We have a standard SMD 0603 capacitor kit from that I picked a capacitor for use in a 868MHz Antenna matching circuit. I picked this 3.3pF capacitor from the kit.

The Q is listed a 1MHz, but I read somewhere that this value is not valid at 868MHz nor can the ESR be calculated from it at 868MHz.

But nowhere in the datasheet can I find the expected ESR or Q at 868MHz, so is this a good choice for 868MHz. ?

  • 3
    \$\begingroup\$ It will probably work but you will have no guarantee that it does. To be sure (and have a manufacturer's guarantee) it is much better to use a capacitor that is specified to work at 868 MHz. If you put this cap in a product and the next batch is a bit worse at 868 MHz and your product stops working, you cannot blame the manufacturer as this cap was never intended to be used at 868 MHz. \$\endgroup\$ – Bimpelrekkie Jun 14 '16 at 10:23
  • \$\begingroup\$ Murata provides graphs versus frequency. Consider the GRM1883C1H3R3BA01# as a possible equivalent. Looks like not much concern @1GHz. \$\endgroup\$ – user2943160 Jun 14 '16 at 16:34

It's a C0G/NP0 dielectric, which is 'pretty good'. That is, it's not a high K ceramic. Unless you have a very demanding application, so a very low noise VCO, a very narrow filter, or very high power transmission, it will probably do for Q. There are higher Q capacitors available, look at ATC for instance, if you need it.

0603 is a reasonable package size to be using to several GHz.

The capacitor is specified as 'multi-layer'. This is where you might have problems if you want to switch between different manufacturers, as the detail of the internal construction will affect the residual inductance. If you stick to the same range from the same manufacturer, you ought to be fairly safe. In practice, 1GHz is unlikely to cause a problem due to construction variation, unless you are really pushing the performance. I certainly agree with Oleksandr's comment that the case size is likely to introduce more residual inductance than the internal construction

Try some in circuit, and see whether they work.

You have specified that the application is 'antenna matching'. Capacitor Q will have two potential effects here in this application. On receive, the loss will increase noise. However the Q would have to very bad to be significant compared to other variables. On transmit, the loss will reduce the radiated power, and cause the capacitor to get hot. Again the Q will need to be very bad to be significant in the transmitted power calculations. How powerful is the transmitter, what loss would be needed in the capacitor to cause its temperature rise to exceed data sheet limits?

Rather than go all the way to ATC's head-banging RF performance, you could try looking at a range of capacitors that do discuss RF Q, for instance Kemet's here, same C0G dielectric, same case sizes.

  • \$\begingroup\$ Good answer, +1. I just note that the pad/trace inductance normally outweighs the internal inductance of a capacitor, which is a lot smaller than one would probably expect. You can get capacitors that are wider than they are long, or that have many interspersed pads of alternating polarity, to help with layout problems. \$\endgroup\$ – Oleksandr R. Jun 14 '16 at 11:51
  • \$\begingroup\$ But, can I from the datasheet alone determine that the capacitor will do for my design. I mean, is there somehow that I can extrapolate the Q listed at 1MHz to 868MHz ? \$\endgroup\$ – JakobJ Jun 14 '16 at 12:11
  • 1
    \$\begingroup\$ @JakobJ you cannot meaningfully extrapolate from datasheets and expect the extrapolated performance to be stable or in any sense guaranteed. Either the datasheet lists performance characteristics under the intended operating conditions (or something representative of them), or it doesn't, in which case all bets are off. Here you wish to extrapolate nearly three decades higher in frequency than the datasheet considers. The only reliable way to determine its suitability in this extreme case is to characterize it for yourself. \$\endgroup\$ – Oleksandr R. Jun 14 '16 at 12:45
  • \$\begingroup\$ Ok, so using this capacitor in production would not be a good idea. \$\endgroup\$ – JakobJ Jun 15 '16 at 4:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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