TDA1545A and Vref

[hollowman]

With the TDA1545A, how should one handle Vref?
From the TDA1545A datatasheet, it's supposed to be fed to the (+) input of the I/V opamp; I_out gets the (-) input .
Most DIY designs for TDA1545A seem to ignore Vref, leaving it dangling (see Examples below). This goes against Philips' recommendation. So, what's the real deal with Vref?
(P.S. If you don't have experience with TDA1545A, how about TDA1543?) Thx!
Examples:
Rbrorer TDA1545A I/V
...and...
Peufeu's TDA1545A I/V:

[wwenze]

My guess is that if Vref isn't 2/3 VDD then there will be DC offset at the op-amp output?

And if the non-inverted input is connected to ground, the op-amp will be outputting positive voltage all the time? That would be bad.

[hollowman]

Originally Posted by wwenze

My guess is that if Vref isn't 2/3 VDD then there will be DC offset at the op-amp output?

And if the non-inverted input is connected to ground, the op-amp will be outputting positive voltage all the time? That would be bad.

I think the 1545A datasheet ckt, per the shown voltage divider, pretty much takes care of 2/3 Vdd.
Your second point is why I raised the query to begin with. But that (not using Vref) is what some DIYers seem to be doing for discrete I/V. Here, again, are examples of this:
Single rail, active I/V for TDA1543, TDA1545A
http://peufeu.free.fr/audio/extremis...plement_4.html

The first one seems to have been primarily designed for the TDA1543, tho' Rudolf does note (almost as an afterthought) that it works for TDA1545A. It does and it sounds fine (I have built it!). But I keep wondering about the dangling Vref. Is this a fundamental part of Philips' design -- or just an afterthough tweak? Is there any way one can use it in the discrete (non opamp) ckts noted above?

[Nazar_lv]

the current output from 1545 must be with potential from 2 to 4V for 5Vpower supply, better 3.3v, resistor from power supply to Iref pin sets midscale output current and full scale output current.
In your schematic potential on current output 1545 will be 0v, input impedance of such i/u stage will be very big, all this is very very bad.

The principle of CC DAC is described in "A Self-Calibration Technique for Monolithic High-Resolution D/A Converters" by D. WOUTER J. GROENEVELD, HANS J. SCHOUWENAARS,
HENK A. H. TERMEER, ANDCORNELIS A. A. BASTIAANSEN
i uploded it here http://www.diyaudio.org.ua/viewtopic.php?f=20&t=205

PS sorry for my bad english

[sigurd]

Originally Posted by Nazar_lv

the current output from 1545 must be with potential from 2 to 4V for 5Vpower supply, better 3.3v, resistor from power supply to Iref pin sets midscale output current and full scale output current.
In your schematic potential on current output 1545 will be 0v, input impedance of such i/u stage will be very big, all this is very very bad.

The principle of CC DAC is described in "A Self-Calibration Technique for Monolithic High-Resolution D/A Converters" by D. WOUTER J. GROENEVELD, HANS J. SCHOUWENAARS,
HENK A. H. TERMEER, ANDCORNELIS A. A. BASTIAANSEN
i uploded it here DIYaudio.org.ua • ????

PS sorry for my bad english

Vref: intern DAC 11K : ground to Vref ( 7 )- Vref extern resistor 11K to 3,3 volt supply rail : so the voltage of Vref is 1,65 volt and in the middle of 3,3 volt, and in the middle of the two resistors ( 11K and 11K )
I use a TL431 adjust to 3,3 volt ( with capacitor 47µF from input TL431 to + rail )
This seems me the best.

[Eric Juaneda]

Hello sigurd,

Originally Posted by sigurd

Vref: intern DAC 11K : ground to Vref ( 7 )- Vref extern resistor 11K to 3,3 volt supply rail : so the voltage of Vref is 1,65 volt and in the middle of 3,3 volt, and in the middle of the two resistors ( 11K and 11K )
I use a TL431 adjust to 3,3 volt ( with capacitor 47µF from input TL431 to + rail )
This seems me the best.

Folowing Philips recommendations you are right. With other configuration, you can move forward.

The sound quality of the TDA1545A is directly dependant on how you drive IREF pin. In my Jundac Five DAC kit, I use a reference tension with very low output impedance to drive IREF pin.

From Diyaudio.com: IREF, the secret of the TDA1545A

[sigurd]

Originally Posted by Eric Juaneda

Hello sigurd,

Folowing Philips recommendations you are right. With other configuration, you can move forward.

The sound quality of the TDA1545A is directly dependant on how you drive IREF pin. In my Jundac Five DAC kit, I use a reference tension with very low output impedance to drive IREF pin.

From Diyaudio.com: IREF, the secret of the TDA1545A

Thanks for your reply.
Instead of the 11K resistor from + rail ( 3,3 volt TLE431 ) to Vref ( 7 ) a BC550c emitter follower : on the basis of the transistor a capacitor and parallel a resistor to ground : MKP? , PIO ?, TANTALUM,? Output emitter voltage can set by this resistor . The quality of the capacitor affects very much the sound.
I try an capacitor direct on the Vref pin , but that seems me not so good.
BTW your DAC is splendid, also the I/V conversion, simple but very effective.
Thanks for your help.

[sigurd]

Originally Posted by sigurd

Thanks for your reply.
Instead of the 11K resistor from + rail ( 3,3 volt TLE431 ) to Vref ( 7 ) a BC550c emitter follower : on the basis of the transistor a capacitor and parallel a resistor to ground : MKP? , PIO ?, TANTALUM,? Output emitter voltage can set by this resistor . The quality of the capacitor affects very much the sound.
I try an capacitor direct on the Vref pin , but that seems me not so good.
BTW your DAC is splendid, also the I/V conversion, simple but very effective.
Thanks for your help.

Also the bypass of the DAC is very important , 0,1µF ceramic, 1µF Wima,
10µF oscon, 220µF BG and with a bead.

[Eric Juaneda]

Hi sigurd,

Originally Posted by sigurd

Also the bypass of the DAC is very important , 0,1µF ceramic, 1µF Wima,10µF oscon, 220µF BG and with a bead.

Sure, the power supply is important too. You are using a very strong configuration.
In my personal experience, the ferrite bead degrade audio rendering. I forget this solution in analog section.