Asking for help with debugging my first DAC

[Alexium]

Hello!

I'm building my first ever DAC (and a first electronic device I intend to finish, for that matter). It is based on PCM1798, which has 4 mA p-p current output. I decided to test the PCM IC with the simplest I/U possible, so I've built this (sorry for hand-written schematic):

An externally hosted image should be here but it was not working when we last tested it.

The OA is OPA2134PA powered with +/- 12V voltage. So far I've only tested the circuit with zero signal, an got 2V DC on the OA output (0.7V DC on R1). The resistors are 5% in precision, but I suppose 5% can not account for 2V. Any suggestions?

P. S. The schematics is not of my design, and I'm a noob in electronics, could you please give me a highlight on the principle of this circuit operation? Particularly, I don't understand the R2-R3/R2-R4 circuits. R2-R3 is a voltage divider (why would one want to divide voltage that's very low already?), but R2-R4 doesn't look like a divider at all (unless output potential is zero which it shouldn't be)...

Thanks in advance!

P. P. S. I've doublechecked the assembly, looks like everything is in accordance with the drawing.

 

[Alexium]

Okay, scratch that. PCM179x has unipolar output; to perform passive I/U with a resistor it should be connected to Vcc and have have such a resistance that ensures nominal current at the center of the scale. For 1798 that's 2 kOhm which is said to kill the linearity.

Sorry for bothering you with such a dumb question.

 

[Calvin]

Hi,

if using a resistor this resistor should be connected to gnd, because the DAC is a current sourcing DAC. If it were a current sinking DAC (if I remember corrctly AD1965 et al) the resistor should be connected to a positive potential.
Don´t forget that the PCMs signal current is centered around the ´centre current´ (-3.5mA, see p. 23 of datasheet). So at idle the (offset) voltage over the resistor takes up the value: centre-current times resistor-value.
Feeding into a difference amplifier, only the difference in input signals is amplified and the common to both inputs signal is supressed. This way You get rid of the offset voltage generated by the center current and the output of the difference amplifier contains just of the amplified signal-voltage.
The probelm with the PCM-DACs is rather, that their outputs are diode protected, which limits the allowed voltage swing. Hence the value of the resistor is limited to small numbers. The gain of the difference amp needs to be considerable to achieve sufficient high output voltage.

jauu
Calvin

 

[Alexium]

Thank you! I've been misleaded by some info from the internet. Have to refer to the datasheet myself next time.
Back to square one then: what's wrong with my circuit?..

 

[Calvin]

Hi,

what´s wrong?
Basically nothing, but:

The problem with the PCM-DACs is rather, that their outputs are diode protected, which limits the allowed voltage swing. Hence the value of the resistor is limited to small numbers. The gain of the difference amp needs to be considerable to achieve sufficient high output voltage.

220Ohms times 3.5mA leads to an Offset voltage of 770mV. Added +-440mV of peakpeak signal voltage. The protection diodes will be conducting. After my knowledge a I/V-conversion resistor should not be considerably larger than 22Ohms to not spoil the low distortion figures the DAC is capable of.

jauu
Calvin