Another PCM2707 problem...

[Akira386]

Hi, I built a PCM2707 based USB DAC, but WinXP don't recognize it .
It says "usb device can't be recognized due to malfunction", or something like .

Attached is the schematic. 2707 is used in self powered mode, feed by two LDO regulators. Voltages on the regulators are ok, but pin 32 (Vcom) is 0V.
I can't believe I've killed the PCM2707, because I soldered it very carefully.

Any hint on what may be wrong? TIA.

[Akira386]

Is pcm2707 very sensible to the ESD? More than other chips, I mean .
My soldering iron is not grounded... I'm starting to think I've killed the PCM by ESD, although I never have made it before, soldering opamps.

Appears good, but doesn't work .

[danb1974]

Other than checking the digital part (grounds, d+, d-, vbus, vdd, host, psel, xtal, associated resistors), no idea.

[Brent Welke]

Hi Akira386.

I took a look and compared your schematic to my schematic. I use the PCM2706 in the I2S mode so there are differences but from what I see yours should work.

Have you had your circuit working as a prototype so you know it will work? Or maybe this is your first prototype.

Could it be that one of the pins is not soldered? I would guess you have checked continuity with a meter already.

If you have a scope check to see if you have any signal on the clock pins. If you do then you know the chip is not dead. If you don't then it tells you nothing.

When I prototyped my DAC I ended up with two or more dead DAC chips. I never knew why and I still don't. I was pretty careless and prototypes had wires everywhere so that is the nature of prototyping. But on production DAC board I have not had a single failure in 7 dacs and I take no precautions with what order I plug in the mains or USB cable or how long the DAC sits with out a USB connection.

What I suggest is that if you have nothing left to try then you replace the DAC chip. I have done this many times. A few times I have soldered the PCM2707 in the wrong way and replacing chips on prototypes. First apply extra solder to the pins. Take a sharp utility blade and cut the pins at the chip body. Clean up with desoldering wick and solder in a new chip. Careful as too much heat to the pads will make them lift. But even then you can fix it with jumper wires. Don't give up. You are very close

Let me know if I can be of any further assistance.

[jkeny]

In self power mode if you connect the power to the chip before/without the USB cable connected you will fry the chip

Have you done this?

John

[jleaman]

Originally posted by jkeny
In self power mode if you connect the power to the chip before/without the USB cable connected you will fry the chip

Have you done this?

John

How is this possible for the way he has it shown.. You can't..

[Brent Welke]

jkeny: In self power mode if you connect the power to the chip before/without the USB cable connected you will fry the chip

Not true for me. I do it every day. I have probably done that 200 times by now and never a problem. I wonder if the problem isn't when the chip is used in BUS powered mode. Maybe the TI datasheet is wrong?

[Brent Welke]

Maybe the problem is related to timing. In my DAC the PCM2706/7 is powered from the mains. The PCM2707 has power long before the USB connection is made. But in Akira386's case it is happening at the same time.

I also do not have a problem if I disconnect the mains power first and then disconnect the USB cable or connect the USB cable before the mains.

But come to think of it... I have a prototype (looking at it right now, just plugged it in and it works) that is rigged in self power mode but gets power from the USB bus. Same as Akira386.

The mystery continues....

[paulb]

I've blown a PCM2702 for no apparent reason. It worked for a while, then phffft! Replaced the part and it has worked for over 6 months now.

[Akira386]

Thank you for the input, guys!

Originally posted by Brent Welke
Have you had your circuit working as a prototype so you know it will work? Or maybe this is your first prototype.

It's my first prototype, Brent. I was confident that it would work, it is so simple, but...

If you have a scope check to see if you have any signal on the clock pins. If you do then you know the chip is not dead. If you don't then it tells you nothing.

The only instrument I have is a DMM... Measuring from the GND, it has 3V in the XTI and 3.3V in the XTO .

When I prototyped my DAC I ended up with two or more dead DAC chips. I never knew why and I still don't.

This is what I fear..
Maybe it is time to replace the PCM.
Thank you for the suggestion and incentive, Brent.

Originally posted by jkeny
In self power mode if you connect the power to the chip before/without the USB cable connected you will fry the chip

Have you done this?

John, the LDO regulators are powered by the USB bus, so it is impossible to provide power to the PCM while dettached.


akira