OS filter Problem

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

jobstens

Member
Joined 2004
jobstens.de
Bernhard said:
There must be a big bug.

Pictures show -60dB and -80dB 1 kHz signal.
-80 dB already looks like a square :(

Wrong data format ?

I had exactly the same problem or even worse when fitting the PCM53 in a JVC player whitch originally has PCM54 DAC.

Edit:

If you look closer, you can see the square from -80dB also in the -60dB signal.
Click to expand...

There are two more things, if you're looking closer:
1: The negative wave has more noise
2: The spikes at the peak of the negative wave

And: If you are replacing the PCM54 by the PCM53 - was the PCM54's MSB adjusted with these resistors ? then you should remove them ;-)
The distortion has the same wave like the MSB should have - that's why I'm thinking about the resistors ...


Regards

Jobstens
 
Bernhard

Bernhard

Member
Joined 2002
jobstens said:
Missing Page 130 ...
Click to expand...

Thanks, this seems to be it, but still can not find the info how data is loaded into flag register... Must be hidden somewhere else in the datasheet.

Last sentence on page 129:

Figure when the attenuator setting is changed

First word on page 131:

flag)




Does this book also include SM5804 SM5806 SM5807 SM5818 ?

I feel I have to watch eBay for NPC databook.
 
Bernhard

Bernhard

Member
Joined 2002
jobstens said:


There are two more things, if you're looking closer:
1: The negative wave has more noise
Click to expand...

Yes.
jobstens said:


2: The spikes at the peak of the negative wave
Click to expand...
Yes.

And also the spikes on zero crossing are occuring only on the falling edge of the sine.

And what does all that tell me ?
Nothing :(

jobstens said:

And: If you are replacing the PCM54 by the PCM53 - was the PCM54's MSB adjusted with these resistors ? then you should remove them ;-)
The distortion has the same wave like the MSB should have - that's why I'm thinking about the resistors ...
Click to expand...

There was a problem: The circuit used Iout of the PCM54 and external opamp for I/V.
PCM53 does not have Iout so I only used the 16 data lines from this player and directly picked the sigal from Vout of the PCM53.
 
jobstens

jobstens

Member
Joined 2004
jobstens.de
Bernhard said:


Does this book also include SM5804 SM5806 SM5807 SM5818 ?

I feel I have to watch eBay for NPC databook.
Click to expand...

NPC Data Book I contains this Filters

4fs:
5807 D/E/F

8fs:
5841 A/B/C/D
5840 JP/JS

High-Efficiency Interpolation Filters:
5803 AP/APT
5813 AP/APT/AF
5842 AP/APT
5843 A

IIS Format Input Interpolation Filters:
5840 FS/FP
5880 P/S
5881 S

Decimation Filters:
5805
5815 A


By the way: I already looking for the SM5840_CP_ Datasheet ...


Regards

Jobstens
 
R

rfbrw

Member
Joined 2001
Bernhard said:


Thanks, this seems to be it, but still can not find the info how data is loaded into flag register... Must be hidden somewhere else in the datasheet.
Click to expand...


The information you need is there even without page 130.
There are 8 modes, MC0 to MC7. These are selected using address lines MS1 to MS3. Data for the selected mode is clocked into the MSIPO serial to parallel register using the /MEN pin with data on the MDT pin. The register you are after,Rmod,is accessed via MC6. See page 133.
 
Bernhard

Bernhard

Member
Joined 2002
jobstens said:


NPC Data Book I contains this Filters

4fs:
5807 D/E/F

8fs:
5841 A/B/C/D
5840 JP/JS

High-Efficiency Interpolation Filters:
5803 AP/APT
5813 AP/APT/AF
5842 AP/APT
5843 A

IIS Format Input Interpolation Filters:
5840 FS/FP
5880 P/S
5881 S

Decimation Filters:
5805
5815 A


By the way: I already looking for the SM5840_CP_ Datasheet ...


Regards

Jobstens
Click to expand...

5807 how many bit output possible ?
5840 has many total different versions I have only what is in the web, no 5840CP.
 
Bernhard

Bernhard

Member
Joined 2002
rfbrw said:



The information you need is there even without page 130.
There are 8 modes, MC0 to MC7. These are selected using address lines MS1 to MS3. Data for the selected mode is clocked into the MSIPO serial to parallel register using the /MEN pin with data on the MDT pin. The register you are after,Rmod,is accessed via MC6. See page 133.
Click to expand...

Thanks, I will study that, but I doubt that my spike problem will be solved if I use 5813 4xOS, because in JVC XL-V1100 it was exactly the same problem, also with 4xOS.
 
jobstens

jobstens

Member
Joined 2004
jobstens.de
Bernhard said:


5807 how many bit output possible ?
Click to expand...

16 bit in/out - but it's for low end systems (they say).

Bernhard said:

5840 has many total different versions I have only what is in the web, no 5840CP.
Click to expand...

me 2 ... :-(


Maybe your spike problem is the result of the switching speed from the previous stage (serial to parallel conversion, os-filter ... what ever) and should be removed by the deglitcher. (?)


Regards

Jobstens
 
Bernhard

Bernhard

Member
Joined 2002
jobstens said:


Maybe your spike problem is the result of the switching speed from the previous stage (serial to parallel conversion, os-filter ... what ever) and should be removed by the deglitcher. (?)
Click to expand...

Yes, the grandma is too slow.

In a DCD1500 ( 2 x SM5801 + 2 x PCM54 ) it works, no spikes but -60dB performance is mediocre like a TDA1543.

And on Vout again the same spikes, so this is removed by the deglitcher.

I was not able to measure something useful on Vout in the TEAC where performance is very good.
Only some kind of noise, nothing that reminds on a sinewave...
Need to follow the traces and draw a schematic, maybe it is witchcraft.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom