Thanks for pointing that out. My comments were made because your question states the D5200 image sensor as being made by Sony and my recollection of this was Chipworks' article on their findings that the D5200's sensor was actually made by Toshiba, so the purpose of my comment was to lead you to the correct facts. But since I haven't previously dug about the later sensors from the D5300 and the D3300 I assumed that just like on the D7100 all were carrying Toshiba sensors. Now, after digging further, I just came across Thom Hogan's findings that Nikon is basically making their own sensors for the D3300 just as they did for the D3200 and D3100 and that they might actually be using a Sony sensor for the D5300.
Now all this makes absolutely no sense to anybody except Nikon, and their reasoning behind this scheme will be nice to know if you can get them to fess up. I mean, they need to spend 3 times the same resources to design 3 cameras that share the same basic sensor specs, for what reason? I doubt that none of the 3 different fabs they are using is able to handle the whole volume of Nikon's APS-C sensors by itself (Canon's fab can, and for the entire Canon line-up for that matter, even while that fab is ages old and should be highly inefficient by today's cutting edge fab standards, but then they are themselves their only customers), unless Nikon is doing this to prevent factory disruptions by forces of nature or politics, in which case they shouldn't even assemble cameras in Thailand at all given how explosive politics have ALWAYS been in that country, or at least divide assembly 3 ways between Thailand, China and Japan for all their cameras instead of placing each particular camera line on a different assembly plant as they are currently doing.
This is highly interesting to me because I have basically gone through all of Nikon's 24mp DX line-up within less than a year and have found meaningful differences in the IQ of each one I've owned: I started with a D7100 I bought last Thanksgiving and it currently holds the highest IQ watermark for me, but I found it cumbersome to use and carry (it was my first DSLR after all, coming straight from cell phones, point and shoots and Nikon Fs) and didn't like how it handled artificial lighting in JPEG, which tended to output greenish, making everybody in the pictures look like little martians. I then got a D3200 and loved its simplicity and lightness, I also held high hopes for its purportedly in-house Nikon sensor, but in actual use I didn't like its IQ as it consistently outputs under exposed and warm (maybe because of its simple 420 pixel matrix sensor, as surely the 2k one on the D7100 should be able to get more proper exposure and WB), plus its video output is very noisy compared to D7100's. So now I have a D5200, which has IQ similar to the D7100 in terms of exposure, WB and noise, but with sharpness more akin to the D3200 and, ugh, buttons all over the body and in the wrong places. I was going to wait until the D5300's price came down lower but there is a great deal going on with the D3300 and I am now waiting for one to arrive next week, just in time to try that 1080p60 on my kids next weekend. I was expecting the D3300 to get IQ and video noise akin to the D7100 as I previously thought that it has the same Toshiba sensor, but now that I know that it comes with a Nikon sensor I am a little pessimistic about this given how Nikon's own sensor has disappointed me in the D3200. Well, I'll see how it fares in a couple of days' time, but if I have to go through the D5300 and maybe a future D7200 or whatever to attain IQ nirvana, then so be it. I have learnt from this experiment that digital cameras shouldn't be held for more than a year because technology is constantly changing for the better and therefore used prices for them suffer huge drops pretty fast, and because of that you should only buy a new camera when it is at its lowest price possible, so when time comes to turn it around you don't end up loosing much money: in my case I have profited on each camera I've sold.