Current differential to voltage line level from PCM1794 et al - diyAudio
Go Back   Home > Forums > Source & Line > Digital Line Level

Digital Line Level DACs, Digital Crossovers, Equalizers, etc.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 17th July 2013, 05:34 PM   #1
JanErik is offline JanErik  Finland
diyAudio Member
 
Join Date: Aug 2010
Location: Vasa
Default Current differential to voltage line level from PCM1794 et al

I have seen different approaches to this;
See this PCM1794A audio DAC and PupDAC at http://www.diyforums.org/PupDAC/sche...chematic-3.gif , both are using only resistors instead of the inital opamp stages in the reference design http://www.ti.com/lit/ds/symlink/pcm1794.pdf , page 20.

What do you think, which is better?
Seems the simpler designs still have some sort of filtering with the capacitor directly on the current outputs.
  Reply With Quote
Old 17th July 2013, 05:55 PM   #2
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
well the manufacturer's app engineers seem to like op amp I/V

and all DAC need some analog low pass - more extreme the lower the sampling or upsampled frequency
  Reply With Quote
Old 17th July 2013, 08:35 PM   #3
diyAudio Member
 
Avro Arrow's Avatar
 
Join Date: Sep 2010
Location: Toronto, eh
Then there's fun stuff like this:
Attached Images
File Type: gif MC Phono Pre Current Input.gif (10.3 KB, 643 views)
  Reply With Quote
Old 17th July 2013, 11:45 PM   #4
00940 is offline 00940  Belgium
diyAudio Member
 
Join Date: Sep 2003
Location: Paris
I took the hint from the Bel Canto DAC2 and use opa1632 as I/V converters after a pcm1798. Sounds good to me.

Two things to watch for :
- the opa1632 doesn't like capacitive loading so it's best to avoid caps in // with the I/V resistors. Put the filter on the next stage.
- the vcom pin of the opa1632 set the output voltage and not the input voltage. So you can't tie it to ground. If you did so, you would have a voltage at the DAC's output equal to your I/V resistor's value multiplied by the pcm1794's current offset (which isn't good). So you need to feed the vcom with a suitable voltage reference to bring the voltage at the pcm1794's output down to 0.*

Btw, if you go down the "simple resistors" road, you shouldn't use more than 22r.

* Which is why I wouldn't use an IVYIII from Twisted Pear on PCM179* DAC.
__________________
Ben.

Last edited by 00940; 17th July 2013 at 11:52 PM.
  Reply With Quote
Old 18th July 2013, 12:50 AM   #5
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
driving Cload doesn't have much to do with feedback caps if you're not using noise gain input C to gnd too, the negative input "virtual gnd" really isn't real gnd

in fact the OPA6132 datasheet has an example with 1 nF feedback C

not knowing the DAC Iout Z can't really say if it would work but the reasoning should be kept straight
  Reply With Quote
Old 18th July 2013, 08:42 AM   #6
JanErik is offline JanErik  Finland
diyAudio Member
 
Join Date: Aug 2010
Location: Vasa
Quote:
Originally Posted by 00940 View Post
Btw, if you go down the "simple resistors" road, you shouldn't use more than 22r.
According to the datasheet, PCM1794 can feed 7.8 mA p-p, would it not bottom out with that 22R if max voltage is 5 V? It seems PupDAC has 22R though, but the other has 220R.
  Reply With Quote
Old 18th July 2013, 01:59 PM   #7
Calvin is offline Calvin  Germany
diyAudio Member
 
Calvin's Avatar
 
Join Date: Nov 2004
Location: close to Basel
Hi,

I´d say read the Datasheets
The 7.8mApp are centered around -6.2mA. So the current swings between -6.2mA+-3.9mA or -2.3mA and -10.1mA.
This current drives the attached impedance to IxR Volts, hence to +0.506V to 2.222V, centered around 1.364V with R=220Ohms.
TI´s slyt360, "Interfacing OP-Amps to high-speed DACs, Part2: Current-sourcing DACs", from Analog Applications Journal 4Q,2009, states a voltage compliance range of +-1.25V for CMOS-DACs (using PMOS-current sources).
The voltage compliance range defines the limits of voltage that may safely be applied to the analog current outputs of the DAC.
Exceeding the upper limit may result in increased THD and linearity issues and eventually shut down of the PMOS-CCS.
Exceeding the lower limit might lead to breakdown of the device.
TI specs the optimum voltage compliance range to max. 0.5V peak (see DS of DAC5674).
The pup stays well within thoase limits, while the Pavouk doesn´t.

Rem: Certain DACs also feature protection Diodes at their outputs, for example the PCM63. Here the voltage applied to the current outputs must stay well below the Diodes turn-on voltage of ~600mV (factor of 1/3 to 1/5).

As to the Q of which is better, it´s certainly the pupDAC.
Besides the too high I/V resistors the Pavouk also uses a insufficient Opamp.
The Opamp mainly will see the I/V-resistors as source impedance. With values of 220R and 22R the noise voltage figures becomes dominant.
Beeing a JFET-input OPamp the TL072 shows high values of voltage noise combined with a high 1/f frequency. Hence its noise performance will be mediocre at best.
Beeing a low-bandwidth, large settling-time, or in short, ´slow´OPamp the fast current/voltage steps at the DACs output would override the TL072, forcing it into overdriven openloop conditions, hence loads of THD. So the current steps need to be slowed down, which is done with the rather largish caps C27 and C28 between the DAC´s current outputs (ceramics, not specced but hopefully NPOs or COGs). Only them will save the day. I guess that from the potentially 24bit of resolution of the PCM1794 maybe only 8-9bits linearity will be left.
In sum: the TL072 is by far not adaequate to the PCM1794´s qualities.
I tend to say the Pavouk is just a sad pinnacle of bad design.

The pupDAC is just looking similar, but is much more carefully/sensible designed.
The I/V resistors are small enough in value to keep the DAC within its specced performance range and the recommended OPamps are far better suited to their task, beeing much faster´, more linear and less noisy. At first glance the OPA2836 seems the best to me.
Consequently the caps C3/C18 are just 1/10 in value and Mica dielectric.
As a major drawback nowadays I regard the PCM2707 which can only work with USB1.0, hence only up to 16Bit, 48kHz, which is at the same the feed limit for the PCM1794. Not up to date any more.

jauu
Calvin
  Reply With Quote
Old 18th July 2013, 02:15 PM   #8
diyAudio Member
 
dirkwright's Avatar
 
Join Date: May 2011
Location: Virginia
Quote:
Originally Posted by Calvin View Post
Hi,


The pupDAC is just looking similar, but is much more carefully/sensible designed.
The I/V resistors are small enough in value to keep the DAC within its specced performance range and the recommended OPamps are far better suited to their task, beeing much faster´, more linear and less noisy. At first glance the OPA2836 seems the best to me.
Consequently the caps C3/C18 are just 1/10 in value and Mica dielectric.
As a major drawback nowadays I regard the PCM2707 which can only work with USB1.0, hence only up to 16Bit, 48kHz, which is at the same the feed limit for the PCM1794. Not up to date any more.

jauu
Calvin
It looks to me that the pupDAC designer merely substituted passive I/V resistors for the active opamp IV of the standard I/V converter from the TI datasheet, with a few changes to filtering.
  Reply With Quote
Old 18th July 2013, 04:58 PM   #9
JanErik is offline JanErik  Finland
diyAudio Member
 
Join Date: Aug 2010
Location: Vasa
OK, I will be using OPA4134 in my design atleast.
  Reply With Quote
Old 18th July 2013, 05:09 PM   #10
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
duals are a lot more popular than quads, more different op amps are available in dual
usually layout is limited by how close to the op amp you can put all the passives - feedback, bypass - so duals often win as it gets too crowded near a quad

and some of best, fastest, op amp I/V candidates now come in singles, very small sot transistor packages, keeping lead inductance lower for their 50+ MHz GBW

Last edited by jcx; 18th July 2013 at 05:18 PM.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Line level SE to differential "fix-it" box Iain McNeill The Lounge 0 15th January 2010 06:09 AM
Descrete Differential Current to Voltage Converter TimS Digital Line Level 11 8th October 2009 04:18 PM
Line level subwoofer out voltage? thefish Subwoofers 2 3rd February 2007 08:29 PM
line voltage level metebalci Tubes / Valves 2 1st March 2004 01:45 PM
Line Level, Balanced Voltage Level... dkemppai Solid State 3 11th December 2002 12:04 AM


New To Site? Need Help?

All times are GMT. The time now is 08:56 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2