LME49860 ??

Shop deals on ebay
Advertise with us
shop deals on amazon
GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Bauman

Bauman

Well-known member
Joined
Jun 4, 2004
Messages
795
Location
BRASIL
Just got an email from National regarding their new audio stuff

one of them
http://www.national.com/pf/LM/LME49860.html


Audio Op Amps
Achieve highest fidelity audio with 0.00003% THD+N, 2.7nV/rt Hz noise 44V dual LME49860, the award-winning 34V dual LM4562, and the 34V single/dual/quad LME49710/ 20/ 40 audio op amps.

High-Voltage Audio Output-Stage Drivers
Simplify your power amplifier design and increase power density with the monolithic 200V LME49810 mono driver with Baker clamp and the 200V LM4702 integrated stereo driver.

looks interesting...

:guinness:
Fabio
 
This part really looks good. But take a look at the data sheet
where they plot distortion VS level. Look what happens to the
distortion at the lower levels. Say 100mV. I'm not sure this is the magic
bullet we've all been looking for. For me, I may try to compare it
to the 5534/5532 in an actual listening environment.
 
See also the recent thread that linked to the Doug Self website, which showed the LM4562 to have some common-mode distortion issues (nothing terrible, just not something they mention when they tout the 300 ppb numbers).

EDIT: Here: http://www.groupdiy.com/index.php?topic=23557
 
Well it's THD+noise. Kinda hard to say but you would think the noise
would not go up if the signal went down if the gain of the amp was constant.
 
actually it is THD + noise to signal RATIO. As the signal level goes down and the noise stays constant, of course the THD+N to signal ratio is going to increase in the direct linear fashion as shown on the graphs.
 
[quote author="squib"]actually it is THD + noise to signal RATIO. As the signal level goes down and the noise stays constant, of course the THD+N to signal ratio is going to increase in the direct linear fashion as shown on the graphs.[/quote]

Yes---if the amp had significant crossover distortion somewhere, then you could see the increase of percentage distortion at low levels that was not due to noise. But one can presume that National has gone to great lengths to avoid such mechanisms.

I'll bet they got teased nearly to death about releasing the 4562 with a no-greater-than +/- 17V rails specification.
 
Great stuff guys, I am taking a better look in the datasheet, I just sent the links a minute after I got their email :)

I asked for same samples too cos in the end what I like it to hear them...

Lets see if they can break my absolute favorites opa627/opa637 ;)

Fabio
 
You might be able to get some insight into an op-amp's comparable shortcomings by configuring it as a simple inverter and then monitoring it's (-) input. In a perfect opamp there would be zero voltage there. In an ideal opamp there would be a tiny replica of the output signal (perhaps -100dB). In a real opamp this error signal will contain noise, distortion, be phase shifted, and vary as gain changes with frequency.

Of course you need to buffer this point with a high quality, high input impedance, high gain stage.

I suspect you will get plenty of noise here, and any signal at this point is subject to much interpretation, it might be interesting to stress an opamp with difficult but real music stimulus and compare the error terms between different devices. The nastier the error term, the nastier the transfer function, and noise will sound different than distortion.

Flat clean signal components in this error term should be innocuous so if you want to get fancy you could try to null this out, coming up with a 90' phase shifted version of the input signal may be a bit more difficult. Perhaps why this test is just a bench curiosity and not popular.

JR
 
One of the techniques recommended for looking at settling time in inverters is to set up a virtual summing node with values trimmed to equal the input and feedback R's (and other components if present). Then you can clamp that node with some low-C back-to-back diodes and not overload the amp you use to look ta it with too much. Pease and some ADI guys have written at times past on the subject.

The idea of looking at a similar node and comparing it to an appropriately phase-shited replica signal and not just a step function, is interesting.
 
I guess you could create a virtual null with a pair of resistors, to eliminate the possibly of changing the result by hanging an impedance off the (-) input.

Perhaps simpler yet feed the input and inverted output to simple summing amp. These two legs could be trimmed to correct for in phase gain error due to loop gain margin, so all that's left is phase shift, distortion, and noise.

I suspect the noise and phase shift will swamp the distortion for these new guys, and probably the old guys too... otherwise this would be the simples THD analyzer on the planet. Again I'm only considering this for comparing opamp "A" to opamp "B" with real music stresses... My suspicion is that real world distortion will be down in the dirt, and phase shift which is ignored by THD analyzers, when put in quantitative perspective is larger than we realize.

JR
 

Similar threads

E
META - headphone amplification
Replies
4
Views
17K
gyraf
gyraf

Latest posts

Back
Top