# What is a good methodology to identify small SMD components?

Having been in the position of trying to identify a strip of mystery 0603 components I found on my floor, I thought a good general question to pose to this community might be what a recommended set of steps could be to identify similar unknown components.
I identified these by using package (0603) and color (black) and then thinking about where they might have come from, and how they might have ended up being on the floor. Luckily I didn't need to use any test equipment to work out they are a 4.7 $\mu$H inductor because I'm not sure how I would have proceeded to get to this end result without my extra clues; hence this question.

I'm looking for a reasonable set of steps and methodology to follow that would allow me to identify most small components (new or recovered) with no (or simple) markings. Presumably identifying package is the first step and the last steps would be using test equipment.

I searched and did find resources like this that mainly deal with through hole and traditional components but very little on these surface mount items many of which share similar packaging.

• You could write "4.7 uH" on the strip before you drop it on the floor. Nov 5, 2015 at 17:48
• This is what I use -> smarttweezers.us . Nov 5, 2015 at 18:04
• @DanSut, for serious work, you throw the parts away and buy new. Even if you have an LCR meter to work out the value of the parts, you won't know the voltage/current/power ratings, whether they're lead-free or not, ... Nov 5, 2015 at 18:59
• Or a component tester like this one: ebay.nl/itm/…, it detects and displays automatically what component type it is and its value. And OH, about 8x cheaper than the tweezer type LCR meters and it can also test transistors, fets. Not so accurate but VERY usefull. Nov 5, 2015 at 19:44
• Going forward, you could do something like this. Nov 7, 2015 at 23:02

I'm surprised noone mentioned the various SMD codebooks [for active components] one can find a google search, e.g. the largest seems this one. I've identified fairly obscure Ricoh chips [voltage regulators, low-voltage detectors] and run-of-the mill SMD transistors using that. Because the codes are not unique, it helps if there's more than one component from the same manufacturer on the board and also if you have some clue what they do in-circuit. For stuff found in a junk storm, it might be harder.

The best general document for identifying SMD passives I found is Wikipedia's SMD page and this IAEA one. I suppose you know SMD resistors have standardized codes. Alas no codes whatsoever for ceramic SMD caps, so you have to ultimately measure them. There is a rough relationship with color and size, but I was never able to find a certain one. Wikipedia proposes some ranges for each color (alas without any reference), but the ranges are pretty large and overlap. The tantalum SMD caps are marked though. Inductors are just as bad as ceramic SMD capacitors when it comes to marking (and the THT inductors, unlike THT, caps are just as bad.) Coilcraft uses a color code for theirs (above 0603 size), but I haven't seen it adopted by others. The thing you can be fairly certain of is that SMD inductors are likely to be black and the capacitors more likely to be light grey or beige/maroon, as you probably know already. Also inductors are usually magnetic.

And that's all I know, as the saying goes.

• Also, some components aren't in any databases, e.g good luck finding this 07 one or an IJ99G (which doesn't even have datasheet anywhere, but one can find it sold on taobao together similar "IJ" chips like ij9da, etc.)
– Fizz
Nov 8, 2015 at 20:11
• Since asking the question I have actually found and started using ElectroDroid which appears to contain at least some of the info in the resources you link to with a semi-automated way of finding them. I was actually looking for more of a flowchart style answer such that the process of identifying could perhaps be coded and maybe even made into a service/app/website... the answers and comments here have given me great info that makes me think this still might be possible even if it would never be perfect. Nov 13, 2015 at 14:58