Hi All,
having a problem with I/V from a the PCM1704. The original DAC design called for a PCM63 and I substituted with the PCM1704 on a conversion board. Everything works fine but I noticed some high frequency loss > 12KHz. On investigation with a test audio CD I find that measuring the voltage at the top of the I/V conversion resistor the output starts at 1.4V @ 16Hz but drops to 400mV @20KHz. I have tried resistor I/V and other topologies such as valve SRPP but nothing seems to help. I also followed the theory and designs at:
http://www.fortunecity.com/rivendell/xentar/1179/theory/vasfda/vasfda.html
I have used a 270R and 0.012uf cap across the output as suggested in the above link.
I can’t seem to pin down why the output is varying so much with frequency? Even if I drop the cap and have nothing but the resistor from I/V output to ground the problem is apparent.
Can someone confirm the response of their I/V circuit and let me know or suggest what is wrong.
My test CD output response is flat if I use the CD player built-in DAC.
The details of the original DAC circuit can be found here:
http://members.chello.nl/~m.heijligers/DAChtml/dactop.htm
having a problem with I/V from a the PCM1704. The original DAC design called for a PCM63 and I substituted with the PCM1704 on a conversion board. Everything works fine but I noticed some high frequency loss > 12KHz. On investigation with a test audio CD I find that measuring the voltage at the top of the I/V conversion resistor the output starts at 1.4V @ 16Hz but drops to 400mV @20KHz. I have tried resistor I/V and other topologies such as valve SRPP but nothing seems to help. I also followed the theory and designs at:
http://www.fortunecity.com/rivendell/xentar/1179/theory/vasfda/vasfda.html
I have used a 270R and 0.012uf cap across the output as suggested in the above link.
I can’t seem to pin down why the output is varying so much with frequency? Even if I drop the cap and have nothing but the resistor from I/V output to ground the problem is apparent.
Can someone confirm the response of their I/V circuit and let me know or suggest what is wrong.
My test CD output response is flat if I use the CD player built-in DAC.
The details of the original DAC circuit can be found here:
http://members.chello.nl/~m.heijligers/DAChtml/dactop.htm
Assuming that you are making the measurement with something you know is accurate over that freq range, then it sounds like the de-emphasis flag is set on the digital filter.
BTW, the PCM1704 sounds much better into a low impedance, preferably below 10 ohms, which makes some type of current mirror the best bet for I/V conversion. Not too hard to get down into the below 10 ohm range without using negative feedback (other than just some local in the current source). Still sounds good with a resistor, but definitely try to keep it well below 100 ohms if you can, and do some controlled tests with closely matched volume once you get it working. You may be surprised how much the sound changes for the better into low impedances.
BTW, the PCM1704 sounds much better into a low impedance, preferably below 10 ohms, which makes some type of current mirror the best bet for I/V conversion. Not too hard to get down into the below 10 ohm range without using negative feedback (other than just some local in the current source). Still sounds good with a resistor, but definitely try to keep it well below 100 ohms if you can, and do some controlled tests with closely matched volume once you get it working. You may be surprised how much the sound changes for the better into low impedances.
Hi thanks for the reply and the advice over the I/V resistor value. I did manage to get this fixed at the weekend and you are absolutely correct. The deemphasis flag was being set incorrectly by the CS8414 - I don't know why but I hard wired it off which has eliminated the problem.
Thanks again.
Thanks again.
Re: Re: PCM1704 I/V response
I haven't personally done any listening tests on the PCM63, but my listening tests on some of the other BB parts would suggest that it will probably sound best at the lowest voltage on the output. This is probably true of all switched current source type DACs, even though the distortion seems to stay pretty low until the protection diodes begin conducting, though I think there will be some non-linearities caused with any voltage on the output. I was originally using a 100 ohm precision bobbin wound resistor on my PCM1704 design (which would be about the same voltage compliance you would have at 50 ohms) and following that with a discrete gain stage, so going to a simple current mirror with under 10 ohms impedance made a very big difference since it provided all of the gain I needed, and I only had to follow that with a single-ended current stage.
But the compliance test is pretty easy. You should use a preamp with buffered volume control to avoid frequency response changes at different volume settings since you will be reducing the output level of the DAC and compensating with the preamp. Take care to match the volume at each step with a test tone before listening. Use female pins on the circuit board so you can easily swap resistors in and out without powering down the DAC. Have fun, it might not make a significant difference to you, especially if you need more gain than provided by your design.
regal said:
I haven't personally done any listening tests on the PCM63, but my listening tests on some of the other BB parts would suggest that it will probably sound best at the lowest voltage on the output. This is probably true of all switched current source type DACs, even though the distortion seems to stay pretty low until the protection diodes begin conducting, though I think there will be some non-linearities caused with any voltage on the output. I was originally using a 100 ohm precision bobbin wound resistor on my PCM1704 design (which would be about the same voltage compliance you would have at 50 ohms) and following that with a discrete gain stage, so going to a simple current mirror with under 10 ohms impedance made a very big difference since it provided all of the gain I needed, and I only had to follow that with a single-ended current stage.
But the compliance test is pretty easy. You should use a preamp with buffered volume control to avoid frequency response changes at different volume settings since you will be reducing the output level of the DAC and compensating with the preamp. Take care to match the volume at each step with a test tone before listening. Use female pins on the circuit board so you can easily swap resistors in and out without powering down the DAC. Have fun, it might not make a significant difference to you, especially if you need more gain than provided by your design.
Like Black Heart I definitely found the smaller I/V resistor beneficial in my Assemblage Dac 1 with PCM1702 D/A converters. I ended up with 10 ohm resistors for the I/V conversion and liked the result much better.
Many of these older dacs have definite recommendations on compliance voltage, they were after all intended to drive virtual earths.. Some AD dacs were specified for as little as 20mV on the outputs before linearity issues set in.
Many of these older dacs have definite recommendations on compliance voltage, they were after all intended to drive virtual earths.. Some AD dacs were specified for as little as 20mV on the outputs before linearity issues set in.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.