Issue with PCM2707

Hanns

Other Parts Discussed in Thread: PCM2704, PCM2706

Hello!

I'm currently evaluating the PCM2707 and I have it hooked up in bus powered mode according to the datasheet page 29. Since I only need i2s out, I've left out the pins for headphone output, i.e. pins 26 - 31.

Now, a couple of questions:

1. Is it ok to leave all the pins 26 to 31 unconnected? I'm not sure where to find info on this in the datasheet.

2. The chip works well for a couple of minutes playing music from the USB host. Then ^SSPND suddenly goes low and the chip produces no sound after that. After disconnecting from USB host and waiting for a couple of minutes and then re-connecting USB, the chip works well for another short period of time, then ^SSPND goes low again. It seems, the longer I wait before re-connecting USB between failures, the longer the time of correct operation. I have a feeling it is something building up over time, like heat but the chip doesn't get hot or anything. Maybe be clock deviation, I'm using an ordinary crystal 12.000 MHz HC-49/S 30 pF.

Does anybody recognize the described behaviour or have any idea as of what could be causing the problem?

Thank you!

EDIT: I didn't have any 1uF ceramic capacitors so I used electrolytes instead. None of the parts (except the PCM2707) are soldered but placed on a breadboard. Maybe this is the cause, especially concerning the crystal.

over 13 years ago

0 Hanns over 13 years ago

Expert 1005 points

Regarding question no 2 above - I measured the power draw from VBus and found it to be some 95 - 98 mA which is very close to the max limit of 100mA with HOST pin low. After connecting HOST high and thus allowing for max 500mA, the device has been playing correctly for hours. The reason for the higher power draw is that I'm also powering an MCU via an external 3.3V regulator from the same VBus. Without the MCU, power draw is about 27mA from the PCM2707 alone. So question no 2 seem to be resolved!

The first question regarding leaving headphone pins 26 to 31 unconnected still remains though. I guess this is OK since the unit seem to work correctly now but it would be good to know for sure.

Thanks.

EDIT: Nope, regretfully issue no 2 remains - it just took a great deal longer for it to appear this time. Still I get the feeling that something is building up over time...but what anyone?

0 Don Dapkus over 13 years ago in reply to Hanns

TI__Guru**** 189399 points

Hi, Hanns,

For Issue #1, doesn't the /SSPND monitor for bus traffic:

The PCM2704/5/6/7 enters the suspend state after the USB bus has been in a constant idle state for
approximately 5 ms. While the PCM2704/5/6/7 is in the suspend state, SSPND flag (pin 27 for PCM2704/5,
pin 11 for PCM2706/7) is asserted. The PCM2704/5/6/7 wakes up immediately when detecting the non-idle state
on the USB bus

For your second question, you should connect the power and ground to those pins.

-d2

0 Hanns over 13 years ago in reply to Don Dapkus

Expert 1005 points

Don Dapkus said:

The PCM2704/5/6/7 enters the suspend state after the USB bus has been in a constant idle state for
approximately 5 ms. While the PCM2704/5/6/7 is in the suspend state, SSPND flag (pin 27 for PCM2704/5,
pin 11 for PCM2706/7) is asserted. The PCM2704/5/6/7 wakes up immediately when detecting the non-idle state
on the USB bus

Yes, but the bus can't be said to be idle when music is playing plus it doesn't wake up again wihout a power cycle so this is clearly something else. I've ordered ceramic capacitors and a crystal socket to see it that helps. In the mean time, any other ideas of what could be the cause of this?

Don Dapkus said:

For your second question, you should connect the power and ground to those pins.

Ok, maybe this is the reason for the unexpected ^SSPND=0. I'll make sure pins are connected and see if that helps.

Thanks.

0 Hanns over 13 years ago in reply to Hanns

Expert 1005 points

As it turned out, my crystal was bad and after replacing it, everything works great!

0 Don Dapkus over 13 years ago in reply to Hanns

TI__Guru**** 189399 points

Hanns,

CRAZY!!! I would have never thought of that! What made you think of that?

-d2

0 Hanns over 13 years ago in reply to Don Dapkus

Expert 1005 points

Well, I had a feeling something was building up over time and since it wasn't heat then clock drift kind of fitted the description. Good thing I had another crystal.

Audio

Audio forum

Issue with PCM2707