How I identified those mystery SMD caps - a method that actually works

Nov 05, 2026 Last reply: 19 hours ago 5 Replies

First reply

HarvestNomad736

I had one of those yellow blobs last month on a dead audio board from a friend, and I learned the hard way that the footprint match only gets you so far. I pulled one, dropped in the same size from my drawer, and the thing buzzed like a fridge untill I realized I'd used an…

Read the full reply

Been down this road more times than I can count. Tiny yellow tantalum cap, almost no marking, code that looks meaningless. Here's what worked for me. First, the code on the body is ussually a date or lot code, not the capacitance or voltage, so stop trying to decode it into specs. Measure the thing instead. Length and width in mm narrow it down to a package size, and the color plus a polarity bar tells you it's almost certainly a tantalum.



Next, work out what the circuit actually needs. Find where the cap sits on the board. If it's on a power rail next to a chip, it's a decoupling or bulk cap and anything with the right capacitance, voltage rating well above the rail, and same footprint will do. If it's in a signal path, be more careful about the exact value.



Then just measure capacitance with a meter if you have one. Comparing against a known good cap of the same physical size from another part of the board helps too, since manufacturers usually use the same part everywhere on one design. Worst case, I replace with a modern polymer part of the same size and a voltage rating comfortably above the rail and it just works. Honestly half these tiny caps are generic once you match the footprint and rating.


I had one of those yellow blobs last month on a dead audio board from a friend, and I learned the hard way that the footprint match only gets you so far. I pulled one, dropped in the same size from my drawer, and the thing buzzed like a fridge untill I realized I'd used an electrolytic-looking substitute that leaked way more. Put a proper tantalum back in and it went quiet. So I'd add: match the chemistry too, not just the width and length, especally anywhere near audio or anything that cares about noise. On the meter side, I've found cheap capacitance readings on these little caps can be all over the place if the leads are long, so I solder the part across a couple of short stubs to get a believable number. Also worth checking for a short with continuity mode before you reuse anything, since a failed tantalum often goes dead short and I once nearly cooked a rail testing one in place.

One thing I'd add on the voltage side, since I got bitten by this years ago on a repair where everything looked right but the replacement kept dying. Tantalums realy want a lot of headroom, so the common practice is to pick a rating around double whatever the rail is, meaning a part marked for 6 volts or so is living dangerously on a 5 volt line even though it looks fine on paper. Polymer parts are a bit more forgiving about this, which is another reason I lean that way now. Also, if anyone is chasing that 650uF figure from the title, check wheather the seller measured it with the meter set wrong, beacuse I've seen numbers inflated like that plenty of times.

Kind of relieved reading this, I've been staring at a tiny orange cap on a dead router board for a week trying to decode the numbers on it. Question though, does the measure and match approach work the same if the part is one of those flat brown ceramic ones instead of the yellow type? That's mostly what I have in my drawer. Also is it okay to go higher on the value or does that mess with the circuit somehow? I only know from swapping fuses that bigger isn't always safer, so I've been too scared to just wing it.

One thing that burned me early on with these yellow ones is the polarity. The bar on the end is the positive side, wich is backwards from every electrolytic I'd ever soldered. Pop one in thinking the stripe is negative like on a big can cap and it'll let you know pretty fast, usually with a small pop and a mess to clean off the pads. I killed two before someone told me. Now I double check the bar against the board marking every single time, no exceptions.

Update: the headroom comment hit home. Went through my drawer stock and half the parts I've been subbing in are rated right at the rail voltage instead of double, which probaly explains a couple of mystery failures over the years. Question though, does that same double rule apply to the newer polymer type, or only the old manganese ones?

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required