cs8414 & wm8740 dac (one dac chip)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi guys,
i am working slowly on a dac mostly based on "Opus" DAC that should use CS8414 input receiver and WM8740 dac chip in stereo mode...
Output of the dac should use 6n6p tube connected as cathode follower...
The idea is to use the parts i allready have at my home from the previous projects (to make it cheaper) and also to make it nice and smiple but also to sound really good... It should run in 16 bit or 24bit mode (which we can selec on the dac chip)...

First sch of the dac is attached in this post and it is not finished - the tube section is not there - i still am considering how to do it since i would not like to use opamps to convert singal to unbalanced (and yes - i would like to have it simple and to use both outputs - ballanced and unbalanced)...so - any ideas for the solution are welcome... and also any comments related the first schematic of the dac....
thnx...
many regards
daniel
 

Attachments

  • dac_sch_rev01.pdf
    28.5 KB · Views: 605
Disabled Account
Joined 2002
That is probably due to the obsolete CS8414 you decided to use. CS8416 (or the DIR) is a better choice. Also a tube or opamp stage is not really needed as the DAC chip can be used passively without further active components. Adding a tube stage will only add distortion and there is no reason for doing so unless you like distortion.

So essentially Opus is:

- more simple ( when you take the tube stage in account )
- it has a more modern receiver with lower jitter sensitivity
- equipped with onboard regs close to the chips !!
- accompanied by a suitable power supply
- equipped with a passive output stage
- thoroughly tested and a proven design
- enhanced by using WM8741
- a very good sounding DAC
- reasonably priced

To surpass that level will be difficult !
 
jean-paul - good to see you here in this thread....
thnx for the answer....
...
i do not want to surpass the quality of OPUS dac...
i only want to build the dac for myself that will use the parts that i have in my junk box and that will (in that matter) be also cheap (because of the parts i allready have).....

so i am stuck with CS8414 and WM8740... that is why i decided to use those two:)
and since i am not that good in digital audio i simply would like that somebody checks the sch so that i can be sure that this will work and at least work good.... that is my only goal here....
the tube stage is less important here for now...
pcb is also less important - no need to check it unless somebody can see a major problem on it - maybe related the grounds or something........
o.k... thnx again for answering .....
 
Hi I had a quick look at your circuit and a few problems are clear.

1. CS8414 pin 16 is an output and needs to be tied to either 0v or VCC otherwise it will float and could draw lots of power. I suggest 0v (it sets what is output on other pins but since your not using them it shouldnt matter)

2. You have shown no power supply decoupling on your Sch. This part will not work correctly without it. You need at least a 100nF cap on both the digital and analogue power supply pins and I would recomend a 10uF can in parrallel with the 100nF as shown in application diagram. Do not think because it is digital it does not need a good power supply. Phase locked loops are very sensitive to power supply, get it wrong and it may not lock or you will get lots of jitter.

3. You have no reset circuit on the CS8414. It will work most of the time however it may operate strangely on occasions if it comes up in an odd state. The reset circuit is shown in the data sheet and basically takes all the mode pins high at the same time. (M0..M3) it achives this usign an OR gate so that the mode can be set at other times. It will probably work without this and I have seen other circuits without it, but you can't be sure.

4. Ensure that the filter components are close to the CS8414 and that the ground returns straight back to the AGND pin. If you get this wrong you will get lots of jitter.

5. For good performance you may need a double sided pcb. If you cant do this then some copper tape on the bottom side connected to ground will help.

6. SPDIF input needs a return ground, it can't float as if you have a transformer at the other end with no ground you will have no ground referance. So connect the Rxn side of the 75R to ground.

7. You have the same decoupling issues on the DAC. Each power supply pin needs a 100nF cap in parrallel with a 10uF cap. These parts are critical to the sound quality of the DAC and it is worth optomising them once it is built as the DAC seems to be sensitive to the characteristics of the caps used here. Ensure the the ground return paths are as short as possible.

8. On the VMID rails it appears that the two caps do not return to teh same ground. If so this is a mistake. VMID also needs 10uF caps, sounds rubbish without and THD is compromised quite alot of current is drawn from these caps in pulses.

9. Pin 27, 26 should be high for no deemphasis (most CDs dont have deemphasis) any that do will sound like the trebble has been turned up. You can wire the SPDIF reciever up to do this automatically but its quite complicated.

10, pin 25 should be wired high to unmute the output. or connected to a switch if you want a mute.

11. Pin 23 should be wired high for I2S 24bit, you may as well use this as any unused bits are filled with 0 anyway.

12 cCearly you will need an output filter. I would filter using an opamp stage and get it working before trying to do anything with valves.

13 You have not shown any PSU for this circuit. Thw WM8740 is very dependant on the quality of the power supply so I would put the regulators on the sam PCB and close to the WM8740 to keep the impedance of the supply as low as possible.

That should be enough to get you going. Good luck.

Andy.
 
gfiandy.... thnx for the kind reply...

1. i was using CS8414 without conecting pin 16 to low or high before in my TDA1543 dac... it was working fine... no problems occured... no unusual currents detected... locked every time - and is still workign with very good sound coming out of it...

2. power decoupling - you are completely right - i am using decoupling a lot... every psu has a decoupling but it is not shown on the sch attched - it is a revision one of the sch... just to ask questions and get some answers... :D my bad - forgot to draw it... :D

3. reset circuitry... well, i don't have it in my TDA1543 dac and it worked fine every time - in few dac's i build.... no problems were detected.... i will think about it... thnx...

4. filter is close to the cs8414 on the pcb....

5. double sided pcb is a problem... can't do it... i have to use single sided....

6. Rxn to ground - ouuupsss... forgot to draw it... you are right - it is going to ground...

7. same as no. 2.

8. VMID's are going to the same ground....

9., 10., 11. pins will be connected to insure 24bit I2S without deemphasis...thnx for the tip - i was reading the datasheet but do not have the table i draw for the pins' right now... i made it adjustable on the pcb just in case i misunderstood anything... :D thnx...:D

12. i'll try without the fitering first... if the sound is bad i will decide upon filtering.... thnx...

13. psu has two LM317 regulators (one for digital and one for analogue)... after that i use shunt regulators with TL431 (that worked nice in TDA1543 dac) to regulate the voltage to 5Vdc and 3.3Vdc that i for powering the receiver and dac....
i tried to make separate grounds - one digital and one analogue and i will connect them on input/output connectors (or, as an option, on one point on the pcb using a jumper....)...

i have a lot of jumpers on the pcb - this is not very good but i am trying to cut down the costs and to have a nice dac.. that is nearly impossible i am aware of it but still......
....
i will attach the current verion of the pcb and schematic - as soon as i update the sch.... :D
comments are welcome :D


gfiandy ... thnx again for the time and effort... much appreciated..... thnx again...
 
You definately need to post filter the output from the WM8740.
It is a very noisy DAC, and on top of that, it performs better with post filtering. Wolfson Micro recommends a lowish 20 KHz LPF for the device, to meet their standards.
The only DAC chip I know of, that could be used with no or at least a very low slope filter is BB´s chips, as they only need smoothing of the analog signal. Most other chips are musch more noisy, as it is the working frequency of the DAC or antialiasing filter, that has to be filtered. This noise can be as loud as only 10-15 dB below full scale at all times and typically @ several hundred KHz. This can result in amplifier clipping even at very low listening levels. As listening level is increased you might even burn your speakers without hearing any sound from them.
When filtered you also need a buffer, or else the characteristics of the filter might change with load.

If you do either as Gfiandy or the application notes sugests, the matter will be solved.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.