Thank you for the clarifying. Maybe this new fabrication process will come without noisy chips.
Welcome and we can certainly hope.
TI lists it as a cheaper package alternative to the metal can LME part.
The data sheets are identical except of course the package.
I wonder, if in part, the 49990 was axed because it would effect sales of TI other products.
Even going from 4" to 6" to 8" wafers in the same fab is painful. The same flow with different implanters, etc. needs tweaking everytime. One of my first tasks 40yr. ago was to figure out why the Micropower fab and our fab had different yields on the same masks/flows. Turned out the beta/VA product here was worse by a factor of 1.5 or so so the open loop transfer function of some current mirrors had more non-linearity.
On the LME49990 it's also possible it was too late to capture the non-audio business. I'm afraid the new design-ins of classic +-15V high performance op-amps is dwindling and folks reorder what they designed in in the first place since re-qual is VERY expensive.
Last edited:
I can see only this code on the chip. On the one older chip I see 39YC.
In the data sheet we can read this:
"There may be additional marking, which relates to the logo, the lot trace code information, or the environmental category on the device"
For some parts thie the LME parts there is a hot counterfeit market. Do be sure you are getting them from a fully qualified source. I have seen counterfeits turn up in places where it would never seem possible. The big players have made serious efforts to control the supply chain but things do happen. You should be able to check with a TI rep on the date codes for validity.
The LM4962 and LME 49720 are the same die. The product manager told me they created a whole team to make "premium" audio parts and got some traction from Sony. The LME was a specific designator for the family but the LM4962 was already shipping and qualified so it stayed in the line. It was actually sorted into I think 5 different products but all the same die.
As Scott said moving a chip is not easy and you need a good reason to do it. That part was not really going to be high volume (not enough on a chip to be interesting for things like sonic imaging).
Actually National started shutting the audio stuff down before TI bought them. The management though the future was elsewhere. Now they are all elsewhere.
The LM4962 and LME 49720 are the same die. The product manager told me they created a whole team to make "premium" audio parts and got some traction from Sony. The LME was a specific designator for the family but the LM4962 was already shipping and qualified so it stayed in the line. It was actually sorted into I think 5 different products but all the same die.
As Scott said moving a chip is not easy and you need a good reason to do it. That part was not really going to be high volume (not enough on a chip to be interesting for things like sonic imaging).
Actually National started shutting the audio stuff down before TI bought them. The management though the future was elsewhere. Now they are all elsewhere.
In some cases it does.
I tested some alternatives to the LME49990 for my Audio Analyzer. Among them OPA1611.
With the LME49990 i can get loop back distortion at 1 kHz below -130dB. Replacing one op-amp (output driver, in a combo with LME49600) gave me a loop-back distortion of -123.7dB. The input amp and the ADC driver still used the LME49990.
Some of the other alternatives performed slightly better than the OPA1611. The only one that was on the same level as the LME49990, and perhaps even slightly better, was the AD797 (once I got it stable).
I find it depends on the application. I use them in my SVO. The 49990 and 1611 do about the same.
You passed the test.
A current equivalent process would have 10X or higher ft's. Though I fly-wired a simple one op-amp, 1kHz oscillator to roll different amps and never had a problem with any being unstable. BTW has anyone measured the noise >100kHz on some of these amps? I published the 5534 spectrum rising, the 1028 does it too both due to the multi-stage design.
Last edited:
Scott- Does the AD797 use dielectric isolation or something like that? That's an expensive process I have read. Not trivial.
My Quan-tech got hijacked into Constellation production a few years ago and I have not been able to get it back yet. The Quan-tech is only good to 100 KHz. Would there be a value in making a test setup with a wider bandwidth? Big problem is the nV/rthz scaling of a general purpose spectrum analyzer (after getting past the noise floor).
My Quan-tech got hijacked into Constellation production a few years ago and I have not been able to get it back yet. The Quan-tech is only good to 100 KHz. Would there be a value in making a test setup with a wider bandwidth? Big problem is the nV/rthz scaling of a general purpose spectrum analyzer (after getting past the noise floor).