The Opamp thread - Page 272

Both are very different chips and I can live with both

I can live with the ADA4627-1brz not the arz, the arz are a little unstable for my liking.  The ADA4627-1 brz I had in a dac and a pre and I lived with it for about 9mths, a chip that as far as Im concern is hard to beat.  Very expensive  I prefer them over the OPA 627 at the price point.   The LME are value compared to them!

 

Im going to try the LME 49860 AND THE 49880 hopefully soon

 

I like your thinking - makes sense to me.  Ron (HiFlight) had convinced me that duals don't suffer any loss of SQ relative to singles, but for op-amps where dual versions don't exist (or are unavailable), the "Dual to Mono" idea seems appealing.

 

Thanks,

 

Mike

Interesting so the Lite DAC-AH is "improved" by removing the op-amps in place for capacitors, changing it to passive I/V - http://www.coemaudio.com.au/Sources/DACs/183/

 

This guy says that sounds really "weak" though - http://lampizator.eu/LAMPIZATOR/REFERENCES/AH%20DAC/AHDAC.html

 

Still looking for DAC's for op-amp evaluation...

Edited by kiteki - 7/5/12 at 2:27am

I would love it if you found one, would make it fun to play with opamps, because you could have the dac its in own compensated circuit, and just swap out opamp circuits (opamps + their own compensation on a separate circuit.)

 

But different chips are made for different things, like the AD797 (Sigma-delta buffer) so you won't really be able to use them all anyway, but the concept is still fun.

Well my AK4396 DAC has exactly 1 op-amp socket and I can listen directly from the line-out, which is what I've done, it's when I connect my amplifier I start worrying about the op-amp in my DAC a little, so a current-out DAC chip into passive I/V into the buffer in my amplifier into [insert op-amp] seems like a more ideal scenario.

 

It seems like there is no consensus on passive I/V or using an op-amp at least - http://www.head-fi.org/t/512389/mini-dac-tda1543-x-4-nos/135#post_7233717

nothing happens; there are roughly 50 nano seconds in each sample at 20khz, which is already above many peoples hearing threshold

 

passive IV with transformers on the above dac is very nonlinear, most people like the 2nd harmonic distortion. passive IV with resistors only works really well if the dac chip has decent output compliance. EVERY design you can come up with presents a compromise of some sort, just pick one and keep it static, while experimenting with the rest. this quest to find something that doesnt have any effect on following stages ie completely colourless is not realistic, particularly as a rather basic level hobbiest, no offense intended.

 

there are no truly modern current out dacs that have high enough current output to drive a resistor by itself and give best performance. the pcm1704 is about as close as you'll get, but a truly reference 1704 build is a pretty big task and not cheap. most passive IV circuits are famous for being 'musical' 'natural' 'unfatiguing' which are all in themselves different colourations

Edited by qusp - 7/5/12 at 6:07am

ADxxxxARZ unstable vs BRZ? haha you can back this up? I recommend you drop the guys at AD an email, I think they would be very interested in your findings

 

I didn't mean exactly 1 nanosecond...

 

Anyway I don't believe in thresholds since they change all the time and usually come from small university studies, with a select number of people, there's plenty of incorrect thresholds on Wikipedia.  Afterall, hasn't the SanDisk Clip+ surpassed all of them in audio?  At least according to the multi-thousand dollar dScope reviews of it, so by that logic nothing in your AckoDac is audible either. ^^

 

Maybe he meant to say "colored". It certainly sounds more colored without extra bypass caps vs. BRZ, but both grades are hard for us to stabilize. We've got overheating 797s and overheating 4627s, lol.

I recently read in the 6moons review of the Meier Concerto that Jan Meier described the AD797s (which were used in both his Concerto and the current Classic model) as being particularly sensitive to power supply quality:  

 

 

This gave me revisit my use of AD797s in the buffer stage of my PB2 (where I had previously reported extreme overheating during an evening's use, but have not been able to reproduce the overheating since then).  

 

I had a vague recollection that I may have been using my Energizer XP8000 external lithium-ion battery pack to power the PB2 at 16VDC, while the XP8000 was being recharged - which normally isn't the case.

 

Sure enough, the AD797s started overheating when the PB2 was powered by an XP8000 battery pack that was itself being recharged via its switching (non-linear) AC adapter.  

 

When I unplugged the XP8000 from its charger, the AD797s went back to normal operating temperatures (they stopped oscillating).

 

Mystery solved.  The XP8000's charger can cause the AD797s to oscillate.  

 

Mike

ADA 4627-1 brz is a higher grade of the arz, I have found that most ADI chips require quality PSU and strick correct bypassing to work.  The ARZ in the same circuit has higher audiable bottom end distortion. You will not notice it if you used the arz version first, it isn't until you used the brz for a few days and then place the arz in place without alter anything.

For some unknown reason the ADI chips run hot accept for the ADA4610-2.  But the ADA 4610-2 is low resolution.

 

I now have a CRO to show that the ADA 4627-1 has low level oscillation even with documented bypassing, this is possibly due to landscape that I am using and not the recommended bypassing.  I even think that using thru-the-hole components are useless to fully eliminate oscillation and noise and SMD is the next step.

 

I now returned to LME devices like the LME 49720 as this chip is stable and  suited to the circuit I am using.

Edited by ecohifi - 7/5/12 at 2:03pm

 

To be blunt...

 

"Not another nosser", "passive I/V sux", "just keep whatever DAC static" and "try harder next time".  Let's say you are hypothetically correct about current-out and passive I/V, I still need to hear the concept first like any real scientist.  I'm asking for advice on finished DAC's, sooo... the skill level is only on the designer, unless you mean my level of consumer savvy.

 

Are using JLH ripple eaters between lithium-ion battery pack and amp are a way for you to solve this problem?

Or a lowpass filter with choke coils and caps?

 

 

http://www.diyaudio.com/forums/group-buys/167286-john-linsley-hood-ripple-eater-3.html

http://rockgrotto.proboards.com/index.cgi?board=m&action=display&thread=5645

http://www.gt6.co.uk/JLHPSRR/John%20Linsley%20Hood%20ripple%20Eater%20-%20Rock%20Grotto%20Charity%20Edition.html

http://www.audiofuture.net/showthread.php?95-JLH-Ripple-Eaters

To stabilize LME49990 in my DAC, should I add the RLC compensation thing that the datashet prescribes? Also, it is driving a MOSFET buffer, so does that have high enough impedance at the buffer's inputs? But wouldn't adding that compensation color the sound due to the inductor and cap? Or does the inductor and/or resistor alleviate that?