4
\$\begingroup\$

I have been looking for an 0805 10uF ceramic capacitor rated for 100v. Mouser and Digikey have only appalling prices for such specs (50 cents/piece at 5k for example).

I recently made this shocking discovery on Aliexpress from a (somewhat) reputable dealer: https://www.aliexpress.com/item/33053004449.html

The price per item comes out to a few cents, and when comparing to U.S. prices, I am wondering if it's a scam. I can see one of these sellers selling 50v caps (much cheaper from U.S. distributors) as 100v caps, and my design blowing up with a bang.

Are these legit? I know it couldn't hurt to test them (100 of these is the same price as three from Digikey), but if there's a 0% chance they will work I would rather look elsewhere and not bother waiting for the shipping.

Perhaps this is an opportunity for someone who knows more than I do about ceramic capacitor manufacturing to explain the cost impact of increasing the capacitor's capacitance and voltage rating. Three factors that come to mind are more expensive materials, specialized design process, and lack of demand to produce economically in large batches. Are any of these remotely reasonable?

\$\endgroup\$
  • \$\begingroup\$ I'm just a hobbyist. Get that right. But I do buy from China, regularly. And the products almost always are fine. I've had some products that weren't well protected and needed coddling to make work well. But even then, the prices were fine. I just had to make sure that I took care of the details. So far, I've not gotten any products that are actual fakes and don't meet the specs I needed for the application. However, I didn't test them for the entire specified ranges, either. I'd be willing to take a chance, myself. But also willing to throw them away if there was a problem. \$\endgroup\$ – jonk Dec 9 '19 at 6:54
  • \$\begingroup\$ Here's the important thing about AliExpress. Anytime you see a super low price there, you should set your expectation that this is the only reasonable price for that thing. Thenceforth, reject all other suppliers who quote a price not in that range. Hold out for that price, no matter what. It is simply unacceptable to pay more! \$\endgroup\$ – Harper - Reinstate Monica Dec 9 '19 at 8:05
  • \$\begingroup\$ Note that ceramic caps are rated for 2x overvoltage excursions typically, so a50V cap will survive 100V for test period, which is what does make them attractive for ripple filters and inductive decoupling. An unscrupulous vendor may simply be selling based on the do not exceed voltage (typically exposure limited to a minute or so) and not the true rating. To qualify a second hand capacitor stock I repeated the MFG test spec, which included a 60second overvoltage condition that resulted in a return to normal capacity within 24 hrs \$\endgroup\$ – crasic Dec 9 '19 at 20:39
  • 2
    \$\begingroup\$ @Harper ... or what you saw is an underperforming/fake part... and by blindly rejecting the other suppliers, you'll end up with a product that will not meet the expected MTBF/perform poorly/explode and you'll have a lot of unhappy customers in the long run. Things aren't that clear-cut. \$\endgroup\$ – dim lost faith in SE Dec 9 '19 at 21:03
8
\$\begingroup\$

0805 10uF 100V ceramic = scam , The dielectric constant is not high enough

For ceramic with very thin , many layers you can achieve 10uF in small area but not 100V. You can achieve nF 10kV in 1 layer , but not 10uF at this voltage and size.

The product of these 3 parameters : size, capacity and voltage is not achievable in ceramic.

Since capacitors are highest stress components of all. Caps have the lowest MTBF, along with high demands for low ESR for performance. It is wiser to design with conservative expectations, add redundancy and derate limits to improve MTBF. Thus use more components to share the stress of ESR, ripple current, and operate < 75% of rated voltage or less. Just as e-caps have a design limit of ESRC for any ,size/family, ceramics have a design limit for VC*size which must be respected and used conservatively.

Never try to use 100% of capacitor ratings. This like buying 6000 mA Li Ion 18650 cells. They do not exist or are extremely unreliable.

1210 10uF 100V ceramic = possible but never perform with full charge Q=CV=10u*100V because charge capacity limits in size cause capacitance to reduce with raised DC voltage in ceramic.

\$\endgroup\$
  • 1
    \$\begingroup\$ +1 - It's interesting to note what you CANNOT buy either from Digikey or LCSC. \$\endgroup\$ – Russell McMahon Dec 9 '19 at 11:42
  • \$\begingroup\$ There are stacked parts from TDK and United Chemicon that are small 1206 ish, but to your point if no major MFG offers this bin its too good to be true \$\endgroup\$ – crasic Dec 9 '19 at 19:54
  • \$\begingroup\$ Yes and reality check on CV=Q limits as Spehro shows with graph for Y5V \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Dec 9 '19 at 20:01
10
\$\begingroup\$

Here is the X7R availability table from AVX:

AVX X7R Table

The highest capacitance for 0805 100V is 0.47uF; 10uF 100V is not available even in a 1206 package (I am willing to concede that the next capacitance value may be available in X5R).

Personally I would give this one a miss; AVX is one of the best sources out there and they don't make anything even close.

\$\endgroup\$
  • \$\begingroup\$ Agreed. Isn't this simple physics? If the size is that small, it's not feasible to keep enough clearance yet still produce a big enough capacitor. So if AVX can't do it, a less tech-savvy company is highly unlikely to do it. Major chance of overspecced (underperforming) parts here. \$\endgroup\$ – Mast Dec 9 '19 at 20:35
6
\$\begingroup\$

Digikey sell Murata 0805 50V 10 uF X5R for 18.8 cents in 100s.

Digikey do not sell 100V 0805 10 uF ceramic capacitors, at all.

LCSC, an as reputable as they come high volume major Chinese distributor (located in Shenzhen) also do not sell 100V, 0805, 10 uF ceramic capacitors, at all.

For commodity items (and much else) LCSC pricing usually (not always) beats typical Western suppliers into a cocked hat.

LCSC will sell you Murata 35V X5R 10 uF ceramic caps for from US 5 cents in 150's, and 4.4 cents in 5000 quantity, and TDKs from about 11 cents in 100s.

I doubt that I need to "join the dots" re the claimed specifications of the capacitors concerned or, probably, the supplier.

LCSC guarantee provenance of products sold with money back offered if goods found not genuine. [I'd feel happier without the money back part, but it's better than you'll probably get from Ali Express sellers].

\$\endgroup\$
2
\$\begingroup\$

Chances are the part, if it actually is a capacitor and actually measures anything like 10uF with no bias, will be more like 0.1 or 0.5uF with 100V across it, assuming it doesn't fail short. Note that it could actually meet the claims - 10uF and 100V, just not at the same time.

The voltage limit on many ceramic caps is determined by the capacitance reduction rather than the breakdown, somewhat analogous to inductor saturation. For example, the smallest genuine part I found in a quick search, the Murata GRM32EC72A106KE05L 1210 package has the following typical voltage characteristics:

enter image description here

At 100V bias it's down to almost 1uF, practically missing in action. And that's the genuine one that costs serious money even in 5K.

This is the equivalent of the 10,000mAh 18650 cell for $3. It fails the sniff test. If world-class companies like Taiyo Yuden, AVX and Murata can't do it at any price..

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.