0
\$\begingroup\$

I'm working on a circuit using the Diodes Inc./Anachip AP1603 boost regulator.

The datasheet (linked above) notes:

Start-up voltage operation is guaranteed with the addition of a Schottky 1N5819 external diode between the input and output.

I'm slightly unclear as to where this should be placed (it's not shown in the sample schematics). Should it be like this:

Schematic: Diode from Vin to Vout

Or like this:

Schematic: Diode from LX to Vout

Side question: Any harm in using ceramic 47µf caps instead of electrolytic or Tantalum ones?

\$\endgroup\$
2
  • \$\begingroup\$ I’d say the first suggestion is correct. As for capacitors, ceramics should be fine but be aware of the dc bias derating spec of the capacitor. You may need a number of capacitors to achieve the required capacitance. \$\endgroup\$
    – Kartman
    Mar 9, 2022 at 6:12
  • \$\begingroup\$ I guess the diode is in parallel with the internal synchronous switch in case the switch cannot be fully enhanced at very low startup voltage. That makes sense. I have seen that low Vin synchronous boost regulators before. \$\endgroup\$
    – user57037
    Mar 9, 2022 at 7:42

2 Answers 2

1
\$\begingroup\$

The demo board shows the diode connected as per your second option.

enter image description here

You can use ceramic but as noted in the comments you may need to use a higher value, voltage rating and/or size to get a true 47uF at operating bias. MLCC caps can lose as much as 80% of their nominal value at bias- you really need to dig into the detailed data.

\$\endgroup\$
2
  • \$\begingroup\$ Is there any chance that the lower ESR of the ceramic cap will throw off the stability of the regulator? Asking because I don't know. \$\endgroup\$
    – user57037
    Mar 9, 2022 at 7:32
  • 1
    \$\begingroup\$ @mkeith Certainly an issue at times with linear regulators. I don't think so- this does not seem to be an issue on switching regulators, probably to do with the compensation method. In the datasheet they say you can use OSC-CON caps which would have likely have typical ESR << 25m\$\Omega\$. A suitable MLCC part for 47uF (49uF effective capacitance at 5V bias) is the "220uF" 6.3V GRM32ER60J227ME05 which has an ESR of more like 1m\$\Omega\$ up to a couple hundred kHz, but PCB traces will add to that. \$\endgroup\$ Mar 9, 2022 at 7:49
0
\$\begingroup\$

I think the second suggestion is correct; for this section you can also see these application notes for best answer:

https://toshiba.semicon-storage.com/info/docget.jsp?did=66549

https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/196/Schottky-Diode-Selection-in-Asynchronous-Boost-Converters.pdf

For side question, my guess mlcc can be good, but 47 μF requires more MLCC and can be expensive.

\$\endgroup\$
1
  • \$\begingroup\$ This is a decent answer but the problem is that links almost always become stale after a short time. So it would be better to provide a graphic (as well as the link). This will insure the answer remains valuable for a longer time. \$\endgroup\$
    – user57037
    Mar 9, 2022 at 7:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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