I came across the MIC920 Op Amp yesterday, a single high speed op amp package, it seems like the slew rate and bandwidth are all reasonable and at $0.48 a chip it seems like a really economic way to design for video. Can anyone think of a reason not to use this IC? I’m sure there must be some glaring reason I’ve not notice!
Datasheet - https://docs.rs-online.com/ce47/A700000008632983.pdf
Strange but the above link in Phosphenes post didn’t work for me the first time. After finding this one below, I then tried again & it worked. Well just in case I’m not the only one experiencing this issue, here’s another link to the datasheet.
Found the link here:
Oh, this might be helpful to find the LM6172 datasheet:
Nice find @phosphenes 
For a really simple circuit that has only 1 out 2 op-amps, this is definitely worth looking at but because it’s an SMT component and there’s the ADA4851-4 (4x op-amps on 1 IC) that’s a way cheaper alternative to the LM6172 & used by LZX, Syntonie and Foxing Hour on their pre-built (Non-DIY) offering.
The key problem that stands out to me regarding the MIC920 is its power requirement. It’s max input Voltage is 9V/±9V, & min V is 2.5V
So even using the ubiquitous 7805 & 7905 5V & minus 5V regulators will push it. I know a positive 3.3V regulator is commonly used on digital Eurorack modules, probably to power a STM32 IC so finding a negative one to double up is surely no problem at all. The 7*05 regulators dissipate a lot of heat due to a somewhat inefficient V conversion (if I remember correctly they require a lot of current).
I think you might be a little confused about this, or I might be confused about what you’re getting at. The datasheet says “Supply voltage range is from ±2.5V to ±9V”, so this means it can be powered from anywhere from +9V and -9V rails down to +2.5V and -2.5V rails. This means powering with +5V and -5V rails (for instance with a 7805 and 7905 regulating the +12V and -12V) is just fine. ±5V is the same voltage range that LZX is powering their ADA4851-4’s with, I believe. So compared to the ADA4851-4, it’s not any more limiting in it’s voltage requirements. (In fact, I think the MIC920 can handle a higher voltage range than the ADA4851-4, since it goes up to ±9V.) Compared to the LM6172, though, both the ADA4851-4 and the MIC920 have a smaller voltage range. Of those three, the LM6172 is the only one that can be powered by ±12V.
I haven’t pored over the datasheet yet but I did notice it looks like it might have an even faster slew rate than the ADA4851-4, which is good. I suspect others with more experience with the ADA4851-4 will be able to speak more to the differences.
IMO I very rarely need just one op amp at a time like this MIC920, so I tend to prefer dual or quad packages.
Yeah agree about the dual or quad packages but for this price it makes it cheaper per op amp than the ADA4851-4 which is very appealing!
It looks comparable specs-wise to video op amps like ADA4851-x and ADA4891-x, and suited for similar purposes. The price point is about the same (Microchip vs Analog Devices is maybe a little cheaper just based on brand name.)
One big thing to notice on this op-amp and ADA4851-x is that the input range of the amp is not rail to rail, but the outputs are. So if doing series gain staging, you need to make sure to attenuate the input between stages. ADA4891-x is rail to rail on both input and output, but needs either single supply or max +/-2.5V bipolar supply. I would like to publish some new reference designs this year.
If you are looking for a cheaper op amp that’s more of a midrange option that’s fast enough to pass SD bandwidths at low gains, the TL974 looks promising.