question about ART digital MPA

antichef

pornk rock
like others on the board, I'm the proud owner of an ART Digital MPA (the gold colored one). Some discussion about these came up on the thread regarding the group buy preamps (which I'm definitely in line for, btw)
Well, let me tell you something about your cheapo MPA. I've had mine for over two years now. It has a five year warranty. Last week I emailed ART because my input gain pots are scratchy. Cleaning hadn't helped. I told them I didn't want to take it out of service, so could they please send me replacement pots, and the VU meter lamps as well. They answered me same day, and the parts are on the way.
so... while we're on the subject...
I love the way mine sounds, and I have found that the quietest and bestest way to get the signal to my DAW is via the digital out (via a MOTU Traveler), but I've had to wrestle to get there. I get a result I'm happy with if I 1) use the coax (not optical) s/pdif out with a nice 75ohm cable, with 2) an external clock (the Black Lion Audio Microclock, in my case), terminating the "clock through" output at ART's direction, 3) at the "2x" sample rates, that is, 88.2khz or 96khz. *All* other approaches I've tried (including all combinations of optical cables, other sampling rates, slaving to the ART clock, slaving to my MOTU traveler clock, AES/EBU, ADAT) result in, at best, clicks and pops, at a worst, big static noise, or no sound.

No big deal, really, since I'm very happy with the the arrangement that works, but recording at 44 or 48 khz with the same conversion/clocking quality would be nice, especially for recording interviews and stuff.

Last night, I opened the box, and stared into the abyss (with the help of some lights in the room), and further research here yielded this:

Mine has OPA2134 *but* a disclaimer, I have the DMPA, and an early model, it also has a top range AKM converter chip, but later versions use a cheaper chip. So I won't swear that the opamp didn't change, although OPA2134 is a lot cheaper than the AKM chip (which was $20), so the cost savings wouldn't be as great. The opamps are socketed too.
Mine is newer than that. I do have the OPA2134 opamps (and still more research has convinced me to leave them be), but I don't see anything on what's obviously the converter board to make me think it has the AKM converter chip. That is, I don't see a chip with "AKM" on it (I do see one with TI -- don't have the numbers with me right now). If I'm right, is this possibly the source of my struggles, and would it be something that's upgradeable for a DIYer?
 
Mine is newer than that. I do have the OPA2134 opamps (and still more research has convinced me to leave them be), but I don't see anything on what's obviously the converter board to make me think it has the AKM converter chip. That is, I don't see a chip with "AKM" on it (I do see one with TI -- don't have the numbers with me right now). If I'm right, is this possibly the source of my struggles, and would it be something that's upgradeable for a DIYer?

I use the coax output using 1/2 of a garden-variety 18" RCA audio cable into my VSL2020 card. No problems with that, ever. But I wouldn't guess your problem is with the converter chip itself. It just sounds like a typical clocking problem, somewhere between bits of your gear.

Swapping out the converter would be less than easy. I've gotten reasonably good at surface mount work, but a 28 pin chip would be a challenge. And I don't know if they used a pin-compatible replacement, or redesigned the board.

Here's a pic:
 
That's encouraging that it doesn't sound like a problem with the chip -- I suppose I'll find the fix eventually.

I looked on my unit where the arrow in your picture is pointing, and in approximately the same place, I see the TI PCM4202

http://focus.ti.com/docs/prod/folders/print/pcm4202.html

I guess that would be the replacement for the AKM chip?

Yep. PCM4222 would be closer in performance, but that's a 48 pin chip. I had a quick look at PCM4202 and AK5394A, they are the same size but don't have the exact same pinout. 5394A specs better, but I don't think 4202 should suck. I'd look for master/slave misassignments somewhere in your setup.

I tend to keep things simple and have everything slave to the ART. SPDIF out to card, ADAT up to my RME converter. If I'm fussy, on playback I'll set the RME to internal clock. If the ART's not on, and I'm lazy, I'll set the VSL2020 to master. I don't think I've ever bother to test for a difference. All the measurements I have taken on the RME, distortion is so low as to not worry about it.

But a clocking problem, that you have to worry about . . .
 
ART Digital MPA optical out to 2408 (MPA=major pain in .....)

Couldn't resist the title.

Anyway, I too had major problems with the optical-out, into ADAT-in on the original 2408. After much hair pulling, broke down and bought a word clock cable(BNC), connected 2408 word-out to MPA word-in -- set Sample Rate knob to SPDIF-WCLK/ optical button pushed in for ADAT -- and 2408 now clocks the MPA; optical out works perfectly. I am also using the 2408's spdif-in/ out as send return for my effects unit. Through Digital Performer, the effects unit is the clock source to sync the 2408.

I had originally expected the 2408 to sync to the ADAT optical, which it did, but there was so much garbage coming out at the same time burying the audio signal which itself was totally garbled -- so word-clock is the answer.

Cheers.
 
Back
Top