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

[edbk]

Si570 interest list:

1. bigpandahk
2. tagheuer
3. hochopeper
4. qusp (of course)
5. AR2 - definitely!
6. wktk_smile
7. hirez69
8. CeeVee - you bet!
9. number9
10. analog_sa - GB maniac
11. edbk

[atom6422]

Si570 interest list:

1. bigpandahk
2. tagheuer
3. hochopeper
4. qusp (of course)
5. AR2 - definitely!
6. wktk_smile
7. hirez69
8. CeeVee - you bet!
9. number9
10. analog_sa - GB maniac
11. edbk
12. atom6422

[MisterRogers]

Si570 interest list:

1. bigpandahk
2. tagheuer
3. hochopeper
4. qusp (of course)
5. AR2 - definitely!
6. wktk_smile
7. hirez69
8. CeeVee - you bet!
9. number9
10. analog_sa - GB maniac
11. edbk
12. atom6422
13. misterrogers - Of Course!

[iancanada]

Originally Posted by MisterRogers

Si570 interest list:

1. bigpandahk
2. tagheuer
3. hochopeper
4. qusp (of course)
5. AR2 - definitely!
6. wktk_smile
7. hirez69
8. CeeVee - you bet!
9. number9
10. analog_sa - GB maniac
11. edbk
12. atom6422
13. misterrogers - Of Course!

Hi guys, I'm being kidnaped .

Ian

[iancanada]

The biggest challenge to the Si570 clock board project is the power supply. Silabs declared that Si570 has high PSRR performance and not that sensitive to the supply noise, but trust me, that’s not true. It needs more than 100mA power consumption, much bigger than normal XO oscillators. Although ADP151 is rated at 200mA output, but it really can’t keep good low noise performance at this current level.

Using a 3.2V LiFePO4 battery cell (together with the management board) powering si570 directly is a good solution, which I’m running right now. It works fine and sounds great. The only problem is that the output voltage of the battery is not fixed. During the whole discharge progress, voltage will keep dropping from 3.4V to 3.0V. Actually that’s not bad for a battery, but the interesting thing is, si570 has a sweet point on power supply voltage. It will take around half hour to make the battery running into that point and then Si570 becomes more and more sweet.

TPS7A4700 seems to be one of the best alternatives. It rated at 1A output, with only 4.17uVrms noise, industrial best so far. I designed a TPS7A4700 PCB last night and placed the order today. Hopefully it works for Si570.

Ian

[hochopeper]

Originally Posted by iancanada

The biggest challenge to the Si570 clock board project is the power supply. Silabs declared that Si570 has high PSRR performance and not that sensitive to the supply noise, but trust me, that’s not true. It needs more than 100mA power consumption, much bigger than normal XO oscillators. Although ADP151 is rated at 200mA output, but it really can’t keep good low noise performance at this current level.

Using a 3.2V LiFePO4 battery cell (together with the management board) powering si570 directly is a good solution, which I’m running right now. It works fine and sounds great. The only problem is that the output voltage of the battery is not fixed. During the whole discharge progress, voltage will keep dropping from 3.4V to 3.0V. Actually that’s not bad for a battery, but the interesting thing is, si570 has a sweet point on power supply voltage. It will take around half hour to make the battery running into that point and then Si570 becomes more and more sweet.

TPS7A4700 seems to be one of the best alternatives. It rated at 1A output, with only 4.17uVrms noise, industrial best so far. I designed a TPS7A4700 PCB last night and placed the order today. Hopefully it works for Si570.

Ian

Ian, your productivity appears limitless!!

[iancanada]

Originally Posted by hochopeper

Ian, your productivity appears limitless!!

That's my daily job, my hobby takes the advantage of .

[iancanada]

Already apply free samples from TI, just hope it works as good as expected.

Ian

[hochopeper]

Ian, that looks like it could make a nice supply for the 'clean' side of the isolator boards too! (for those who may not want to add batteries).

Though I have to ask, why only one ground connector? Or is that a position for connecting ground wire between C6, C7 and R9?

[iancanada]

Originally Posted by hochopeper

Ian, that looks like it could make a nice supply for the 'clean' side of the isolator boards too! (for those who may not want to add batteries).

Though I have to ask, why only one ground connector? Or is that a position for connecting ground wire between C6, C7 and R9?

You are correct, running clock from a isolated clean PSU is one of the goals.

That three pin connector is to connect directly to the on board socket. Isolator will be powered from Si570 clock board by the FPC/FFC cable, no need any additional connection.

Ian