E-mu 1616m soundcard
ImageShack® - Online Photo and Video Hosting
I'm not sure if the opamp is use as LPF only since the cs4398 dac has an opamp inside and is spitting voltage..
I plan on removing the caps in the path and i wonder about the jrc2068.. LME49720, LME49723 and LME49725 seem to be interesting.. or AD8066?? what is you opinion
this soundcard and internal dac rock.. i wish somebody would just take it, mod it as hell to bring it to the level it need to be..
I was looking also at clocks to replace those in the box.. but dam.. there is 4 clock on this product.. two on the pci card and two in the dock.. i wonder what are the clock i should replace..
I want to buy two crystek CCHD-957 to replace those but i wonder if I need to buy 4.. kinda crazy if only two are clocking the dac.. i really wonder why 4 clock anybody ??
or any other clock you recommend instead of crystek CCHD-957?
ImageShack® - Online Photo and Video Hosting
I'm not sure if the opamp is use as LPF only since the cs4398 dac has an opamp inside and is spitting voltage..
I plan on removing the caps in the path and i wonder about the jrc2068.. LME49720, LME49723 and LME49725 seem to be interesting.. or AD8066?? what is you opinion
this soundcard and internal dac rock.. i wish somebody would just take it, mod it as hell to bring it to the level it need to be..
I was looking also at clocks to replace those in the box.. but dam.. there is 4 clock on this product.. two on the pci card and two in the dock.. i wonder what are the clock i should replace..
I want to buy two crystek CCHD-957 to replace those but i wonder if I need to buy 4.. kinda crazy if only two are clocking the dac.. i really wonder why 4 clock anybody ??
or any other clock you recommend instead of crystek CCHD-957?
Last edited:
Ah in that case you want an IC which has good tolerance for RF - try LM6172 but first check the supply voltages are lower than +/-18V. It won't be so low noise as 2068 but might well sound better due to lower IMD. I have no experience of those other opamps you mention but as a guideline, look for ones with high slew rate where the environment has a lot of RF.
I would replace it with another low noise bipolar opamp if you must. I think LME49720 is actually noisier than JRC2068 can't remember exactly off the top of my head, but it should be a good choice overall.
The EMUs only have 0.1uF caps for local decoupling on the op-amps, so it's likely that anything fast or exotic will oscillate (ADA4898-2 was unstable in an 1820m for me).
LME49720 and OPA1612 worked for me in an 1820m but I did not get around to verifying with a fast oscilloscope yet, but no obvious signs of instability. I suspect the ADA4898 was unstable because I saw abnormal current draw and heat.
I would not be surprised if LM6172 is unstable on that board.
The EMUs only have 0.1uF caps for local decoupling on the op-amps, so it's likely that anything fast or exotic will oscillate (ADA4898-2 was unstable in an 1820m for me).
LME49720 and OPA1612 worked for me in an 1820m but I did not get around to verifying with a fast oscilloscope yet, but no obvious signs of instability. I suspect the ADA4898 was unstable because I saw abnormal current draw and heat.
I would not be surprised if LM6172 is unstable on that board.
My experience vis-a-vis sound quality is that the lower the noise the (bipolar) opamp, the worse it sounds in adverse environments (RF ingress). I put this down to the LTP input stage - just a hypothesis at this stage, no firm data. Hence my suggestion of LM6172 as that topology dispenses with the LTP.
Its true enough that LM6172 is not the easiest opamp in the world to get stable so its worth sucking it to see.
Its true enough that LM6172 is not the easiest opamp in the world to get stable so its worth sucking it to see.
anybody think the same?
what happens when an opamp oscillate?? how can i know by ear? do i need an oscilloscope?
Sometimes an oscillating opamp can be told by ear, sometimes even with the oscillations the SQ is improved over the previous device, it all depends. A scope is the only reliable way to tell. Rolling opamps is always somewhat liable to produce an oscillating device, its just a hazard to take in one's stride 
The alternative is keep the opamp and fix up the environment its in. Which means paying attention to RF filtering on the power supplies and inputs and outputs. I get decent sound from NE5532s this way but its a lot more work than just swapping out the opamp, it means reverse engineering the circuit, then patching it.
thx both for your help.. then i'll try LME49720 and find somebody with knowledge to fix RF filtering on the power supplies and inputs and outputs..
i'll change also the crystal..
also someone said that i could decouple the voltage input rails of the CS4398 DAC chip.. not sure what decouple mean hehe
i'll change also the crystal..
also someone said that i could decouple the voltage input rails of the CS4398 DAC chip.. not sure what decouple mean hehe
'Decouple' here might mean filter - put some series impedance in the supplies and then provide extra capacitance to ground directly at the DAC chip. Its one way to filter out some RF, the series impedance can be ferrite beads or inductors or both, or simply a resistor if there's enough supply voltage headroom available and you're not confident enough in selecting appropriate magnetics
Last edited:
If 1616m is the same as 1820m, the op-amps already have an RC filter in both the supply lines, at least they do in the headphone section of the 1820m. Adding a 10uF tantalum or electrolytic across the 0.1uF ceramic might do enough to make some picky amps stable.
The CS4398 DAC already has several decoupling caps, but they are crappy Chinese electrolytics and 0.1 X7R ceramics in the exact CS4398 datasheet configuration (pretty sure). You could replace the Jamicon / Teapo crap with some decent (Panasonic, Nichicon, etc) caps that will at least probably last longer.
The CS4398 DAC already has several decoupling caps, but they are crappy Chinese electrolytics and 0.1 X7R ceramics in the exact CS4398 datasheet configuration (pretty sure). You could replace the Jamicon / Teapo crap with some decent (Panasonic, Nichicon, etc) caps that will at least probably last longer.
And you need a fastish analogue dual channel scope. Dual trace even better but these are relatively rare and much more expensive.
Aim for 100MHz and 2mV/div if you can find one at a secondhand price you can afford.
Dear audiolic, pls try OPA1612 from TI/BB. it's a straight swop for the double 'over bass with bad top' BJT input 2068. I found it amazing having 1st tryied the AD8599. The AD was a definate improvement, but the BB made all a not more natural. I don't personally like the National chip series much, since I find them a bit musically 'lifeless' - very accurate and clean, but somewhat 'to good' - bit like some Luxurious Sino cars - probably 2 perfect... Matter of taste as well - indeed...
All the best of luck
DocO
All the best of luck
DocO