dac I/V convertion with very low distortion

[smms73]

Just because much of you guys are curious to see the schematic of the I/V converter, this is what i have right now. Possible i will change some values but the final circuit will be very close to this.


second image is distortion with input impedance of 1Kohm 8ma p-p, third is 3Kohms 8ma p-p.

As you can see distortion is very low even at low input impedances.

[smms73]

Forget to post this one that is with 3 volts rms output, 3 kohms input impedance. no visible distortion, distortion just start to became visible at 4.5 V rms.

[smms73]

Today i have not much time to show you another test that i made with different´s output connection, but tomorrow i will post that.

[Ken Newton]

Sergio, very cool. Thanks, for posting the schematic. I'll offer further comment after I've had opportunity to model the circuit in my simulator. Specifically, I'm curious to examime how sensitive the circuit is to producing D.C. output offset, due to small amounts of parameter mismatch between the two complementary current sources.

[regal]

Originally Posted by smms73

Forget to post this one that is with 3 volts rms output, 3 kohms input impedance. no visible distortion, distortion just start to became visible at 4.5 V rms.

This is amazing. Can you show a photo of you setup? Did you have to go through hundreds of transistors find a low dc offset?

Very elagant looking circuit, really hope it pans out.

[georgehifi]

Originally Posted by Ken Newton

Sergio, very cool. Thanks, for posting the schematic. I'm curious to examime how sensitive the circuit is to producing D.C. output offset, due to small amounts of parameter mismatch between the two complementary current sources.

Yes this would be interesting, hate to see coupling caps used with this, then a dc servo wouldn't be too bad if needed.
Also Serg you showed 1k at 8mA PtoP source, how does it behave with something less like a PCM1704 which is 1Kohm but only 1.2mA PtoP? even better hopefully? Also what is the output impedance of this setup?
PS sorry so many questions how does it also behave over 100khz and above, input impedance wise and output distortions?
Cheers George

[Ken Newton]

Originally Posted by georgehifi

Yes this would be interesting, hate to see coupling caps used with this, then a dc servo wouldn't be too bad if needed.
Also Serg you showed 1k at 8mA PtoP source, how does it behave with something less like a PCM1704 which is 1Kohm but only 1.2mA PtoP? even better hopefully? Also what is the output impedance of this setup?
PS sorry so many questions how does it also behave over 100khz and above, input impedance wise and output distortions?
Cheers George

George, the output impedance is essentially set by the value of the I/V resistor utilized. The I/V resistor value itself is determined by the fullscale DAC output current available for developing a 2VRMS signal level via Ohm's law. So, for the schematic shown by Sergio, the output impedance is 375 ohms.

For the PCM1704, I believe the fullscale output current is 1.2mA peak. The I/V resistor value needed to develop 2VRMS would be: 2.8Vpk / 1.2mApk = 2.35kohms. Therefore, the output impedance essentially would also be 2.35kohms.

[MiiB]

SMMS... do you balance the circuit by unbalancing the current sources, or do you inject a counter current at the input... or do you use a transformer to get rid of the current offset...??

[georgehifi]

Thanks Ken, so this is not an ideal circuit for a PCM1704 then? as to get 2v out, the loading for the dac would be too high at 2.35kohms? When the ideal should be much lower, around 10-100ohms. Is this correct? still learning about I/V's.

Cheers George

[regal]

Originally Posted by Ken Newton

George, the output impedance is essentially set by the value of the I/V resistor utilized. The I/V resistor value itself is determined by the fullscale DAC output current available for developing a 2VRMS signal level via Ohm's law. So, for the schematic shown by Sergio, the output impedance is 375 ohms.

For the PCM1704, I believe the fullscale output current is 1.2mA peak. The I/V resistor value needed to develop 2VRMS would be: 2.8Vpk / 1.2mApk = 2.35kohms. Therefore, the output impedance essentially would also be 2.35kohms.

So let's say we have minimal DC offset. It may be a good idea to put some filtering after this and then a buffer. I'm thinking the DCB1.