Inertial
These look to be an excellent way to avoid solid state and negative feedback:
http://www.berliner.netfirms.com/machines.htm
If you interested in the mechanics of the Colinear dacs I've put a copy of the patent covering the Colinear method in pdf format up here.
These look to be an excellent way to avoid solid state and negative feedback:
http://www.berliner.netfirms.com/machines.htm
If you interested in the mechanics of the Colinear dacs I've put a copy of the patent covering the Colinear method in pdf format up here.
BPO
I'd note also that the DAC output isn't "processed" or "run through" by the BPO. BPO is an offset current derived from the same buried zener reference used by the DAC sections. Without the BPO the PCM1704 would output 0 to -2.4mA. The BPO biases the output with 1.2mA current to make the output symetrical around zero: -/+ 1.2mA.
You can see this clearly on the PCM63 where both the raw DAC output and the Bi Polar offset current are available as separate pins.
I'd note also that the DAC output isn't "processed" or "run through" by the BPO. BPO is an offset current derived from the same buried zener reference used by the DAC sections. Without the BPO the PCM1704 would output 0 to -2.4mA. The BPO biases the output with 1.2mA current to make the output symetrical around zero: -/+ 1.2mA.
You can see this clearly on the PCM63 where both the raw DAC output and the Bi Polar offset current are available as separate pins.
Thanks Spzzzktt !
Very apppreciate documentation!
Now I turn the pdf to my partner who is more inclined to electronic...
I have to take my time to understand the lesson!
Anyway guys, to make a long story short: do you realize that despite all this technology , 99% of the cdplayer sound like dogs ?
What is your diagnosis? ( excuse the OT)
I think the analog part has a great part in this ...
Cheers,
Paolo ( LP still rules!)
inertial said:
Anyway guys, to make a long story short: do you realize that despite all this technology , 99% of the cdplayer sound like dogs ?
What is your diagnosis? ( excuse the OT)
I think the analog part has a great part in this ...
The most common problems:
-poor power supply
-poor output stage (yes, the analog part as you said)
-sigma-delta converters
-dirty signal path and RFI
Hi Bernhard,
A normal implementation for an I/V resistor would be connected from the DAC output pin to analog common. You would then feed an op amp from the DAC output pin as you would any other audio signal.
The main difference between the two methods is that an I/V resistor returns to a hard analog common point. The resistor in the PCM63 returns to a virtual ground, hopefully referenced to the analog common, from the op amp output.
I can see your point that the audio voltage is developed across this resistor in each case. The circuit location is completely differnt though.
-Chris
Well, that's not really a classic "I/V" resistor anyway. It's more like a current limiting resistor that sets the voltage gain. You send a current into the op amp and get a current out to null against the output current. In order to see a voltage from this, the resistor causes the output of the op amp to rise in voltage enough to cause current through that resistor to cancel the current coming in. In other words, your standard, every day feedback resistor. It's on the chip so that the value can be closely controlled.
A normal implementation for an I/V resistor would be connected from the DAC output pin to analog common. You would then feed an op amp from the DAC output pin as you would any other audio signal.
The main difference between the two methods is that an I/V resistor returns to a hard analog common point. The resistor in the PCM63 returns to a virtual ground, hopefully referenced to the analog common, from the op amp output.
I can see your point that the audio voltage is developed across this resistor in each case. The circuit location is completely differnt though.
-Chris
Re: Re: BPO
Yes, in parallel.
nagaesan's attachment shows a block diagram of the pcm1704 architecture. The connection of buried zener reference -> BPO Amp is to the Iout. You can see why the ref and BPO decoupling caps make a difference to sound - any excess noise is dumped straight onto Iout.
Paul
inertial said:
Yes, in parallel.
nagaesan's attachment shows a block diagram of the pcm1704 architecture. The connection of buried zener reference -> BPO Amp is to the Iout. You can see why the ref and BPO decoupling caps make a difference to sound - any excess noise is dumped straight onto Iout.
Paul
PCM1704 size and type of capacitors
Nagaesan kindly offered up the architecture of the PCM1704 which was most illuminating.
I note that rather large electrolytics are specified for the relevant capacitors in the TI datasheet (large for those of us who prefer 100% polypropylene film at least):
REF DC 47uF electrolytic
SERVO DC 47uF electrolytic
BPO DC 100uF electrolytic
I am considering going 100% film and am wondering if the actual values have much importance.
My thinking is that the value of REF DC is is unlikely to be of importance to the operation of the part particularly if the supply to the chip is very well bypassed, but that the value of SERVO DC and BPO DC does have some significance on the operation of the part where one might actually benefit from going bigger. On the flip side, it seems reasonable that the noise quenching qualities of smaller film caps should be way better than the larger electrolytics specified.
Your comments would be most appreciated.
Petter
Nagaesan kindly offered up the architecture of the PCM1704 which was most illuminating.
I note that rather large electrolytics are specified for the relevant capacitors in the TI datasheet (large for those of us who prefer 100% polypropylene film at least):
REF DC 47uF electrolytic
SERVO DC 47uF electrolytic
BPO DC 100uF electrolytic
I am considering going 100% film and am wondering if the actual values have much importance.
My thinking is that the value of REF DC is is unlikely to be of importance to the operation of the part particularly if the supply to the chip is very well bypassed, but that the value of SERVO DC and BPO DC does have some significance on the operation of the part where one might actually benefit from going bigger. On the flip side, it seems reasonable that the noise quenching qualities of smaller film caps should be way better than the larger electrolytics specified.
Your comments would be most appreciated.
Petter
Re: Re: Re: BPO
Thanks Paul !
All become to be more clear to my newbie's eyes
And very thanks to kind Nagaesan !!
P.S.: abot various possible modds, what do you think about new PSU for 1702/4 module ? I am thinking at a no-compromise shunt regulated type, maybe chocke input also!
Plus wima Mks little capacitors also as Petter has suggested...
Cheers,
Paolo
spzzzzkt said:
Thanks Paul !
All become to be more clear to my newbie's eyes
And very thanks to kind Nagaesan !!
P.S.: abot various possible modds, what do you think about new PSU for 1702/4 module ? I am thinking at a no-compromise shunt regulated type, maybe chocke input also!
Plus wima Mks little capacitors also as Petter has suggested...
Cheers,
Paolo
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.