Sonic Frontiers SFD-1mk2 Vintage DAC for $750 That Stomps Most Modern DACs

‘yeah, it looked white to me. Your pic also looks silver not gold, but that’s not important. It says UA20400A on both those pics. The IV you are describing has always been op-amps on the SFD models regardless of the designation, can pull the original specs and it indicates this.

 

I wouldn't say your thinking is off per se but I would caution against letting the "I must have the bestest version of this thing" nervosa take over. I've been there too many times and it's rarely productive. I didn't know my SFD-2 has a different chip until after I bought it* and I'm glad for the same reason. The comparison against the SFD-1 is what sold me and if it's possible for you to do the same comparison I'd encourage it and if you're happy stop there. The concern for "but this chip could be .5% better" is an endless pit, one of the reasons tubes are such a pain in the ass.

I'm not picking on you, I'm encouraging you to step away from a deep hole. A hole I've been down too many times.


* I am in no way suggesting the seller was misleading, he didn't realize it either until looking closer at his receipt from PartsConexion.

 

I do recall that, years ago around early 2000s, when the D20400 chips started going bad, PCX had replacements for SFD owners. I believe PCX followed the UA DAC module architecture but used PCM1704 chips (like the ones also used on the 24/96 SFD2 Mk3).

Later, there were a couple of cheaper knock offs from other sources.

More recently, a guy on the DIY Audio forum developed a better replacement module based on the same UA architecture but using AD1862 chips. There is a long thread on that forum if you search. Eventually, Chris at PCX adopted/updated them and that is what they sell now as replacements.

Someone did do a comparison between the OEM and AD1862-based replacements.

https://www.threshold-lovers.com/2022/07/02/parasound-dac-200-ultra/

 

I would, of course, confirm with Chris or Glenn at PCX but I believe these modules had the PCM1704.

 

Goddamnit, I knew they weren't UA. I've emailed them/PCX 3 times already, I'm not gonna write them again and be "that guy". I'll take your word for it. Thank you for resolving this! It's OG, vintage modules for me.

 

Interesting, thanks for the link! He seems to like the replacement board. I also saw on diyAudio that guy, Antonio first showcasing his AD module, but I didn't make the connection to PCX's version, which is p much identical (I guess you now have your choice of opamps w PCX tho).

I also saw the empty PCM63 module he was talking about at the top, and I was about to buy it, until someone wiser told me that it was indeed empty of actual DAC chips, and that I'd have to source my own PCM63s lol. Would only that they actually had a ready-to-go module with dual 63s...

Def not buying that SFD-2 with 1704s tho. On goes the search.

 

To each, his/her own, but if it were me, I'd get the SFD2 and have PCX put in the AD1862 DAC modules and enjoy.

Plus, the DAC is only part of the equation. IMHO, the output section (and tubes in this case) are just as important.

 

For what it's worth, a few years ago I put together an AD1862 dac using a board I bought from diyinhk based on a suggestion from Hands. I added a couple of their low noise power supplies, one for the digital side and one for the analog/IV circuit. Actually I can't say I built it because I screwed everything to a scrap piece of plywood just to hear how it worked. Then I rolled a few opamps that are the IV and output stage. I compared it to my SFD-1 and I would have to say it was pretty close. Minor differences in voicing, but that's to be expected.

I guess what I am saying is that the PCX AD module may very well sound good. After all, they understand how to design it as well as the guy in Hong Kong (although I have been impressed by diyinhk products multiple times. He's good). And there is nothing wrong with the AD chips.

Regarding PCX setting up for production, I expect it was a small batch build. These days it's pretty easy to get circuit boards built in small numbers for very reasonable cost. From there you can build it by hand.

I should say that I needed to add a couple small transformers, diode bridges, and power supply caps to make everything work. The interesting thing is that you can still build this. Diyinhk still sells the board with USB and i2s input and it costs $170 if you have him supply the AD1862 chips. The power supply boards are $60 each or $120 for 2. I was able to feed it with i2s from the buffered outputs of my Pi2AES and I used OP!1656 opamps for the IV/output stage, but you can roll to taste. I never did get around to finishing the project, but it's a little over $300 without a case, so not bad. Looking again, it seems he is currently out of the +/- 12V power supply regulators but I suspect he will get them again.

https://www.diyinhk.com/shop/audio-...fo-reclock.html#/37-dac-dac_pcb_with_dac_chip

https://www.diyinhk.com/shop/audio-...e-dac-power-supply-regulator-33v55v-1ax4.html

https://www.diyinhk.com/shop/audio-...oise-dac-power-supply-regulator-1215v-1a.html

Jac

 

Anyone know what the possible upgrades are for the Processor 3? I know I could email PCX, trying to not do that at the moment as they are working on my SFD-2.

Thanks

 

Series signal resistors
Interstage coupling caps
PS caps
PS diodes
Tube sockets
DAC board caps
PS regulators on DAC board
New wiring
New RCAs, XLRs, fuses, tubes, and DAC modules (as needed).

 

Thanks, that reads similar to the upgrades done on my SFD-2

 

Correct. They all have a family resemblance internally.

 

So I finally found some time and butchered the relay caps, then soldered new caps (C501=33uF, C502=47uF) to the leads of the old ones sticking out of the board. I'm pretty sure I got decent connections by checking for continuity.

I'm now getting an off/on/off/on cycling every 5 seconds (I think someone may have mentioned similar behaviour earlier?).

• power on > 5 sec > loud click, music
• 5 sec > soft click > silence
• 5 sec > loud click > music
• etc.

Any suggestions on what to try next? At least it sounds good for the 5 seconds that it's on ¯\_(ツ)_/¯

 

Have you tried replacing the relays themselves?

 

No, I was instructed to replace the caps first, as that's what's most likely. I'm not up for pulling the PCB out, so that's not an option for me to try at home. If it's that, I'll be sending it to PCX.

Forgot to mention. Voltage across 33uF cap is a constant 4.96V, Voltage across 47uF cap starts around 2V and gradually drops when I put a voltmeter across it.

Attached is the schematic provided by PCX.

 

I suspected the new 47uF cap at C502 was wonky, so I replaced it with another one. Back to total silence. I don't think either of them were actually dead as there was still some oil in them when I cut them off.

Glenn and Chris at PCX got back to me (as well as @dBel84), and I tried a few other things like checking some voltages, resistors and diodes, and then reflowing the solder on all the joints in the mute relay circuit. After the reflow, it would output sound for a few seconds after power up (and the standard mute period) and then drop out again. I guess the opamp at U7 crapped out and isn't able to provide enough voltage to charge C502 and reflowing got rid of a bit of resistance so it could charge it enough initially after power up.

Anyway, I asked PCX how to bypass the mute circuit and it's a simple short from base to emitter of one of the transistors. It's working now, and all I have to do is remember to power up the DAC before the amps and power them down before powering down the DAC to avoid any spikey nasties. I've put a piece of tape with "AMP?" written on it above the power switch.

When we move (back) to Ontario, I'll send it in to PCX and have them fix it properly. Packing this beast (and paying) for shipping back and forth across the Pacific is just too much. At least I can still listen to it (for now).

 

I'm glad it worked out and the guys at PCX were so helpful. 30 years ago, when they designed the SFD series, I bet they never would have so many in service and never guessed that the circuit that would be giving the most problems would be the mute circuit.

Being a human soft start isn't really a problem. I already do that with the active crossover and amp in my system. It just takes a little more care first thing in the morning when I'm still half asleep.

Congratulations on getting it working.

 

Heard back from Chris at PCX, my SFD-2 mkii issue is a failed AES-21 module which they currently do not have any service parts to replace it. He mentioned they may have some being made but that is a few months away, an alternate suggestion was to look for an Ultra Jitterbug or Assemblage UJB-1 to harvest the AES-21 module. Unfortunately after a brief search have not come across any used.

Luckily I picked up the Processor 3 to hold me over until this gets sorted out.

 

pls do a brief (or not) shootout when you have the time. V interested.

 

I will try my best from memory, would rather have both in house before making a proper comparison but that could be a while given the mercy of waiting on the busted module.