High Performance WM8741 Upsampling DAC New Version build thread (show 'n tell too)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Code done...

as far I will take it now.

Arduino 1.0
Apple remote (Vol+/-, dim)
Rotary encoder (plz check if the debouncing delays are in the right places)
Switch for filter changing

Didn't do: Remote filter selection, remote or rotenc rate selection, brightness or contrast...

Brgds

The code:

The spiderDAC:
 

Attachments

  • WM8741_DAC_4-9-2012.zip
    6 KB · Views: 59
  • bild.JPG
    bild.JPG
    549.1 KB · Views: 136
Last edited:
Looking good Turbon.

It looks like my 2707 is dammaged. Probably ESD when assembling it. I lifted pin 17 and still have no output.

I can plug/unplug the USB cable and see HOST toggle.

I can feed S/PDIF input from the CD player and trace the signals through the SRC4192 which generates data and clock from the S/PDIF signal. These pass to the WM8741.

However, I also get no output from the WM8741. (Actually I get a lot of digital noise.) Measuring V3.3D I only get 1.8V so something is wrong there also. Removing the ferrite bead in line with the 3.3V supply, it goes to 3.3V at the regulator.
 
The day I was installing the ICs was cold and dry here. I had a wrist strap on, but the plug pulled out of the ground socket on the bench and I didn't realize it till I had finished installing the parts. Even with antistatic wax on the floor, rolling a chair around a bit can generate a considerable static charge.

I may wire in an additional 3.3V supply to the WM8741 with greater current capacity to see what happens.

Otherwise I'll just have to wait till I can get replacement parts.
 
I'm going back through the schematic and data sheets and I'm wondering if I'm trying to put the board in an invalid state.

Has anyone successfully used the board without the Arduino connected and driven audio out to either another device or headphones?

If so, what was your data source? USB or S/PDIF?

In addition, can anyone hazard a guess as to why HOST is connected to D+ instead of VBUS as recommended in the TI data sheet?

Thanks.
 
I'm going back through the schematic and data sheets and I'm wondering if I'm trying to put the board in an invalid state.

Has anyone successfully used the board without the Arduino connected and driven audio out to either another device or headphones?

Yes, I have ;-) But I didn't like the SQ from the connector before the volume potentiometers. I will revisit this now when I can steer the SW filters. The headphone output sounded better.

If so, what was your data source? USB or S/PDIF??

USB only...

In addition, can anyone hazard a guess as to why HOST is connected to D+ instead of VBUS as recommended in the TI data sheet?

Thanks.

I don't know. I have to check this...

Brgds
 
Last edited:
In addition, can anyone hazard a guess as to why HOST is connected to D+ instead of VBUS as recommended in the TI data sheet? [\QUOTE]

The SN74AHC1G08 seems to be the one creating the HOST signal when having 3.3V and VBUS available. I can only think of that either it is communicating upwards giving an indication that something has been connected to the USB port or wakening the PCM2707. I also selects the 74HC157.

Now how does the circuit decide which input to use if both are active? LIFO or FIFO?

Brgds
 
Last edited:
I missed the G08 gate.

I would have preferred a schematic with interconnects in place. It would make it easier to read.

The HOST signal indicates to the 2707 that a USB device is connected during Self Powered Operation.

I've got a headache from reading the spec sheets and staring at the schematic. It is interesting that the SRC4192 data sheet states it only supports four data modes (TDM, and I2S, Left Justified and Right Justified), but it shows support of S/PDIF in figure 12.

I'm still hoping I have something goofy wrong that is easy to fix when I find it, rather than damaged parts.
 
Last edited:
I missed the G08 gate.

I would have preferred a schematic with interconnects in place. It would make it easier to read.

The HOST signal indicates to the 2707 that a USB device is connected during Self Powered Operation.

Yes, but what happens if you want to use your coax input while USB is connected... I understand from the PCM2707 datasheet that it goes into suspend mode when no data comes on D+. Should I understand that USB has preference over coax or that you cannot use coax if USB is connected?

If the later is true then the circuit is wrong. It would have been better to use the SSPND on pin 11 of 2707 to provide the logic for when it's safe to use coax. But I haven't scrutinized the whole circuit yet.

Brgds
 
Well, I got a bit of luck (possibly) as I found the WM8741 is not loading down the 3.3V supply. It measures 1.8V, but the regulator is bad and not putting out voltage. I believe the 1.8V is from I/O voltage back-feed through the protection diodes in the WM8741 and showing up on the power pin.

I wired up a LM317 to put out 3.3V and patched it in. I put a current meter in line to monitor it and it only pulls 38mA, which is close to nominal of 40mA. So possibly the WM8741 is not damaged.
 
On board regulator replaced. WM8741 now has 3.320V at the power pin.

When I turn on the headphone amp I get hum in the left ear. As I rotate the control CW, the hum goes down for the first 1/4 turn, then buzzing starts coming up in the right ear up to a little past half way, then both sides continue to increase with both hum and a buzzing (digital noise). This happens with or without any input.

No audio with USB input. No data from 2707.

No audio with S/PDIF although I can trace the signal to the SRC4192, and what looks like I2S from it to the 8741.

I've got 2 new 2707s on the way.
 
Turbon, do you still have chips running hot?

Mine are all slightly warm at best. However, after reading the regulator spec sheets I decided to run a pre-regulator.

I set up a LM317 for an output voltage of 6.15V which gives enough margin for the 5V regulator and minimizes the dissipation in the 3.3V regulators.

What is your input voltage?

The regulators will withstand the voltage up to 16V, but it is the dissipation that is a concern. Theta JC for the SOT-23 is 220°C/W.

At 6.15V in the 3.3 V regulator will dissipate 0.114W for a temp rise of 25°C. If you increase the supply to 8V the temp rise goes up to 41°C.
 
Yes I have.

7.5V - the WM8741 being the hotest. I have tried to understand what its nominal temp should be from the datasheet but I don't understand yet how to calculate it.
Nominal current seems to be ~50mA for 5V and ~40mA for 3,3V which would give
a total of 0.382W or ~ 0.4W. And then what would I look after to get the temperature?

Brgds
 
Just remember you are calculating the die temperature which is the critical value for reliability. The package temperature will be significantly lower, although still quite warm.

I should probably check my analog current to the 8701 as mine is barely warm. I've measured the digital current at 38-39mA.

I'll measure the analog current today and take a temp reading with a FLIR IR camera if I can use it at lunchtime.

The purpose of Theta j-a and j-c is to be able to calculate the die temperature from the ambient temperature or case. Unfortunately it is not as accurate for estimating the package temperature from dissipation. This would require a different thermal coefficient.
 
Last edited:
[...]I can keep my finger on it for a while but the sting forces me to remove the finger - max 20s. So no cooling needed?
No, thats OK. If you can keep the finger for 20s, it may have 50-55°C at the case.

New math:
Junction to case is 35°C/W
Junction - Ambient is 98°C/W

So, Case - Ambient is 98-35=63°C/W
0.4W*63°C/W=25.2°C, if it's 25°C in the "ambient" of the chip, the case has around 50°C.

You do not have to worry, the chip being around 65-70°C is ok.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.