How far can IV converter be from DAC?

[philpoole]

Hi All,

I'm going to build a TDA1541A dac module, and I want to experiment with different I/V conversion techniques (i.e. passive, opamp, discrete). If I could have the I/V conversion on an adjacent board, that would be great.
How short a trace must the current output of a DAC be? or, IOW, what could I get away with?
I assume it would be particularly sensitive to interference before I/V conversion.

Cheers,
Phil

 

[lcsaszar]

Since the I/V converter represents a very low load resistance, it is not susceptible to stray fields. The DAC drives it with high output impedance, so the serial resistance of the tracks has not much influence either. The stray capacity between the tracks is also shunted by the low I/V input resistance. I would say 10" should not be a problem. Care should be taken to proper ground paths (I/V ground to DAC analog ground).

 

[philpoole]

Hi Oshifis,

Thanks for your reply.
I was hoping to get away with about 6", so you're reply is great news.

On my last/first DAC, the I/V converter was about an inch away from the DAC, on the same circuit board, so I've had no experience of what I can get away with.

Thanks again,

Cheers,
Phil

 

[TimA]

Phil,

Do you have any interesting findings regarding different I/V circuits with the TDA1541A?

Tim.

 

[philpoole]

Hi Tim,

Alas, I keep getting distracted.
When I get the time, I'm juggling between building a dedicated PSU for my DAC, trying to get oversampling going with an SAA7220, or building my new TDA1541A DAC that will be powered by the new PSU (will probably be a balanced output with several DAC chips - and hopefully oversampling if I get it working).

Anyway, I hope the new power supply will be suitably quiet enough to try different IV stages.
So far, I've only done the standard Opamp IV conversion (which sounds good to be fair - without much frame of reference), and investigated different opamps. However, I want to do a discrete stage, based on a common base amplifier, and maybe a FET equivalent common gate amplifier.
Currently, I'm not too bothered in a passive stage, but I am interested in removing feedback - to see what happens.

If I ever get round to it!
Cheers,
Phil

 

[TimA]

Phil,

All that should keep you nicely occupied.
FWIW, in my NOS TDA1541A DAC the AD844 is used as common base I/V. The DAC has two outputs, one from the AD844's internal buffer, the other from a simple(r) FET buffer. To my ears both have a sound which is more 'organic' than a standard op-amp. The FET buffer can significantly improve upon the AD844's internal buffer but I acheived this only after minimising/maximising the FET buffer's component count/component quality.

Tim.

 

[philpoole]

Hi Tim,

That's interesting.
I've heard the AD844 mentioned on a few threads, but not really investigated it. That'll be another entry n my to do my list!

Its good to hear of people who've tried different types of IV with positive results.

I must get going with some of this.

Cheers,
Phil.

 

[jcx]

you might recall that current always flows in loops, all of dac output I has to have a low Z return path to the dac analog ground

this return current loop goes through the i/v op amp power pins - so now you need a low Z connection to the clean anaog power at the dac - or well isolated i/v power (batteries?) with a low Z gnd return

Draw the Loops! think about how to minimize the area(s) to minimize Z (inductance is porportional to loop area) and reduce coupling of external fields

the dac i out has high frequency switching glitch energy which make high frequency design consideration important - an unanticipated high frequency resonance could be excited by the switching spikes and ring - giving the i/v op amp input a signal it can't deal with linearly and causing intermodulation distortion that could fold down into the audio band