Some of Analog's still rock. I used this chip a few years ago and its datasheet kicked total ass. https://www.analog.com/en/products/ad9739.html
And yes, that /is/ a 2.5 GSPS 14 bit DAC. :D
@zetasyanthis what is even the use case for that
-F
Lots of things. DOCSIS to your cable modem, for one. Cell towers, satalite communications, etc...
Can run it in 1st, 2nd, or 3rd nyquist, too, and get out usable signals to 3.75GHz with the right filters.
If you need a fast ADC to receive on one end you need a fast DAC on the other end to transmit. :P
@zetasyanthis I'm thinking of ADCs for oscilloscopes, not for receiving transmitted data
-F
@zetasyanthis we're a test engineer, we don't work with *functional systems*!
-F
@zetasyanthis speaking of, did you see our thread on hotspotting the other day? Seems like stuff you'd find interesting
-F
That is absolutely fucking wild. I did not know those crystals could be made to that. XD
@zetasyanthis we didn't either until a week or two ago!
-F
Also, now I have to tell you that one of the power supplies at my second internship was named the 'tracebreaker' for similar reasons. You took boards with a short, and it would find them for you, ideally with the thermal camera, but it was a 30A linear job... :P
Bonus: I once had a board it couldn't find the problem with! Turns out the PCB and assembly house both fucked up and managed to skip e-test on a set of boards. This one had one corner under a large connector just... not etched. Confused the hell out of us until I hacked the thing off to find out what was under it.
Double bonus: I once accidentally made a board that passed all the tests, but had like 20A of current slated to pass between the pads of an 0805. This caused a bit of ground ripple for the thermocouple amp also on the board, even though it was semi-isolated. :P
Since that was a prototype, I ended up scraping some solder mask off on both sides of a ground plane, drilling a hole in the board, and punching in an aluminum rivet (with a little solder for proper connection). Problem solved. Rev B board fixed it. :P
@zetasyanthis while we were trying to figure out how to use the hotspotting technique with liquid crystal, we thought we'd killed a die when we saw smoke coming up from it... the liquid crystal went completely black over the entire surface, too. but it was working perfectly fine afterward, somehow
when cleaning up at the end of the day, it turned out that there was a perfect black rectangle in the cardboard we'd used to keep the liquid crystal from getting all over the probe station
-F
I do not quite follow. Was the cardboard on the die?
@zetasyanthis the die was sitting on the cardboard, so that liquid crystal spilling off the edge wouldn't get into the vacuum system (intended to hold wafers down)
-F
@zetasyanthis we don't really have a proper setup for working on individuated dice right now
-F
Microscope slide maybe?
But yeah, I get you. Those are /tiny/.
@zetasyanthis this one is pretty large, actually; it's nearly a centimeter on its long side! and like 3 mm on the short side
-F
@zetasyanthis it actually has an interesting thing we haven't seen before; several 01005 capacitors are to be mounted *on* the die
-F
Am I reading that right? 1mm by 0.5mm? I've never even thought of something that small being a standard package. XD
@zetasyanthis no, 0.01 by 0.005 inches
roughly 0.4 by 0.2 mm
-F
I forgot those were in inches. Gosh I hate some of the things we do with units sometimes. (Looking at you, mils!) XD
And yeah, that's still pretty damned small.
@zetasyanthis afawk it's the smallest standard size currently available, though i suspect metric 0201 will be available at some point (if it isn't already)
-F
All I know is that anything smaller than 0805 is not something I want to do by hand. XD
(I've done 0603s, but that was miserable, and I have no desire to go smaller. XD)
@zetasyanthis https://www.digikey.com/en/products/detail/taiyo-yuden/JMK021BJ103KK5W/6198379 ah, yes, it is, here's 10 nF somehow squeezed into an 8 mil by 4 mil package
-F
I mostly play in the uF range myself, but even those are ridiculous for the density. (As long as you don't make them unhappy and have them go kaboom...)
Oh right, I forgot what type of chip you were working on for a moment. Makes sense that'd be a little larger. :P
@zetasyanthis some of the pads are visible without a microscope, as long as you have really good eyesight!
-F
Ah, yeah, that'd do it. :P
@zetasyanthis it's really weird when you see magic smoke and then everything seems to be working perfectly! (as perfectly as it was before, anyway, what with the large quiescent current problem we were trying to pinpoint)
-F
@zetasyanthis there's a 750 volt, 40 amp power supply in our coworker's lab
we have no idea what they're doing with it, but it's a Lot
-F
That is quite frightening, but I suppose I did just request, uh... another 5MW of power for an upgrade at one of my sites? XD
@Felthry
I did not! Link?