DIYHiFi.org

Well, you have made a visually appealing part, but you might be on the wrong forum for extolling the virtues of non-os TDA1543 . Many people come/came here to escape the sea of non-os / gainclone posts over at diya.

With regard to the whole "ipod functionality" thing, this applies to every dac ever made for the most part... so I give up.

For comparisons you might want to look at one of the trillion other PCM270x + TDA1543 NOS reviews. Not that I doubt yours sounds different, but when everyone uses the same rx chips and the same dac with no filter there are bound to be many similarities.

I'm sure you'll sell lots of kits though, those unsleeved capacitors + blue pcb and dual transformers/IEC will drive the average asylum digital drive reader nuts.

Hello All,

Someone brought to my attention that D+ pulled-up to a self-powered bus (i.e. not VBUS coming from USB) is dangerous and can ruin the IC if the USB bus is powered-down (e.g., when disconnecting the USB connector) and the D+ is still pulled high via its direct connection to the self-powered bus. My knee-jerk reaction was to disagree given so many circuits here and elsewhere showing pull-up to a self-powered bus.

But in looking at page 29 of the PCM2706 datasheet (http://focus.ti.com/lit/ds/symlink/pcm2706.pdf), it seems that D+ may have to be pulled up to VBUS instead of a self-powered bus. For example, D+ is pulled up to the "I2S I/F Audio Device" which VBUS also is connected to. Also, the text at the bottom of the diagram on page 29 indicates "D+ pullup must not be activated while detaching from host." Anyone care to comment?

Also, I've tried emailing TI and have not gotten a response. I believe Gordon (or John?) earlier in this thread talked about having extensive email communications with someone from TI. Can someone email me contact info of someone at TI that may be able to shed some light on these issues?

Thanks,
Bryan

A friend believes he may have fried a PCM2707 because of this very issue. In order to protect myself, I've wired up a 5VDC coil SPST-NO power relay, in series with the main power switch. The coil is connected to VBUS in parallel with the DAC, so that it only closes the relay is only closed when the device is connected to USB. Now, in order for the DAC to work - both the main power switch has to be turned on and it has to e plugged into a USB port.

I do have some concerns however that leaving it plugged in to USB only may also cause damage. I was having some difficulty initially with the DAC. I disconnected it from mains power, and just left it plugged into USB - thinking everything would be okay, but the chip ended up getting very hot. It turns out I did have some small bridges that I hadn't noticed at the time which I cleaned up after replacing the initially PCM2707, so one of those brides may have caused the problem - but it also may not be. I just don't take the risk. I turn the DAC off and disconnect the USB cable to be safe. Removing 32QTFP chips is just too much of a pain in the ass with a $8 40 watt soldering iron.

Yeah, in reading the datasheet again and thinking more about it, this is an issue that should be addressed. So if the line is pulled up by a supply that is not de-powered when USB is disconnected, something should be implemented to let the line down when USB is disconnected.

I was pointed to Deode's I2S board, which uses an AND gate... it seems simple enough to implement. To minimize "pollution" to this thread and in hopes of stirring a broader discussion, I started a new thread (here: http://www.diyhifi.org/forums/viewtopic.php?p=11277) to discuss the use of an AND gate.

Regards,
Bryan

What do you have the USB 5V routed to? If I understand you correctly, it should only be connected to HOST...

Regards,
Bryan

I have some experience with the "D+ pulled-up" problem or lack of it.

I don't have a problem. I have 7 of these dacs that I plug/unplug in any random order without care as to what order. Main power on or off, USB connected or not. I do lots of plugging/unplugging as I compare differences between dacs. The DACs that are not currently being tested are left plugged into mains power to help with the break in process.

When I was prototyping I had a couple dead pcm2707 chips but that could have been for many reasons. Perhaps this is really more of a problem when the chip is BUS powered. Maybe the TI datasheet is wrong.

_________________
Brent Welke

I do the same thing with my 2704s. And they are bus powered.

jh

PS - I just got a 2906 fired up.

_________________
Jim Hagerman