Looking at the web page for LM4562 this morning I noticed that TI is marketing a device called OPA2891 as an alternative device.
OPA891 is a single, while 2891 is dual op-amp.
I have attached the datasheet, more info can be found here https://www.ti.com/product/OPA891
Looking at the datasheet is is dated November-23 and the devices seem to have similar "simplified schematic, noise and current spec compared to the AD797.
The single opa891 version is available for purchase at least from TI and Digikey.
Anyone tried this device or know where I could order a couple of the dual version. I'd love to try it in my upcoming Pearl-3 build
OPA891 is a single, while 2891 is dual op-amp.
I have attached the datasheet, more info can be found here https://www.ti.com/product/OPA891
Looking at the datasheet is is dated November-23 and the devices seem to have similar "simplified schematic, noise and current spec compared to the AD797.
The single opa891 version is available for purchase at least from TI and Digikey.
Anyone tried this device or know where I could order a couple of the dual version. I'd love to try it in my upcoming Pearl-3 build
I'm waiting for samples at the moment. This new series is really exciting, notably as there are de-compensated versions (OPAx892) and the singles have offset nulling pins.
Current noise and its 1/f corner are on the very high side, though, as is bias current, but that is what one would expect from such a design. Best used in a very low impedance environment.
As a replacement/upgrade for other OpAmps, you really have to know what you are doing here, this thing is damn fast and will likely be unhappy in any design/layout that isn't true RF-style.
Current noise and its 1/f corner are on the very high side, though, as is bias current, but that is what one would expect from such a design. Best used in a very low impedance environment.
As a replacement/upgrade for other OpAmps, you really have to know what you are doing here, this thing is damn fast and will likely be unhappy in any design/layout that isn't true RF-style.
That input bias current will catch some out, 33µA max at temperature (and it runs hot with 11mA per amp (worst case at temperature)). Aimed at ultrasound and audio like the AD797, but falling just short of its noise performance.
A nice bandwidth, but not everything is ideal as noted above.
I find it disappointing the the distortion (THD+N) increases above an output of around 2 Vrms. That didn't happen with the good old LME49990 and also less pronounced with e.g., the OPA1612.
I find it disappointing the the distortion (THD+N) increases above an output of around 2 Vrms. That didn't happen with the good old LME49990 and also less pronounced with e.g., the OPA1612.
Its surprizing the output current limit is as high as 66mA when it has difficulty driving 2k loads...
This looks like a CFA on the face of it but the block diagram on page 18 is very much a folded cascode VFA device. The input bias current is very high (can't remember what the AD797 was off the top of my head). For low-noise bipolar input stages, designers have to run the input long tail pair at high currents, and this is reflected in the high supply current. I suspect that the short circuit current is 66mA because I see that very low-value feedback resistors are used in the application diagrams so the opamp has to drive this (50mA peak for a total feedback network value of 300 ohms at 15V peak for example). That doesn't leave much current left over for driving the 2k load. Given the very wide bandwidth and high slew rates, any capacitance from the feedback node to ground, or across the feedback resistors, is going to result in peaky square wave response, which is what you see in the frequency plot graphs.
I would not consider this part for audio or as a replacement for the AD797. It might work as an MC input amp, but then the high bias currents and 0.2-1mV offset voltages mean you would have to AC couple the input or servo the offset out. I do not understand why you would need these slew rates and bandwidths for audio.
The OPA1611 and 1612 have a tad higher input noise and input bias currents of 250nA max and device supply current of 4.5mA per channel. I used it in an opamp-based MC amplifier and the performance was quite good all things considered.
I would not consider this part for audio or as a replacement for the AD797. It might work as an MC input amp, but then the high bias currents and 0.2-1mV offset voltages mean you would have to AC couple the input or servo the offset out. I do not understand why you would need these slew rates and bandwidths for audio.
The OPA1611 and 1612 have a tad higher input noise and input bias currents of 250nA max and device supply current of 4.5mA per channel. I used it in an opamp-based MC amplifier and the performance was quite good all things considered.
Last edited:
Just stumbled over it. It has some interesting behaviours - like THD is mainly H2! And it's linear over the audio band. Both is very nice to see and not common for an audio OPA. Over all it reminds me at the good old AD797 - but with huge output current!
Anyone gave it a try? A mic pre with some servo and transformer output could be nice to do with it ...
Anyone gave it a try? A mic pre with some servo and transformer output could be nice to do with it ...
according the internal simplified schematic on page 19 under
https://www.ti.com/lit/ds/symlink/opa891.pdf
it is only one gain stage inside of the NFB loop (a folded cascode + buffer stage - I like the sonic character of this topology).
Just like the AD825 and the AD797 - go to
In the datasheet of LM4562 I don't find a simplified schematic - go to
https://www.ti.com/lit/ds/symlink/lm4562.pdf
https://www.ti.com/lit/ds/symlink/opa891.pdf
it is only one gain stage inside of the NFB loop (a folded cascode + buffer stage - I like the sonic character of this topology).
Just like the AD825 and the AD797 - go to
In the datasheet of LM4562 I don't find a simplified schematic - go to
https://www.ti.com/lit/ds/symlink/lm4562.pdf