Asynchronous I2S FIFO project, an ultimate weapon to fight the jitter

[iancanada]

Since batteries were widely approved as one of best low noise and easy to use power supply for oscillators and other low jitter applications, I decide to design a battery management board.

I couldn't put up with battery as a testing hookup anymore. I need my battery based power supply working as standard equipment in my system. This battery management board may have the following features:

1. Comes with an on/off control interface. Could use the system power or a switch as the control signal.
2. Comes with an external charger interface
3. Compatible with different type of battery or different voltage.
4. Isolate output with charger and control circuit.
5. Pure battery operating without internal ground connection.
6. Support multi-board configure to setup battery based power supply groups.
7. Could stack on top of the clock board or stack over each other as an array.
8. Batteries from different board could share a same charger if they come with same voltage and chemical type
9. Optional battery monitoring interface
10. Simple pure passive design avoids any additional noise introduced by active components.
11. DIY friendly and many options for different applications.
12. Batteries in parallel are available.
13. Charger will be bypassed during working time.
14. Easily being integrated into system.

Any suggestions or comments?

Ian

[qusp]

ok so this is allowing to break the i2s ground from fifo to XO board? i'm off to bed, i'll have a think and send you some ideas and designs, like the circuits i'm using for the same purpose in my portable dac

uses ltc2935 battery monitor, adjustable thresholds for lifepo4, lipoly, low charge indication relay driver.

I use external 10 cel fast charger/balancer from FMA, the cellpro 10XP

[1audio]

Ian:
Most of the charging circuits for Li batteries are much more complex and active than a simple trickle charger. Reading through the specs I see hits of DC-DC converters which can be a major noise source. You can synchronize them with the system clocks, which gets complex when the clocks change. Linear Technologies has a decent low noise converter that doesn't regulate but is expensive to implement. You also need magnetics in this case transformers and rectifiers. It will require a lot of careful management of emi to get good results but it can be a really useful device. I would start with TI's portfolio of battery management chips. They also have a good selection of charge pumps that can be synchronized.

I started on a very similar DC management circuit for a USB audio application, not DC isolated but needing multiple split supplies. I can supply what I had done if it helps. . .

[soundcheck]

Batterie-charger:

1. I'm also running a Buffalo with LiPoFes. It would be nice to have multiple
outlets coming witch such a charger. This way you can charge your
device and e.g a DAC supply with the same charger. You just turn the
switch and all batteries are charged at the same time.
2. I'm also running bridged LiPoFes to feed a 5V regulator.
Some kind of bridging option might be nice.

Such a charger is definitely missing in the DIY word.

Cheers

[AR2]

Thank you Qusp for detailed answer.

Best
AR2

[AR2]

Originally Posted by iancanada

Panel interface schematics just for reference.

Ian

Thank you Ian.
AR2

[AR2]

Originally Posted by marce

I cant understand the fasination with i2S, it is the audio equiv of i2c, both designed to be a local digital bus on a single PCB. If people want to send signals long distance use a deidicated ethernet interface, or SPDIF, its surprising how many interfaces use manchester based encoding for digital transmission where a seperate clock line is not provided. Of course trying to get a clock signal long distance on its own is throught with problems, hence LVDS and manchester type encoding are frequently used. If the supporters of i2S look around you will see very few digital interfaces designed to go off board (ie some distance)these days that have a seperate clock line (the days of RS 232 are long gone).
ian has provided one a well thought out and proffesionaly implemented solution to your demons (ie Jitter)that I have seen in the DIY community. i2s would cause you much more problems, you would still have to try and recover what was left of the clock, never mind the problems of sending the data long distance.
Again ask yourselve why ALL digital interfaces that go off board these days are manchester based encoded and LVDS.

Originally Posted by qusp

the fascination has its roots in fact, but as per usual its been blown out of proportion and ignores significant factors. if you are having to actually USE the clock recovered from spdif for the MCK and its fairly local transmission, it makes sense to avoid the extra conversions (this is where the fabled higher jitter comes from) but this assumes a well implemented and LOCAL i2s connection to beat good spdif, this is disregarding the fact that the result is the same here unless you need speeds higher than 192 and/or only have a USB->i2s convertor as source

i2s is far more prone to problems, is more difficult to isolate and when we have such a powerful leveler as we have here, it really makes no sense. Sure if you have to send the data from 20M away and the easiest output is i2s, the transporter (which is basically LVDS marce) is the logical choice, but otherwise just using this rather fantastic spdif board, is the only choice that makes sense IMO.

I am joining both of you guys with this opinion. It makes all sense with internal I2S source ( USB to I2S board), but for any transportation longer than few inches benefits are lost. I2S has a clear advantage when conversions to and from S/PDIF are voided and that is where popularity stems from. But, short distance requirement, coupled with need for very particular high quality wiring and coupling, opens up I2S as a wrong choice in many situations. As for the choice of S/PDIF signal and its integrity I would always prefer balanced approach, which kind off makes it AES/EBU.

[AR2]

Here is an interesting project for Si570, but more interestingly check at the bottom of the page, some very impressive measurement.

Si570 Kit from K5BCQ

[TheShaman]

Ian, your swiss-army-knife-of-chargers would really come in handy for a variety of projects!

[iancanada]

Originally Posted by 1audio

Ian:
Most of the charging circuits for Li batteries are much more complex and active than a simple trickle charger. Reading through the specs I see hits of DC-DC converters which can be a major noise source. You can synchronize them with the system clocks, which gets complex when the clocks change. Linear Technologies has a decent low noise converter that doesn't regulate but is expensive to implement. You also need magnetics in this case transformers and rectifiers. It will require a lot of careful management of emi to get good results but it can be a really useful device. I would start with TI's portfolio of battery management chips. They also have a good selection of charge pumps that can be synchronized.

I started on a very similar DC management circuit for a USB audio application, not DC isolated but needing multiple split supplies. I can supply what I had done if it helps. . .

Hi Demian,

I don't need the battery being charged while it's powering the oscillator. So, I think all the functions could be implemented by a DPDT relay .

I will use a standard external charger or a trickle charger and will disconnect it if the battery power is on. I'm trying to make a very simple design with all passive components to avoid introducing EMI noise. Just to see if I could.

Regard,

Ian