it dosnt work!!!(TDA1543)

[takashi]

i want to ask some question about TDA1543.
is it only support I2S format?
i use a AD1892 to work with TDA1543, it dosnt work.

then i use CS8414, it work best in I2S mode. but the Analog output of the TDA1543 is too small(about 100mVp-p), and the wave have many high frequency in it.
(i use a 1K hz test CD to do the test.)


i use 1K resistor to connect pin 8 and 7 to GND, the supply voltage is 8V.

please help!

 

[jean-paul]

TDA1543 is I2S only and when CS8412 or CS8414 is used in the right mode ( 2 ) you should have an outputvoltage of at least 0.7 V. This is with 5 V supply and 3 x 1 k resistors. Try to lower your supply voltage to 6 V and see what happens. Not all TDA1543 like 8V. Tried it but it clipped severely.

BTW did you play the right track for measuring output level ?

Please see this site for correct setting of the I2S mode:

http://www.geocities.com/rjm003.geo/rjmaudio/diy_dac.html

 

[vuki]

Have a look at this thread - might help:
http://db.audioasylum.com/cgi/m.pl?forum=tweaks&n=32623&highlight=tda1543&session=

also

http://www.diyaudio.com/forums/showthread.php?s=&threadid=10838&highlight=tda1543

 

[scottnixon]

Where's pin 6?

Takashi.. you need to load pin 6 (left out) also with a 1k and drop your voltage to 5V if you want to use that value 1k for vref and IV. Vref needs both dac outputs loaded, otherwise it clips severely and is not at all usable. 8V supply requires slightly taller resistors.

 

[fedde]

Re: Where's pin 6?

Takashi.. you need to load pin 6 (left out) also with a 1k and drop your voltage to 5V if you want to use that value 1k for vref and IV. Vref needs both dac outputs loaded, otherwise it clips severely and is not at all usable. 8V supply requires slightly taller resistors.

What would you suggest using ? I now use my TDA1543 dac with 8V and 3x 1K. The sound is quite ok, though the sound could be more controlled. (I need a buffer)

Thanx,

Fedde

 

[takashi]

thanks a lot !

it work finally!

i use 2k in pin7
and 4.7k for pin 6 and 8

 

[jean-paul]

4k7 and no clipping ?!?!?!

 

[rbroer]

Takashi, what are you doing

You either got a very strange TDA1543 or it is screwed up (can happen btw).
The Riv is way too large (4.7k) to stay within DC voltage compliance of the DAC.

Read the datasheet and you will understand.

You must limit the AC voltage swing to stay within 1.8V and Vsupply-1.2V.

In your case with Vsupply of 8V, your maximum swing is 1.8V to 6.8V, that is 5V.
With an assumed output current of 2.3mA, this results in a maximum Riv of 2k17 (5V/2.3mA).
You need a biascurrent of 0.828mA (1.8V/Riv), so Iref needs to be 0.414mA, resulting in a Rref of 5k31 (2.2V/Ibias/2)

I would advise to select Riv < 2k17 and to use a pot for the bias.
Adjust the pot such that with no music playing the output voltage on pin 6 and 8 is in the middle of your DC voltage compliance range, in your case (8V-1.2V-1.8V)/2 + 1.8V = 4.3V

 

[fedde]

Thank you Rudolf...

I already had the plan to use three potmeters to find the best sounding configuration. This is a new motivation to actually try that

I wonder if it would be beneficial to increase the 100 K resistor after the cap to 200-300 K !? Did you ever drive your passive I/V dacs into amps with an input impedance of 10K-20 K?

Fedde

 

[jean-paul]

I wonder if it would be beneficial to increase the 100 K resistor after the cap to 200-300 K !?

Why ? If you want to change the value you will hear a loud plop when changing interconnects assuming you use big output caps. When a low value is used this won't be a problem but it will not have a great influence on the DAC when you raise the resistance.

Did you ever drive your passive I/V dacs into amps with an input impedance of 10K-20 K?

You obviously need a buffer / amplifying stage Fedde ! BTW 3 x 1 K is not optimal with 8 V supply voltage as you can calculate with the excellent example Rudolf gave. 3 x 1 k is OK with 5 V supply voltage.

BTW Thank you Rudolf for the very good example. You obviously have the ability to understand Philips datasheets

 

[fedde]

I just tried 1.5 K. Funny, the output of the DAC has exactly 4.3 V DC with the 1K resistor on Vref I already had ;-)
So the configuration 2x1.5 K and 1K should work OK! With 3x1 K maybe there was some clipping !? At least that setting was suboptimal.

It seems to sound better, tomorrow I'll be able to tell more...

Fedde

 

[Pedja]

Hello fedde and jean-paul,

Low impedance should not be a problem when you have a current source and passive I/V conversion. Current IS converted to voltage, but nevertheless DAC's current output remains the only source. So, amp’s low input impedance (or that 100k resistor) is not a problem, DAC is already loaded by 1k and it’d like even less impedance.

I’m not against the (current or voltage) buffer, but it seems the necessity for it comes from some other side.

Pedja

 

[fedde]

Hello fedde and jean-paul,

Low impedance should not be a problem when you have a current source and passive I/V conversion. Current IS converted to voltage, but nevertheless DAC's current output remains the only source. So, amp’s low input impedance (or that 100k resistor) is not a problem, DAC is already loaded by 1k and it’d like even less impedance.

I’m not against the (current or voltage) buffer, but it seems the necessity for it comes from some other side.
Pedja

I've to think longer about this topic. Interesting view. I hope that the new resistor configuration will help the tightness at loud levels. I can't check because the time (2:36 am). At low listening levels the tightness seems to be MUCH better. Tomorrow I'll give my final judgement

I wonder what makes the sound bad in the case of low impedance coupling. I think that you get some sort of compression. Don't know what the cause yet.

Fedde

 

[takashi]

i am so sorry, i am new in tda1543.

i find that 1k resistor is used, the sound clip and very loud noise.
then i use 3 variable resistor to tune the sound, i find that pin 7 use 2k, and pin8, 6 use 3-5K sound best.

i think i need to check one more time...

and one more question, what is the relationship between resistor of pin 7 and Ibias, and what is the relationship between Ibr/Ibl
and Ibias.

thanks

 

[jean-paul]

You must limit the AC voltage swing to stay within 1.8V and Vsupply-1.2V.

In your case with Vsupply of 8V, your maximum swing is 1.8V to 6.8V, that is 5V.
With an assumed output current of 2.3mA, this results in a maximum Riv of 2k17 (5V/2.3mA).
You need a biascurrent of 0.828mA (1.8V/Riv), so Iref needs to be 0.414mA, resulting in a Rref of 5k31 (2.2V/Ibias/2)

I would advise to select Riv < 2k17 and to use a pot for the bias.
Adjust the pot such that with no music playing the output voltage on pin 6 and 8 is in the middle of your DC voltage compliance range, in your case (8V-1.2V-1.8V)/2 + 1.8V = 4.3V

Says it all Takashi, just like the hard-to-read datasheet does.

 

[fedde]

Well after listening for a day with my upgraded dac I/V configuration I can only say it's MUCH better with the 1.5 K resistors!!! For a while I had one channel with 1K and 1.5 K to compare. It's a night and day difference (besides from the volume change). More life, dynamics, tighter bass, more control etc. Well just try it if you still have 3x1 K and 8V supply...

Maybe I'll later try something like 2x1,67 K and 1.2 K ref, could even be better...

Fedde

 

[jean-paul]

Low impedance should not be a problem when you have a current source and passive I/V conversion. Current IS converted to voltage, but nevertheless DAC's current output remains the only source. So, amp’s low input impedance (or that 100k resistor) is not a problem, DAC is already loaded by 1k and it’d like even less impedance.

I’m not against the (current or voltage) buffer, but it seems the necessity for it comes from some other side.

Hi Pedja, the DAC is nevertheless very sensitive to low(er) impedance loads. It sounds best with high input-impedance amps.