminium engraved boxes which contain one or more ultracaps and which will give you only the optimum results with special cables and fuses. The '' energy '' leaving these caps will meet a lot of obstacles before reaching the circuit demanding the energy.
You can use a smaller brother of the Maxwell right next to the circuit so with 3 inches of wire or a slick 700 $ upgraded one at 2 feet distance with the big one.
Greetings, Eduard
Really happy with my Raspian music streamer, great work Ian 
Raspian (Ian Canada) streamer (full options) (Spotify / MoOde)
Audio-GD R8 (2021) DAC (i2s)
Audio-GD Master 1 Vacuum XLR preamplifier
Mordaunt Short Performance 6 Limited Edition loudspeakers (converted to full active)
6x NCore poweramp's (Two Hypex FusionAmp's FA253)
Rythmik F18 subwoofer
Little Fugue in G minor - YouTube
Raspian (Ian Canada) streamer (full options) (Spotify / MoOde)
Audio-GD R8 (2021) DAC (i2s)
Audio-GD Master 1 Vacuum XLR preamplifier
Mordaunt Short Performance 6 Limited Edition loudspeakers (converted to full active)
6x NCore poweramp's (Two Hypex FusionAmp's FA253)
Rythmik F18 subwoofer
Little Fugue in G minor - YouTube
@dennusch
FifoPi will make the sound as rich and detail as the original muse. But it doesn't make the music "thick". If you want that style, you can try suitable amplifiers, cables and speakers after DAC. The most significant thing of a FifoPi is to keep the music as close as possible to what it should be
.
Regards,
Ian
FifoPi will make the sound as rich and detail as the original muse. But it doesn't make the music "thick". If you want that style, you can try suitable amplifiers, cables and speakers after DAC. The most significant thing of a FifoPi is to keep the music as close as possible to what it should be
.Regards,
Ian