I'm using CS8412 (with TDA1453 DAC) in master mode and it is ok. When I use it in slave mode with external oscillator 11.2896 and 4040 divider it sometimes happens that after power on there is no sound. So I power it on and off several times and in the end it sounds. In other workds it is not stable. I'm quite sure it is due to slave mode - external oscillator as in master mode it always works. Any idea what can be wrong? (power supply, start-up sequence, decoupling, ...)
Attachments
When the DAC fails to work, make sure the oscillator is oscillating and oscillating at the correct frequency, not some harmonic.
If the oscillator is OK than the most likely problem is the 8412 initializing before the input clocks are stable. You can eliminate this problem by applying a delayed reset to the 8412.
But why bother using slave mode at all? It doesn’t appear that you are also slaving the digital source to the master clock. The result will be missed or duplicated samples.
And why use a 4040? It’s a ripple counter. Jitter introduced at each stage is added to the jitter introduced by each of the preceding stages. Use a synchronous counter, instead.
Finally, why use an 8412? Any of the more modern DIRs are better in every way.
If the oscillator is OK than the most likely problem is the 8412 initializing before the input clocks are stable. You can eliminate this problem by applying a delayed reset to the 8412.
But why bother using slave mode at all? It doesn’t appear that you are also slaving the digital source to the master clock. The result will be missed or duplicated samples.
And why use a 4040? It’s a ripple counter. Jitter introduced at each stage is added to the jitter introduced by each of the preceding stages. Use a synchronous counter, instead.
Finally, why use an 8412? Any of the more modern DIRs are better in every way.
Thanks for all your comments. As I mentioned in slave mode 8412 is ok and it never appeared to be a problem with playing music after power on. However I wanted to upgrade to master mode for CS8412 - there seem to be good reasons to do it (Description of the DDDAC1543 project - overclocking using external oscillator shows some benefits).
So when I set CS8412 to master mode it sometimes happens that after power on there is no sound. I use the same power supply for transport and DAC so there is no delay between them. The same with clock -CS8412 and DAC TAD1543 - no delay. Do you think this might be a problem? What should be the correct sequence of starting these parts and how to easily apply delay?
Thanks
So when I set CS8412 to master mode it sometimes happens that after power on there is no sound. I use the same power supply for transport and DAC so there is no delay between them. The same with clock -CS8412 and DAC TAD1543 - no delay. Do you think this might be a problem? What should be the correct sequence of starting these parts and how to easily apply delay?
Thanks
I’m confused by your terminology. In your first post you wrote: ”When I use it in slave mode with external oscillator 11.2896 and 4040 divider it sometimes happens that after power on there is no sound. Now you say: ”As I mentioned in slave mode 8412 is ok…”
Master mode is when the 8412 provides the clocks derived from the S/PDIF input. Slave mode is when an external oscillator provides the clocks. IMO, slave mode is of no value unless the external clocks provided are better than the S/PDIF derived clocks (not the case when using a 4040 divider) and the S/PDIF source is also slaved to the same clock.
Most oscillators have a startup delay of 1ms or more. During that time the output of the oscillator is undefined. That’s why most clocked components carry an advisory similar to this one: ”On initial power up, RST must be held low until the power supply is stable, and all input clocks are stable in frequency and phase.”
The 8412 enters the reset state when M0, M1, M2, and M3 high. Simply add a circuit that asserts those inputs for 100ms or longer after the power is turned on. On the other hand, you could have a defective oscillator. That’s why I suggested you first check it for proper operation.
Master mode is when the 8412 provides the clocks derived from the S/PDIF input. Slave mode is when an external oscillator provides the clocks. IMO, slave mode is of no value unless the external clocks provided are better than the S/PDIF derived clocks (not the case when using a 4040 divider) and the S/PDIF source is also slaved to the same clock.
Most oscillators have a startup delay of 1ms or more. During that time the output of the oscillator is undefined. That’s why most clocked components carry an advisory similar to this one: ”On initial power up, RST must be held low until the power supply is stable, and all input clocks are stable in frequency and phase.”
The 8412 enters the reset state when M0, M1, M2, and M3 high. Simply add a circuit that asserts those inputs for 100ms or longer after the power is turned on. On the other hand, you could have a defective oscillator. That’s why I suggested you first check it for proper operation.
Tam Lin: Thanks for your comments. Of course it is my mistake. Slave mode is when external oscillator is used (problem with start up), master mode is when CS8412 uses SPDIF to create clock using internal PLL (this works fine). Do you see any easy change to my circuit how to delay / hold RESET of CS8412 after power on? I do already have PCB and it is not easy to modify the design.
theAnonymous1: How have you used this voltage regulator to solve this problem plese?
Thx
theAnonymous1: How have you used this voltage regulator to solve this problem plese?
Thx
Thanks for your comments. Wouldn't it be possible to delay power supply of CS8412 by adding some reset controller? (e.g. TPSxxxxx mentioned above) I have seen the page 38 og CS8412 datasheet but it seems to be quite complicated to disconnect M0-M2 pins, add a couple of gates etc. Is there any easier solution?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.