0
\$\begingroup\$

I have no background in electronics.

I'm trying to understand Sharp Memory Display. I read the device specification to the best of my limited abilities, but still confused about a few things.

  1. Why does it have analog power pins (VDDA and VSSA)? From my understanding, it communicates only with SPI which is digital, so there's no need to have separate digital and analog power inputs. I randomly searched for some other lcd modules like this one, none of them seem to have it. What am I missing?

  2. This is Adafruit breakout board for Sharp Memory Display. In the schematic, it shows a ferrite bead between power and VDDA, also between ground and VSSA(AGND). These ferrite beads are also present in the design of nice!view, another breakout board. In the Sharp Memory Display specification there is recommendation about decoupling capacitors, but there is no mention of ferrite beads. I only vaguely understand ferrite beads are used to reduce noise. Are they needed here? Or is it just good practice to put a ferrite bead between VDD and VDDA?

schematic for Adafruit breakout board

\$\endgroup\$
1
  • \$\begingroup\$ Only the digital interface (SPI) works from the digital supply. The analoge part (The display Cells) are powered by a different supply (Analog). The ferrit beads are good practice but for hobby applications not mandatory so to speak. In product designs, they are used to reduce EMI and stuff which can lead to problems with the displayed image (Flicker e.g) and also reduces the coupling effects from the display to the board when the display has power spikes (e.g) while changing display contents) \$\endgroup\$ Commented Feb 25 at 7:46

1 Answer 1

0
\$\begingroup\$

The information why there are two supplies, what they do and why they are called analog and digital cannot be found in the data sheet you posted. And, it really does not matter if you intend to use the part. Simply connect both supply pins and it will work. If you then, find out there are problems, then you start to troubleshoot the problem, but sometimes, it is good to prepare for potential problems in advance, and draw extra components that could prevent problems, but each approach increases the cost of making the device.

However, if we start guessing, compare it to an MCU. It might be that the digital supply is used for powering digital logic and it can cause some voltage ripple/noise or tolerate voltage ripple/noise without loss of performance or function. The analog supply might be used to more sensitive parts that can be thought as analog, like oscillators, power-on reset circuits or maybe even internal ADCs and DACs, so this supply needs to have less ripple/noise and also does not cause much ripple/noise.

So that could be a reason why there two supply and ground pins on the connector, so that you can use separate wires for sensitive and noisy circuits, because if there was only one supply/ground wire pair, the noisy circuits could cause noise to the supply wire and disturb the analog performance. With separate wiring, you can use bypass caps and good power supply filtering on PCB so the noisy power wire cannot disturb the sensitive wire. If the analog supply is used in some way for driving the screen pixels, any noise on the analog supply could be seen as flickering picture on display.

2)

About the ferrite beads, there is no one true answer. Yes, ferrite beads are used to filter noise. Yes, it is a good practice to filter noise, if you need to. However, we only see the schematics, not the engineering desicions or measurements leading to the ferrite beads being in the schematics. So again, we cannot guess if they are really needed here. We cannot guess if the ferrite bead values and capacitor values have been properly chosen, or just put there because all the designer saw that it requires both a digital and an analog supply. The ferrite bead values and part types are not even shown in the schematics so we cannot see how much it filters and at what frequencies, so we cannot replicate the design. The ferrite bead between GNDA and GND is also weird, it can even be wrong, and cause more problems, but we don't know that. Usually you want the grounds to have low impedance and connect them directly together, but again, the designer might have made this based on display manufacturer support people suggestions, example schematics, or based on measurements, or just drawn whatever is the first thing that comes to mind in 5 minutes and decides the product is good enough and ready to be sold to hobbyists. So the ferrite beads might not be necessary and depending on their value they could also make the circuit worse.

Having said all this, there indeed are chips that have multiple supply and ground pins, that are not connected inside the chip, so that different submodules on the chip can be powered separately and have different supply and ground return paths that only connect on PCB, in a way that makes most sense. Some sensitive circuits need more filtering than others to guarantee good performance, and even if the chips are digital, the interfaces they have to outside world may be in fact analog, like a gigabit ethernet PHY interface chip, and these high-speed interfaces need noise-free supplies with good filtering in order to pass tests that the interface is within specification and that it also passes tests for radiated and conducted noise emission tests.

\$\endgroup\$

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.