Need Help on DAC, anybody

[cdeveza]

I have been trying to copy the DAC End Andrea Ciuffoli,

http://www.audiodesignguide.com/DacEnd/index.html

For some reason I could not get it to work. Forgive me, this is
my first time in making a DAC and does not have much experience in digital. Please if anybody can help me try to figure out what I am doing wrong, I would greatly appreciate it. This are what I have, I have a CS8414, 74hC04 and AD1865. The input transformer I am using is from Scientific conversion.

I guess my problem is that I am not sure I understand how digital works and thus do not know how to troubleshoot it correctly. I thought I would try to isolate every stage first, so this is what I did and please correct me if I am wrong. I started on the CS8414 first and connected just the DIG power and DANA power, set the mode to normal, meaning M2= 1, M1=1, M0=0 and M3=0. then connect the LED just like it is in schematic, meaning from pin 28 and ground with resistor to limit current flow. Now, as soon as I turn the power on, will the LED light up right away, meaning HIGH? because it does, and when I try to connect the input to my digital out of my CD player, nothing happens, nothing comes out of pin 26. I checked my wiring so many times already and they are per schematic.

I do have a scope, but it is only 20MHZ, I try to use it, but do not even know what I should be seeing and which pin of the CS8414 to check.

Am I doing this correctly? Any help would be greatly appreciated,

Thanks,
Alex Deveza

 

[abraxalito]

I do have a scope, but it is only 20MHZ, I try to use it, but do not even know what I should be seeing and which pin of the CS8414 to check.

Am I doing this correctly? Any help would be greatly appreciated,

You'll need to connect your scope ground to the circuit ground, and probe the pins on the CS8414 with the scope probe. Remember to select DC coupling on the scope.

Since your LED is on, you must have some digital power, but check its within spec on +5V (pin 7). Next check that the analog power is correct - pin22 - it also needs to be close to +5V.

Assuming those two are correct, then double check with an ohm meter that you got the DGND and AGND connected (pins 8,21) to the circuit ground.

Next signal to check the presence of is the input SPDIF which must appear at pin9 (receive +ve input). You'll be wanting to see around 500mV p-p of bi-phase signal here for the receiver to lock on. Check that a signal appears here when your CD player digital out is connected and isn't present when its not connected. A 20MHz scope should be able to make a reasonable stab at this signal, the bit rate is about 6Mbit.

 

[cdeveza]

Thanks. I do have +5v at pin 7 and 22, and the grounds, pin 8 and 21 are connected.

I will check tonight for the presence of SPDIF at pin 9. But, should the LED be off when there is NO SPDIF? If so, then why is it on?

Alex

 

[abraxalito]

From what I can make out from the CS8414 datasheet, pin28 is a logical OR of validity data from the SPDIF stream with the error flag. Since your LED is on, we can safely assume this is as the result of the error flag being set - probably because of no lock established with the incoming stream (if it is indeed present). So no, the LED should be _on_ with no connection to your CD player.

 

[cdeveza]

Ok, I checked presence of SPDIF. I connected the cable from my CD player digital out to my DAC input transformer, connect my scope to pin 9 of CS8414. I turn on my DAC, when I turn on the CD player, I do get about 500mv peak to peak at pin 9. So does that mean my CS8414 is OK? The LED still stays on, is this correct?

I put a music cd in the player, but still no sound. I do get something on pin 11 and 12, but nothing on pin 26. What should I check next?

Thanks,
Alex

 

[abraxalito]

Can't tell if your CS8414 is ok just from the signal at its input. It would be wise to check there's no signal at pin10 - that's the -ve side of the differential input.

Since the validity bit is in actuality a not-valid bit, then if your CS8414 is correctly decoding a valid audio SPDIF stream your LED will be off. The fact its on either indicates your CD player is either putting out non-PCM audio data (Dolby Digital perhaps) or that your receiver has not locked. I'm inclined to believe its the second of these.

Zero output on pin26 means the data bits are all zero. What are you getting on pin11 (frame sync) and pin12 (serial clock) ? I'd also check pin19 (master clock) to see if there's a stable 11MHz clock there. If the receiver has locked then pin11 will have a squarewave at 44kHz on it.

 

[cdeveza]

Thanks again. Well, my circuit is showing that pin 10 is grounded. I know that my source is SPDIF as I have one of those cheap chinese dac that accepts SPDIF and it works. So that means the receiver has not locked.

Sorry, I am not quite sure I understand your question on pin 11 and 12, can you elaborate please?

I will check on pin 19 and see what shape is on pin 11.

Assuming its not locking, (I think it is NOT locking), what could be causing this? How does one fix this?

Thanks,
Alex

 

[cdeveza]

Ok, I get strong square wave signal at pin 19, but no square wave on pin 11, with or without and SPDIF input. So, I guess its not locking in.

Here is where I am stuck, what do I do now? What am I doing wrong?

I am not quite sure about the input transformer, is there any orientation? I do not know which is input and which is output.

Thanks again,
Alex

 

[abraxalito]

Ok, I get strong square wave signal at pin 19, but no square wave on pin 11, with or without and SPDIF input. So, I guess its not locking in.

I'm a bit suspicious that you're getting a 'strong squarewave on pin19' as it shouldn't look at all square on a 20MHz scope: if its locked then an 11MHz signal will look to you as a slightly flattened sinewave. So I conclude its not locked and the signal you're seeing there is considerably lower than 11MHz.

In answer to your question what you're doing wrong I'd say you're trusting in the schematic (and therefore the designer of that schematic) blindly rather than learning for yourself. The datasheet for the CS8414 is freely available but it seems you've not checked it against the schematic you've built for yourself.

I suggest you have a look at appendix B of the CS8414 datasheet and compare the suggested circuit shown there with the schematic you've built. Notice any differences ?

As for the input transformer I doubt its an issue as you're getting a 500mV signal at the input. 1:1 transformers are reversible and the SPDIF signal is not polarity sensitive.

 

[cdeveza]

Thanks again, I do have the datasheet of CS8414, however, the appendix B of the datasheet that I have only shows reseting after power-up. Meaning pulling all mode select 0 to 3 high, and I also tried this set-up to reset it, but nothing happen.

You are right as to the shape of the square wave on pin 19, as it was really not totally square, but I thought it was close enough. Pin 11 has a very small signal, kind of sawtooth shape. But no matter what the shape of the signal is, it does not respond to any SPDIF input that I put in, this tells me its not locking-in.

I read the entire datasheet, but have to admit that I do not fully understand most of it, and therefore do not know how to troubleshoot it correctly. Right now I could not get the CS8414 to lock-in into the SPDIF, and do not know what I am doing wrong.

Again, this is what I have so far. I have (2) +5v power supply, one I call DIG and one DANA, DIG is connected to pin 7 AND 18, DANA is connected to pin 22, ground pin 8 and 21 are connected together. Pin 16(SEL), Pin 17(M3), pin 23(M0), are low, meaning grounded, pin 24(M1) is high, connected to DANA, this sets the mode to 18 bits LSBJ as per pg 21 of the datasheet. I have an LED connected to pin 28 with 500 ohm resistor to limit currect. I have the SPDIF transformer connected to pin 9 and 10, I have 47nf and 1k in series connected to pin 20(filter) and to ground. All ground are connected together. No connection to pin 26, 11, 12. This are as far as I got. With these connections I thought it should lock at the presence of SPDIF, but it does not. Are my assumptions correct? If I am missing something, I would appreciate some guidance.

Thanks again,
Alex

 

[abraxalito]

Thanks again, I do have the datasheet of CS8414, however, the appendix B of the datasheet that I have only shows reseting after power-up. Meaning pulling all mode select 0 to 3 high, and I also tried this set-up to reset it, but nothing happen.

OK so did you implement this reset stuff as in the appendix or in some other way? Please describe the circuitry you included because its not shown on the schematic you posted up.

You are right as to the shape of the square wave on pin 19, as it was really not totally square, but I thought it was close enough.

Please make a note of the frequency on pin19, both with and without the CD player connected.

Pin 11 has a very small signal, kind of sawtooth shape.

How small? Please make a note of the amplitude peak-peak.

Again, this is what I have so far. I have (2) +5v power supply, one I call DIG and one DANA, DIG is connected to pin 7 AND 18, DANA is connected to pin 22, ground pin 8 and 21 are connected together. Pin 16(SEL), Pin 17(M3), pin 23(M0), are low, meaning grounded, pin 24(M1) is high, connected to DANA, this sets the mode to 18 bits LSBJ as per pg 21 of the datasheet. I have an LED connected to pin 28 with 500 ohm resistor to limit currect. I have the SPDIF transformer connected to pin 9 and 10, I have 47nf and 1k in series connected to pin 20(filter) and to ground. All ground are connected together. No connection to pin 26, 11, 12. This are as far as I got.

I've found a disagreement, but I have no clue which is correct, the description or the schematic. Just from memory, your diagram showed the SPDIF transformer connected to the CS8414 via some caps - in this you don't mention those caps. So are they there or not?

With these connections I thought it should lock at the presence of SPDIF, but it does not. Are my assumptions correct? If I am missing something, I would appreciate some guidance.

According to the datasheet the part needs a reset every time its powered up. You haven't got any way on your schematic to apply this reset so that's my best guess right now on what you're doing wrong. My second best guess is that somehow the chip has become damaged.

 

[cdeveza]

Thanks for your reply. The way I tested the reset is this. Since pin 23(M0) and pin 17(M3) are normally low, I just manually connect the two together and just manually touched DANA +5v right after power-up, then see if any changes at all to the status of the LED or pin 11, 12 and 26, but there are no changes. Also, I have searched everywhere and everyone that uses the CS8414 does not implement this scheme, including Audio Note DAC, although I tried it anyway.

I will try to measure the frequency on pin 19 and pin 11. However, as I mentioned, the shape, amplitude and frequency at pin 11, 12 and 19 does not change WITH or WITHOUT SPDIF input. Meaning the chip does not respond at all at any presence of SPDIF input. As for the input cap, yes, I forgot to mention, I do have that cap, but again other schematic that I saw does not have these caps. Even fig. 21 of Appendix A of the datasheet does not show this Cap.

I thought I have a damaged chip, so I got another from ebay, and it too is the same. Now I could have damaged them both, because these surface mount chips are a pain to install.

Thanks,
Alex

 

[abraxalito]

Fig 21 is the suggested circuit for a professional AES/EBU input - you're using a consumer grade CD player to drive it. The amplitudes of the signals are different and the AES/EBU signal is balanced. However you could try that circuit since you already have a transformer (which in your consumer application is really unnecessary) and see if it makes any difference. The 110R termination resistor shown would need to be changed to 75R.

Since you've mentioned that the frequency of pin 19 doesn't change then its not necessary to measure the frequency. However the amplitude (not the frequency) and level at pin11 is of interest since it might be that the CS8414 is not seeing a properly conditioned signal. You only mentioned one cap in your most recent text, but you schematic shows two caps. Do you indeed have both? Also the average DC voltage level at pin10 & pin11 is of interest - it should be about 1.7V.

Does the cheap Chinese DAC you mentioned use a CS8414? if so you might try comparing the signals you're seeing at the input of yours with the signals on that one, particularly pins 10&11.

Another suggestion - to change the values of the components connected to pin20 back to the datasheet suggested values (470R, 68nF). These values will have some effect on whether the CS8414 is able to achieve lock.

My last point is about your method for resetting - I suggest installing a pull-down resistor of 4k7 from pins 17&23 to ground. I agree its strange that other manufacturers get the chip working without such circuitry.