EEVblog Electronics Community Forum

Electronics => Beginners => Topic started by: tehtehteh on June 17, 2021, 04:41:33 pm

Title: Decoupling caps different between data sheet and evaluation kit.
Post by: tehtehteh on June 17, 2021, 04:41:33 pm
Using a MAX7044 transmitter the data sheet typical application circuit has a single decoupling capacitor and this statement: "Supply Voltage. Bypass to GND with a 100nF capacitor as close as possible to the pin."

https://datasheets.maximintegrated.com/en/ds/MAX7044.pdf (https://datasheets.maximintegrated.com/en/ds/MAX7044.pdf)

Yet the evaluation kit has 2 caps, 220pF and 0.01uF.

https://datasheets.maximintegrated.com/en/ds/MAX7044EVKIT.pdf (https://datasheets.maximintegrated.com/en/ds/MAX7044EVKIT.pdf)

I am searching for as much space as possible on the PCB and I wonder what your thoughts are on just using the 100nF capacitor. It's in the data sheet so it should be fine right?
Title: Re: Decoupling caps different between data sheet and evaluation kit.
Post by: Siwastaja on June 17, 2021, 05:02:04 pm
Being a high-frequency device, low impedance at high frequencies may be important. Maybe the devkit designer plays with the old classic myth of paralleling different capacitor values for lower impedance across the frequency spectrum. In reality, all that matters is the ESL. Paralleling multiple has advantage if they can get the 220pF in a smaller case and place it closer to the pin, providing a "quicker cache" for the energy. This was important back when MLCCs were not available in high capacitances and small packages.

In reality, simply for physical manufacturing reasons, you are likely limited to using 0402 or just maybe 0201 parts and have some courtyard placement limit like 0.2mm; and you can definitely get a 100nF cap in 0402 and even 0201 just fine, so it will provide same or lower impedance over the whole spectrum compared to strange mixes of parallel caps. Also, parallel mixes have a risk of resonating with each other.