Ironically the 5534 appears to be the major noise contributor.
I sat down one night and calculated all the noise contributions for the dual
mono circuit in data sheet. It's quite complex because there are many noise
sources all adding RMS however I came to a figure that showed the 132dB DR
in mono mode was actually (slightly) I-V dominated.
Having said that, no one is going to be using 9V RMS out so in real world
use, getting 132dB DR will not happen.
Under 600 is fine, it just depends on V swing. +-4mA / phase is within most
opamps capability and easily driven by the better ones.
If you raise the I-V opamps FB resistor to 1k you will actually get slightly less noise.
The reason: For double the resistance value, noise multiplies by 1.414 but there will be 2 x OP signal so SN will be slightly better all other things being equal.
The distortion will be dependent on a few things. a/ opamps current load so same in all cases but with greater OP swing, given most of the distortion will be OP stage then overall should be slightly lower with more swing.
b/ opamp gain - this will be determined by DAC OP Z versus FB R. The 1794 is almost perfect current source so OP Z is very high resulting in I-V being close to unity gain. Also the I-V has no common mode induced distortion (see Samuel Groners opamp measurement papers.
What this all amounts to is that the 1794 has a good mix of parameters for very low distortion I-V with good dynamic range.
So THD will improve if voltage swing is 3Vrms vs 4.5Vrms ?
I have an active stage and I am trying to fine tune it. My feedback R is 1.5K (IV stage) and I have 2.8ma. Would I have any benefit to lower the voltage swing by 30% ?
Also on the summing , I have a Rin of 1k. Is that a bit high ?
Last edited:
I also find that passive I/V works very well the PCM1794A. It's super transparent sounding. I've measured 0.002% THD at the output pins of my PCM1794A DAC utilizing 75 ohm I/V resistors. In addition, Audio Research utilized 200 ohm passive I/V resistors in one of their recent PCM1794A based commercial DACs. Also, the popular DIY DDDAC1794A project utilizes even larger value passive I/V resistors. So, there doesn't appear to be protection diodes on the outputs of the PCM1794A. The THD does increase as the resistor value is increased.
Search this site for the excellent posts by contributor [smms73], who has performed extensive spectrum analyzer measurements of the PCM1794A driving various I/V circuits, both in single-ended mode and in differential.
The problem with active virtual ground transimpedance amplifiers is that DAC chips feed them a wideband, fast slewing, fast settling signal current which is difficult to track without dynamic errors. In addition, the full amplitude of the DAC chip's output current must simultaneously be handled by the amplifier. This isn't simply an issue of whether the amplifier can output the required current at D.C., or at audio frequencies. It's an issue of the amplifier outputting the required current at high the frequencies of a fast slewing DAC output.
So, with the PCM1794A, an op-amp transimpedance stage must swing up to +/-3.9mA while attempting to accurately track the fast slewing output signal. It makes no difference to the current swing demanded whether the output voltage amplitude is set to a low level. The current which must be processed/handled is always the same as is output by the DAC chip. Such a signal places maximum dynamic demand upon an feedback based transimpedance amplifier.
How does passive I/V help? For some DAC designs, passive I/V is utilized to develop a full-scale amplitude signal, eliminating the need for any output amplifier - along with it's potential dynamic errors. For some other designs, passive I/V is utilized to develop a relatively low amplitude signal voltage, which is then amplified by a zero feedback voltage amplifier, either solid-state or vacuum tube. Passive I/V can even help lower the dynamic errors of feedback based amplifiers, by enabling use of standard inverting, or non-inverting voltage gain stages, thereby significantly lowering the magnitude of the current swings which must otherwise be handled by the output driver of an virtual-ground transimpedance amplifier.
diyralf was saying the contrary (lower thd with higher voltage) but I suspect he meant THD+N and not simply THD.
THD+N improves with higher voltage swing, not THD by itself. If you look at graphs of THD+N vs output voltage (at a constant load), it goes down first because the noise component improves at higher voltages. Then it usually starts to flatten because the distortion component rises as the opamp is asked to provide more power. And it finally rises when the output nears the PS rails and clipping finally sends THD trough the roof.
Which is good but not great for a pcm1794 (-94db). smms73 suggested keeping a passive I/V resistor under 50ohms (he suggested 22R) to achieve closer to 0.0002% thd and repeatedly said that 200R was a bad idea.Ken Newton said:
Last edited:
I wasn't suggesting that 200 ohms is good, but merely pointing out that a well respected DAC vendor utilizes that value. As I stated, lower is better for the resistor value. Never the less, -94dB is an inaudible level of THD. As I had offered, THD shouldn't simply be the main concern with I/V circuits, but rather the overall dynamic behavior.
Last edited:
Hi, ralf. I'm uncertain of whom your comment was addressed. If it was to me, the I/V resistor values I've been mentioning are for a totally passive implementation. The I/V resistor is connected between the DAC chip's current output pin and ground, but is not driven in conjunction with an active stage.
No, not necessarily. Refer to the DDAC1794A for an example. Also, there are a number of TDA1541 based, and other types, which don't either.
Even if utilizing an active voltage amplifier after the passive I/V, it makes possible the use of a traditional inverting, or an non-inverting voltage amplifier configuration. This then easily enables implementation of a much higher value feedback resistor, eliminating concern about overloading the op-amp's output.
You mean something like 330 ohm and then a low pass diff to single converter?
I mean, something like 10k ohm, or however high you desire it to be. Depending on the noise trade-off you're willing to accept.
Last edited:
That's only true for a transimpedance (current to voltage) configuration. I'm talking about an voltage amplifier (voltage to voltage) configuration to be located just after the passive I/V resistor.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.