Take a DAC chip, say the PCM1794, it has a +ve and-ve output for both the left and the right channel.
These in turn get fed into individual opamps that perform I/V conversion.
Next the +ve and -ve voltages get fed into another opamp as a differential.
Naturally one connects the +ve output from the I/V into the +ve input pin of the differential opamp. Then the -ve output from the I/V into the -ve input pin of the following opamp.
My question is can you connect the +ve I/V output into the -ve input pin of the differential and likewise -ve I/V into the +ve input pin?
I can't think of any reason why this shouldn't work.
These in turn get fed into individual opamps that perform I/V conversion.
Next the +ve and -ve voltages get fed into another opamp as a differential.
Naturally one connects the +ve output from the I/V into the +ve input pin of the differential opamp. Then the -ve output from the I/V into the -ve input pin of the following opamp.
My question is can you connect the +ve I/V output into the -ve input pin of the differential and likewise -ve I/V into the +ve input pin?
I can't think of any reason why this shouldn't work.
kevinkr said:
Yes exactly that's all I thought would happen. The phase inversion doesn't bother me and physically on the PCB the outputs from the I/V match up with the inputs of the differential in that fashion.
+ => -
- => +
I'm already using the THS4031 for I/V conversion and I couldn't be happier with the sound. I believe you were the one who responded, suggesting I place a resistor in series with the feedback cap, to help cure highish distortion.
Is there any reason why the THS4131 should be better then the THS4031 in this application? I am thinking about following them with a TPA6120.
I've used the OPA1632 before and it seems very similar to the THS4131, just marketed in a different way.
Hi,
Your explanation is a bit difficult to understand. It looks as if You´re mixing up with the currents and voltages (The PCM1794 doesn´t feature ve+ or ve- outputs but Iout+ and Iout-)
Referring to Fig. 24 of the Datasheet (You know, references are a big help to get everybody ´in phase´ with the problem
)
ve+ should be the voltage at Pin6 of U1 and v- the voltage at Pin6 of U2. Since the outputs of U1 and U2 see equal impedances You can interchange them. So Pin6/U1 going to R4 and Pin6/U2 going to R3 will only change the phase of the voltage at the output of U3.
You have to do the same tweak of course with the second channel, so that they stay in phase with each other, but just the overall phase changes.
The main difference between THS4031 and THS4131 is that the THS4131 is fully differential, i.e. it features 2 out-of phase outputs, whereas the THS4031 is a classical one output device. If You run the PCM1794 in stereo mode (Fig. 24+25) and You want symmetrical outputs the THS4131 would be the choice. Running the PCM1794 in mono mode (Fig.26) the THS4031 would be the first choice.
Other than that, the THS4031 is a bit ´faster´. The settling times are shorter, which might be a factor when used as I/U converter.
All other data is either similar or exceeds the demands anyway.
There is a dual variant of the THS4031, the THS4032. Might be interesting for even more compact signal routing and smaller layouts.
jauu
Calvin
Your explanation is a bit difficult to understand. It looks as if You´re mixing up with the currents and voltages (The PCM1794 doesn´t feature ve+ or ve- outputs but Iout+ and Iout-)
Referring to Fig. 24 of the Datasheet (You know, references are a big help to get everybody ´in phase´ with the problem
)ve+ should be the voltage at Pin6 of U1 and v- the voltage at Pin6 of U2. Since the outputs of U1 and U2 see equal impedances You can interchange them. So Pin6/U1 going to R4 and Pin6/U2 going to R3 will only change the phase of the voltage at the output of U3.
You have to do the same tweak of course with the second channel, so that they stay in phase with each other, but just the overall phase changes.
The main difference between THS4031 and THS4131 is that the THS4131 is fully differential, i.e. it features 2 out-of phase outputs, whereas the THS4031 is a classical one output device. If You run the PCM1794 in stereo mode (Fig. 24+25) and You want symmetrical outputs the THS4131 would be the choice. Running the PCM1794 in mono mode (Fig.26) the THS4031 would be the first choice.
Other than that, the THS4031 is a bit ´faster´. The settling times are shorter, which might be a factor when used as I/U converter.
All other data is either similar or exceeds the demands anyway.
There is a dual variant of the THS4031, the THS4032. Might be interesting for even more compact signal routing and smaller layouts.
jauu
Calvin
+ve and -ve are short hand ways of writing positive and negative, there was never any mention of voltage or current. I am fully aware that the DAC has current output, hence the I/V stages.
Yes which is what I originally thought, I was just checking to see if there was something else going on that I might have missed.
This I'm having trouble understanding however.
The THS4131 or 4031 are being used in the I/V stage, not as the differential.
The idea being that you feed the positive and negative current into either, one 4131, or two 4031. After I/V you've got positive and negative voltages. I'm assuming that you meant at this stage use a 4131 to preserve it being balanced, or use a 4031 to convert the balanced signal into single ended.
This isn't how the chip is used, which is why I'm not fully understanding what you meant.
If the DAC chip is used in stereo mode for two single ended outputs, or mono mode for balanced outputs, the configuration for the analogue circuitry is exactly the same.
In the first case you end up with two voltage outputs, one carries the left channel of audio, the other carries the right channel of audio.
In the second case, the 'left channel' handles the positive side of the balanced signal and the 'right channel' handles the negative side of the balanced signal.
This is why you need two DAC chips for balanced operation. One chip handles the left channel and one chip handles the right channel.
Are we on the same page now?
I think I understand where the confusion lies, the PCM1794 is a differential stereo dac - i.e. there is a differential current output for each of the two channels.
In mono mode these may be summed in the current domain (i.e. ahead of the I//V converters) however in mono one output pair is inverted relative to the other and the current outputs need to be swapped on one channel.
Infact I use the PCM1794A dac ic in mono mode and sum the differential output currents as described above and use them to drive a THS4131 which in turn drives a transformer to convert from balanced to unbalanced output. Accidentally the best sound I have ever heard from any PCM dac I have owned. (Assemblage 1, PS Audio Ultralink, and Zhaolu 2.5A)
In mono mode these may be summed in the current domain (i.e. ahead of the I//V converters) however in mono one output pair is inverted relative to the other and the current outputs need to be swapped on one channel.
Infact I use the PCM1794A dac ic in mono mode and sum the differential output currents as described above and use them to drive a THS4131 which in turn drives a transformer to convert from balanced to unbalanced output. Accidentally the best sound I have ever heard from any PCM dac I have owned. (Assemblage 1, PS Audio Ultralink, and Zhaolu 2.5A)
Hi,
Your prob was about an issue of the filter stages (U3 of fig. 24) and maybe the following differential-to-SE stage (when in mono mode, fig. 26) and not about the I/V-stage (U1 and U2). My A referred only to Your Q. But rereading that part...forget it....I thought wrong
The PCM 1794 features a quite high current offset at its outputs. The I/V-converter must supply this current to the DAC. I assume that a single THS4131 might heat up very much since it has to supply the current for both outputs (-6,2mAx2). A dual of THS4031 need only to supply the current for one output (6,2mA) each. Depending on the supply voltages of the OPamps this means a difference of several tens of mW that the IC has to get rid of or not. So with the THS4131 I would think of using kind of a current compensation circuit, e.g a resistor from the DACs Iout to the positive supply (as a most simple variant). If the current compensation works precise, even the differential-to-SE-converter (U3) could be omitted with.
jauu
Calvin
Your prob was about an issue of the filter stages (U3 of fig. 24) and maybe the following differential-to-SE stage (when in mono mode, fig. 26) and not about the I/V-stage (U1 and U2). My A referred only to Your Q. But rereading that part...forget it....I thought wrong
The PCM 1794 features a quite high current offset at its outputs. The I/V-converter must supply this current to the DAC. I assume that a single THS4131 might heat up very much since it has to supply the current for both outputs (-6,2mAx2). A dual of THS4031 need only to supply the current for one output (6,2mA) each. Depending on the supply voltages of the OPamps this means a difference of several tens of mW that the IC has to get rid of or not. So with the THS4131 I would think of using kind of a current compensation circuit, e.g a resistor from the DACs Iout to the positive supply (as a most simple variant). If the current compensation works precise, even the differential-to-SE-converter (U3) could be omitted with.
jauu
Calvin
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.