This will be more of an impressions paper rather than a full fledged review. I haven't got much experience with DACs and a Buffalo3+Ivy is the best I've heard. This is the first multibit DAC I've heard so I cannot tell if it has *that* multibit sound or not. I just know that it sounds different to all of the DACs I've heard before.
First of all some facts- this DAC was made in the year 1995 which wasn't the golden age of audio (I regard 80-ies being the Golden age) but one could say it was the time when the CD was at its height. The D/AC-1600 along with its cousins was meant to accompany disc transports and maximize the sound quality that can be extracted from the data they provide. All of these units offer a no frills design that I happen to love that focuses on ruggedness inside and utilitarian looks. On the paper the D/AC-1600 ticks all of the boxes that a vintage DAC seeker may look for- quad K grade PCM63 chips and an equally venerable PMD100 receiver. I'm quite sure that five of these chips alone may fetch a pretty hefty sum at the bay. Other than that the device features four input options- optical TOSLINK, TS optical, AES/EBU XLR and SPDIF RCA. Truth be told I'd rather have a proper 75 Ohm BNC for the coax but as you've seen- very few consumer DACs feature those. Outputs range from digital SPDIF RCA for daisy chaining, analog RCA and XLR for single ended and true balanced outputs. If we look at the datasheet for the device you'll see that the output is fairly hot at 3VRMS and twice that on the XLR's. Most of the amps and pre-amps will be okay with this but, for example, the input stage of an O2 happily goes into clipping. Then again I'm not sure how many of the O2 crowd will choose this DAC, but it's something one must keep in mind. The front is pretty spartan with dedicated input selection buttons and a curious button that switches absolute polarity for those that care. I couldn't hear any difference with flipping polarity and the fact that the switch on my unit is broken seems to indicate that the previous owners also tried plenty to hear it. The front panel also features LEDs for indicating the selected input and the frequency it is being fed in. Let's pop the hood, shall we-
As you can see the unit has a pretty hefty power supply with separate transformers for the digital stage and one for each analog stage. On top of that there is heavy filtering with decent quality electrolytic capacitors that have been shunted with film caps. Actually I'm pretty sure that all of the electrolytic caps on this device are shunted with film caps which shows that costs were not spared for this device. I'll also mention that switching this device for 220V operation is a trivial matter of cutting three jumpers and soldering another two. Now let's focus on the conversion stage-
You can see that each channel uses two PCM63K chips, however when listening single ended you will utilize only two of them. Contrary to my initial thoughts the I/V stage is not discrete as it uses op-amps. The radiators that dominate the board are for the MOSFET follower current buffer. My first objection would be directed at this current buffer. First of all the MOSFET's used are not meant for audio use which means that they are not the most linear. Secondly- my unit had bias current a tad off on one channel which meant about 0.5V of DC offset. Re-biasing the MOSFET's without spectrum readings is pretty hard so I just set an arbitrary quiescent current of 40mA and hoped for the best. Luckily there are dedicated test wires for current readings and cutout holes for reaching the bottom PCB so the actual tweaking was a breeze. It seems that the output stage has a DC protection circuit that cuts off at some point but 0.5V didn't trigger it, so hadn't my amp had a DC servo my HD650 would've been fried. Good thing I had a relay based attenuator that gives out a purr when you are having DC at input. Pretty handy. All in all the output stage is pretty pedestrian save for the MOSFET buffer which I find silly. I'm sure one could omit the buffer entirely and get better results. I'll try a simple opamp/BUF634 composite circuit for the output buffer just for the kicks in place of the MOSFET's but honestly- one could do a lot better with a discrete output stage like our member Colin used to make. There's plenty of room for that. A simpler mod would be just to swap out the op-amps with something better as the op-amp tech has improved since the nineties. On the topic of modding Audiotuning.de has some drop-in kits for the digital receiver stage and DAC stages but other than adding the ability of receiving higher resolution signal I very much doubt that it would yield any actual improvement. The PCM1704 is only marginally if any better than a K grade PCM63 and PMD100 is still one of the if not the best digital receiver for 16/48 content. After all the bottleneck of this device, methinks, is the output stage- swap that out and then you could talk about hearing what the PCM63K has to offer. Also the PCM1704 is one of the more expensive DAC chips and I don't really see the point of buying the D/AC-1600 to overhaul it completely for this kind of money. More can be had for less if one goes for a clean slate via DIY.
Now for some practical matters. When you buy a [digital] audio device that has almost reached the legal drinking age you must keep in mind that it might have some quirks. First of all- check at what voltage it is wired. Second- keep in mind that the output stage is always on regardless of the power switch position. This has its advantages as many of the devices doesn't have to go through on/off thermal cycles. Then again the MOSFETs in my unit operate at 50 degrees Celsius and the case has almost no ventilation. That can shorten the life of electrolytic capacitors which there are many, however I can't see any signs of them going bad. When you look for the electrolyte leaks keep in mind that most of the caps are glued to the PCB for stability during transport and these brown splats are the aging adhesive. With age many of the zip locks have gone stiff and are really brittle. I recommend swapping out at least the ones on the trafos. When plugged in the top of the unit will get quite warm as it relies on the case to dissipate heat, sure it won't get stove-like but I'd recommend letting it breath. Heck, one day I might try CNC'ing some vents for it. The trade-off would be dust. My unit was completely dust free inside which is pretty remarkable for a device this old.
Now for the sound. My audio journey started with the FiiO E10 and then I upgraded to the Pico DAC. The E10 was really great for the price but didn't have the grunt for the HD650 to shine. It was better when I built myself a Dynalo amp and after a while I even bought the Pico DAC as it was highly regarded in the headphone society. I must, however admit that the transition from E10 to Pico didn't yield the improvement I was looking for. I know some folks talk about the Pico sounding better than the Benchmark DAC1 which I can't comment. Pico did sound very good but there was this feeling of remorse that haunted my listening sessions. I guess there is only so much you can do with an USB powered DAC. The final nudge was a listening session when I tried out the Buffalo DAC which blew my mind in many aspects. It had an immense soundstage among other things and showed that my system is DAC limited. At least I was convinced that there's more to be had.
For months I had eyed the classified in the FS section where Currawong was trying to sell this Parasound and finally when he threw in the shipping and an Audio-GD Digital Interface I pulled the trigger.
As I said earlier- rewiring the D/AC-1600 for EU voltage was a breeze, however setting up the ADG interface wasn't. There is a dedicated thread on HF about the woes that are involved in living with the Tenor chip Kingwa decided to use so that's another story. Let's just say that from three USB cables I have there is only one that lets the interface to connect in asynchronous mode to my PC. When it does, the result is nothing short of amazing.
First of all- coming from the Pico the Parasound had extremely realistic timbre. I know that Currawong did mention this in one his DAC reviews- the D/AC-1600 excels in portraying music in a very realistic manner. The rig I use consists of a second revision HD650 and a self-made Dynalo complete with a Joshua Tree 128 step relay-based discrete attenuator and a Sigma 22 PSU. I know of people calling the Gilmore design dry, cold and very analytical sounding but I can say that my Dynalo sounds anything but that. I have compared it with a Beta 22 and with both amps volume matched I found the Dynalo to be less aggressive and maybe a tad less resolving. In my opinion it is a perfect match to the HD650. Both of these devices are really smooth sounding, some might say- forgiving. The D/AC-1600 sound signature perfectly complements my Dynalo/HD650 rig. After some weeks of listening I will say that it might not be perfectly neutral. In my opinion it emphasizes bass a little, not much of sin with open headphones. Also the Parasound does treble in a very polite manner- it doesn't try to impress you with fake detail by emphasizing treble. I wouldn't use this DAC for reference purposes but for everyday music it's nearly faultless. It might not be the best setup for aggressive electronic music where more crunch and bite is needed, but it doesn't make the mistake of sacrificing resolution for musicality.
I'll conclude by saying that despite boasting big names in its parts list this DAC isn't high end (and never was). However if you can live with its limitations and size then it's a hell of a deal for 300 bucks. Especially if you take into account that it can do true balanced and you could improve it substantially by modding the output stage. The biggest price one has to pay when choosing this DAC is convenience, not money. Heck, I'm just happy that I have a set of rare digital chips if I ever decide to make my own DAC. I'm pretty sure that this DAC will never win any measurement contests, but what counts is that it has a type of distortion that never stands in the way of my music. Will, it work for you? It might and if worst happens then just sell it for the same cash you bought it.
UPDATE
After some surgical intervention that involves completely amputating the MOSFET output buffer I finally may have tapped into what this DAC has to offer. Also I realized how foolish it was of Parasound engineers to use the previous buffering solution. After installing the OPA134 with a BUF634 in the feedback loop the DAC came alive with improved soundstage, more controlled bass (the bass with the MOSFET buffer was pretty flabby and exaggerated). The next mod most likely will be modifying the digital filter.
First of all some facts- this DAC was made in the year 1995 which wasn't the golden age of audio (I regard 80-ies being the Golden age) but one could say it was the time when the CD was at its height. The D/AC-1600 along with its cousins was meant to accompany disc transports and maximize the sound quality that can be extracted from the data they provide. All of these units offer a no frills design that I happen to love that focuses on ruggedness inside and utilitarian looks. On the paper the D/AC-1600 ticks all of the boxes that a vintage DAC seeker may look for- quad K grade PCM63 chips and an equally venerable PMD100 receiver. I'm quite sure that five of these chips alone may fetch a pretty hefty sum at the bay. Other than that the device features four input options- optical TOSLINK, TS optical, AES/EBU XLR and SPDIF RCA. Truth be told I'd rather have a proper 75 Ohm BNC for the coax but as you've seen- very few consumer DACs feature those. Outputs range from digital SPDIF RCA for daisy chaining, analog RCA and XLR for single ended and true balanced outputs. If we look at the datasheet for the device you'll see that the output is fairly hot at 3VRMS and twice that on the XLR's. Most of the amps and pre-amps will be okay with this but, for example, the input stage of an O2 happily goes into clipping. Then again I'm not sure how many of the O2 crowd will choose this DAC, but it's something one must keep in mind. The front is pretty spartan with dedicated input selection buttons and a curious button that switches absolute polarity for those that care. I couldn't hear any difference with flipping polarity and the fact that the switch on my unit is broken seems to indicate that the previous owners also tried plenty to hear it. The front panel also features LEDs for indicating the selected input and the frequency it is being fed in. Let's pop the hood, shall we-
As you can see the unit has a pretty hefty power supply with separate transformers for the digital stage and one for each analog stage. On top of that there is heavy filtering with decent quality electrolytic capacitors that have been shunted with film caps. Actually I'm pretty sure that all of the electrolytic caps on this device are shunted with film caps which shows that costs were not spared for this device. I'll also mention that switching this device for 220V operation is a trivial matter of cutting three jumpers and soldering another two. Now let's focus on the conversion stage-
You can see that each channel uses two PCM63K chips, however when listening single ended you will utilize only two of them. Contrary to my initial thoughts the I/V stage is not discrete as it uses op-amps. The radiators that dominate the board are for the MOSFET follower current buffer. My first objection would be directed at this current buffer. First of all the MOSFET's used are not meant for audio use which means that they are not the most linear. Secondly- my unit had bias current a tad off on one channel which meant about 0.5V of DC offset. Re-biasing the MOSFET's without spectrum readings is pretty hard so I just set an arbitrary quiescent current of 40mA and hoped for the best. Luckily there are dedicated test wires for current readings and cutout holes for reaching the bottom PCB so the actual tweaking was a breeze. It seems that the output stage has a DC protection circuit that cuts off at some point but 0.5V didn't trigger it, so hadn't my amp had a DC servo my HD650 would've been fried. Good thing I had a relay based attenuator that gives out a purr when you are having DC at input. Pretty handy. All in all the output stage is pretty pedestrian save for the MOSFET buffer which I find silly. I'm sure one could omit the buffer entirely and get better results. I'll try a simple opamp/BUF634 composite circuit for the output buffer just for the kicks in place of the MOSFET's but honestly- one could do a lot better with a discrete output stage like our member Colin used to make. There's plenty of room for that. A simpler mod would be just to swap out the op-amps with something better as the op-amp tech has improved since the nineties. On the topic of modding Audiotuning.de has some drop-in kits for the digital receiver stage and DAC stages but other than adding the ability of receiving higher resolution signal I very much doubt that it would yield any actual improvement. The PCM1704 is only marginally if any better than a K grade PCM63 and PMD100 is still one of the if not the best digital receiver for 16/48 content. After all the bottleneck of this device, methinks, is the output stage- swap that out and then you could talk about hearing what the PCM63K has to offer. Also the PCM1704 is one of the more expensive DAC chips and I don't really see the point of buying the D/AC-1600 to overhaul it completely for this kind of money. More can be had for less if one goes for a clean slate via DIY.
Now for some practical matters. When you buy a [digital] audio device that has almost reached the legal drinking age you must keep in mind that it might have some quirks. First of all- check at what voltage it is wired. Second- keep in mind that the output stage is always on regardless of the power switch position. This has its advantages as many of the devices doesn't have to go through on/off thermal cycles. Then again the MOSFETs in my unit operate at 50 degrees Celsius and the case has almost no ventilation. That can shorten the life of electrolytic capacitors which there are many, however I can't see any signs of them going bad. When you look for the electrolyte leaks keep in mind that most of the caps are glued to the PCB for stability during transport and these brown splats are the aging adhesive. With age many of the zip locks have gone stiff and are really brittle. I recommend swapping out at least the ones on the trafos. When plugged in the top of the unit will get quite warm as it relies on the case to dissipate heat, sure it won't get stove-like but I'd recommend letting it breath. Heck, one day I might try CNC'ing some vents for it. The trade-off would be dust. My unit was completely dust free inside which is pretty remarkable for a device this old.
Now for the sound. My audio journey started with the FiiO E10 and then I upgraded to the Pico DAC. The E10 was really great for the price but didn't have the grunt for the HD650 to shine. It was better when I built myself a Dynalo amp and after a while I even bought the Pico DAC as it was highly regarded in the headphone society. I must, however admit that the transition from E10 to Pico didn't yield the improvement I was looking for. I know some folks talk about the Pico sounding better than the Benchmark DAC1 which I can't comment. Pico did sound very good but there was this feeling of remorse that haunted my listening sessions. I guess there is only so much you can do with an USB powered DAC. The final nudge was a listening session when I tried out the Buffalo DAC which blew my mind in many aspects. It had an immense soundstage among other things and showed that my system is DAC limited. At least I was convinced that there's more to be had.
For months I had eyed the classified in the FS section where Currawong was trying to sell this Parasound and finally when he threw in the shipping and an Audio-GD Digital Interface I pulled the trigger.
As I said earlier- rewiring the D/AC-1600 for EU voltage was a breeze, however setting up the ADG interface wasn't. There is a dedicated thread on HF about the woes that are involved in living with the Tenor chip Kingwa decided to use so that's another story. Let's just say that from three USB cables I have there is only one that lets the interface to connect in asynchronous mode to my PC. When it does, the result is nothing short of amazing.
First of all- coming from the Pico the Parasound had extremely realistic timbre. I know that Currawong did mention this in one his DAC reviews- the D/AC-1600 excels in portraying music in a very realistic manner. The rig I use consists of a second revision HD650 and a self-made Dynalo complete with a Joshua Tree 128 step relay-based discrete attenuator and a Sigma 22 PSU. I know of people calling the Gilmore design dry, cold and very analytical sounding but I can say that my Dynalo sounds anything but that. I have compared it with a Beta 22 and with both amps volume matched I found the Dynalo to be less aggressive and maybe a tad less resolving. In my opinion it is a perfect match to the HD650. Both of these devices are really smooth sounding, some might say- forgiving. The D/AC-1600 sound signature perfectly complements my Dynalo/HD650 rig. After some weeks of listening I will say that it might not be perfectly neutral. In my opinion it emphasizes bass a little, not much of sin with open headphones. Also the Parasound does treble in a very polite manner- it doesn't try to impress you with fake detail by emphasizing treble. I wouldn't use this DAC for reference purposes but for everyday music it's nearly faultless. It might not be the best setup for aggressive electronic music where more crunch and bite is needed, but it doesn't make the mistake of sacrificing resolution for musicality.
I'll conclude by saying that despite boasting big names in its parts list this DAC isn't high end (and never was). However if you can live with its limitations and size then it's a hell of a deal for 300 bucks. Especially if you take into account that it can do true balanced and you could improve it substantially by modding the output stage. The biggest price one has to pay when choosing this DAC is convenience, not money. Heck, I'm just happy that I have a set of rare digital chips if I ever decide to make my own DAC. I'm pretty sure that this DAC will never win any measurement contests, but what counts is that it has a type of distortion that never stands in the way of my music. Will, it work for you? It might and if worst happens then just sell it for the same cash you bought it.
UPDATE
After some surgical intervention that involves completely amputating the MOSFET output buffer I finally may have tapped into what this DAC has to offer. Also I realized how foolish it was of Parasound engineers to use the previous buffering solution. After installing the OPA134 with a BUF634 in the feedback loop the DAC came alive with improved soundstage, more controlled bass (the bass with the MOSFET buffer was pretty flabby and exaggerated). The next mod most likely will be modifying the digital filter.
