Experience with this DIY DAC ?

RichLund, I'm crossing my fingers that the problem lies in the analog section and not the digital! Ian Ma did say each unit is tested before it's sent out.

pchw, let me know if the different opamps cure the crackling. I'm leaning toward ordering some spare opamps, what the heck, they're cheaper than the Edcors...
 
I was skeptical about swapping out the opamp. However, it seems to be at the right direction. I put in a pair of opa2134 that I original bought to replace the, cough cough, NE5532's in my Adcom GDA-700 DAC.

I replayed the material that I heard the crackling clearly before. Now, I don't hear the same distortion unless I pay extra attention, then occasionally I can hear little once for a while. This artifacts might be just bad recording that I didn't hear before. It is definitely a very drastic improvement. 20 minutes has passed, so far so good!! lbj, give swapping opamp a try, that can be the ticket to ride.
 
Well my DAC is running with the upsampling board now. i noticed that also the board for the upsampling function is too high to fit in my case. i will have to change a lot of things to make the dac fit in my case.. 🙁
the 35v/3300µF caps are also too high. i think i will unsolder them and lay them down on the pcb. for the upsampling board i will use a ribbon cable, that should work.. maybe xD
 
Another 3 hours of play time has gone. The Big DAC with opa2134 is quite promising. The crackling is mostly inaudible. I suspect it is something in the analog circuit causing it. Somehow the NE5532 is more sensitive to this "something".

Anyhow, the NE5532's are junk. lbj, if you are going to buy a new pair, I suggest to go for a pair of LM4562 per other's suggestion. I bought my LM4562 from ebay for a little over $10 shipped. He is in NY, and shipping was reasonably fast. That's one option for you.
 
HI sorry to jump in. I had the same problem with mine. I think its the socket for the opamps. try applying some pressure to the opamps and se if it goes. That fixed it for me. I did resolder the sockets in but didn't fix it only applying a bit of pressure did. might change the sockets over completely on mine.

Hope that helps

G
 
Anyhow, the NE5532's are junk. lbj, if you are going to buy a new pair, I suggest to go for a pair of LM4562 per other's suggestion. I bought my LM4562 from ebay for a little over $10 shipped. He is in NY, and shipping was reasonably fast. That's one option for you.
They cost about $4.50 from Digikey or you can get the, supposedly better, "can" version for about $15. The first mod I did with this board(CS4397 version), before jumping on the trafo bandwagon, was to bypass one of the opamps and replace the other with an LM4562. It sounded great. I then tried some "Burson type" discreet opamps but kept coming back to the LM4562 which I found to be at least equal if not better.
An externally hosted image should be here but it was not working when we last tested it.
 
Last edited:
Went ahead and ordered a set of LME49860 last night from Digikey. It's supposedly either a relabeled LM4562, or slightly better than the LM4562, depending on who you listen to. Anyway it was cheap and I'm really only checking if the stock opamps are faulty; the eventual output stage will involve transformers. Graemedench, thanks for chiming in. I think the NE5532s are fully seated but I'll doublecheck.
 
Graemedench,
I did clean, bend the pins, and reseat, no dice.

McCrackers,
You are right, I could buy from Digikey, too. It was one of the cases that I was watching something else on ebay and came across the item - impulse buying which always waste me a few $$ once for awhile. As a matter of fact, I actually have had one LM4562 all along. It was inside the same tube that holds the 4 opa2134's. I didn't remember there was one until I did the opamp swap last night!! Now, you can guess I am no youngster, it must be the beginning of my senior moment 🙂 Guess what, there is one LM6172IN, 2 x AD797, and 13 x opa627ap in the same tube, and I only remembered the opa2134's!! Talking about embarrassment 😱

Eventually I will ditch the second opamp at minimum and may go transformer if the Edcors that I already own (or any appears in my impulse buying path 😉) will work to my liking. In short, all the opamp will probably be back to the tube until 2020, LOL.
 
oh, didnt check that, the pcb is already mounted on some distance bolts.

do i have to unsolder the parts at the R and L pins?

and do i need an earthing wire connected to the 0V connector between the 13v connectors? because the transformator for the opamp section is not required anymore
 
Last edited:
do i have to unsolder the parts at the R and L pins?
I would take out the de/coupling caps and take your feed from there. It will become more obvious when you see under.
This way nothing else will interfere with the output from the dac.

and do i need an earthing wire connected to the 0V connector between the 13v connectors? because the transformator for the opamp section is not required anymore

Not to the best of my knoledge
 
Anyone notice how high the output volume is when using these boards stock or with OPA627s?

Volume is a fair bit higher when comparing coax into this dac against rcas directly from the CD player I'm using as a transport. SO much higher that on some CDs I get a fair bit of distortion when things get loud. Should this be happening or does it suggest there are other problems?

Another reason to try out the output transistors I suppose...
 
Seems like the latest version has some problems, audible distortion just should not exist regardless of opamps used. I didn't even try the stock output circuitry out when I bought mine, went straight to output trafos. I have ver.5

I have a pair of Tamura trafos out of a Neve console that someone might be able to talk me out of.

Bill
 
If the distortion is not present with output transformers, it would suggest that there is a problem with the output stage design.

Has anyone tried attenuated interconnects to see if its just the output thats too high or if the overload is within the output stage itself.

Can someone post a diagram of the output stage. I'm pretty sure changing a few component values will sort the problem for those who don't yet want to go for a fully redesigned output stage or traffos!