4
\$\begingroup\$

On page 72 of the AD7124-4 datasheet it says:

Decouple AVDD with a 1 μF tantalum capacitor in parallel with a 0.1 μF capacitor to AVSS on each pin.

Why would this capacitor be specified to be a tantalum?

At this low capacitance a ceramic MLCC is cheaper and has superior ESL and ESR.

Is there a reason for this?

\$\endgroup\$
  • \$\begingroup\$ General note: ESL and ESR aren't the only “nonidealness” measures for a capacitor. One that's often overlooked is nonlinearity, usually quantified via THD. In that regard, both tantalum and ceramic are among the worse types, with some ceramics being really really bad. \$\endgroup\$ – leftaroundabout Aug 7 at 21:44
7
\$\begingroup\$

The LDO built into the chip needs an ESR that is high enough to ensure stability under all conditions.

You can use a 1uF ceramic capacitor with a series resistor of 5 or 10 ohms in place of the tantalum capacitor. And keep the 100nF capacitor in parallel.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Is 5 to 10 Ohm not a bit much? Most tantalum datasheets seem to specify an ESR of <1 Ohm at DC *edit I was looking at Tantalum-Polymers, 5-10 Ohm for standtard 1 uF tantalum seems right \$\endgroup\$ – Miomio Aug 7 at 12:05
  • \$\begingroup\$ @Miomio Looking at Digikey for 1uF ESR varies from 1.3 to 25 ohms. If it was a much larger value I would have suggested 2-5 ohms. \$\endgroup\$ – Spehro Pefhany Aug 7 at 12:07
  • \$\begingroup\$ @SpehroPefhany: Does anyone simply add small series resistors to MLCCs? \$\endgroup\$ – JDB Aug 9 at 2:06
  • \$\begingroup\$ @Miomio: TI has a wonderful app note on LDO stability with respect to MLCCs: ti.com/lit/an/snva167a/snva167a.pdf \$\endgroup\$ – JDB Aug 9 at 2:08
  • 1
    \$\begingroup\$ @JDB yes, it’s not uncommon. I’ve done it to use dirt cheap 1117 type regulators as well as shunt regulators. Here is a TI app note where they specifically discuss it. \$\endgroup\$ – Spehro Pefhany Aug 9 at 2:17
4
\$\begingroup\$

It's pretty much legacy information. For instance, the AD7192 on page 36 says the same thing and that dates back to 2009. The AD7124 is 2015 vintage and it appears to have just copied over some of the same words: -

enter image description here

If I could be bothered to go back further I'm sure I'd find that pretty much all similar devices that are older say the same thing.

You can use MLCC capacitors.

| improve this answer | |
\$\endgroup\$
2
\$\begingroup\$

MLCCs have mostly better ESR and ESL compared to tantalum capacitors, but tantalum capacitors are a bit stronger against voltage and temperature fluctuations. For example, an MLCC's capacitance may change 40% at 50% of rated voltage.

Plus, some converters need a bit high ESR and a bit high ESL for stabilization (e.g. to prevent unnecessary oscillations).

| improve this answer | |
\$\endgroup\$
1
\$\begingroup\$

Tantalum capacitors are specified to make life exciting and to help remind you of the bad old days :-).

A Tantalum capacitor has good capacitance per volume and low ESR, at the expense of a propensity to (any or all of) smoke, shriek, burst into flame and explode when subjected to small voltage excursions above rated value when connected to an other than low energy source. In this application it may be safe enough. Or not.

This answer Are tantalum capacitors safe for use in new designs? explains (at great length :-) ) the relevant considerations.

In this case, as others have noted, using an MLCC, possibly with a small series resistor, will do as good a job and avoid the more exciting aspects of tantalum capacitor failure.


From Kemet "APPLICATION NOTES FOR TANTALUM CAPACITORS"

Capacitor failure may be induced by exceeding the rated conditions of forward DC voltage, reverse DC voltage, surge voltage, surge current, power dissipation, or temperature.
As with any practical device, these capacitors also possess an inherent, although low, failure rate when operated within the rated condition. The dominant failure mode is by short-circuit.
... Catastrophic failure occurs as an avalanche in DC leakage current over a short (millisecond) time span. The failed capacitor, while called “short-circuited”, may exhibit a DC resistance of 10 to 104 ohm.

| improve this answer | |
\$\endgroup\$

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.