Part 1: PCB design basics （1）

Published:2011/8/23 1:26:00 Author:Li xiao na From:SeekIC

By Karel Walraven

What do most electronics engineers enjoy a lot? Designing printed circuit boards, of course! In an irregular series of articles we would like to familiarize you with this subject. We start as simple as possible and it will become more difficult as we progress. Feedback and requests to cover certain topics are encouraged and welcomed!

When you’re just designing a small PCB, it often is not that important where The actual components ore placed. However, there are a few rules of thumb that, when followed, quickly and easily result in a much better PCB. Belter in this context means lower noise in the desired output signal and loss interference with other equipment. The last sentence means, translated into practice: "your amplifier provides a better sound, makes no additional noises when the fridge turns on and does not interfere with your mobile phone".

The following statement is perhaps surprising, but is actually true: "The optimum position of a component on o PCB is for nearly 100% determined by its function in the schematic diagram." That means you cannot avoid thinking about the function of each and every component in the circuit.

Example

We use an amplifier stage as an example. What are the different functionalities that can be distinguished?

In Figure 1, C5 is a decoupling capacitor, which decouples the power supply voltage. The current drawn by ICI will change during operation and C5 ensures that the peak current demand can be delivered.



CI is a decoupling capacitor as well, but in this case it serves the purpose of preventing any hum and interference from the power supply from reaching the signal.

C3 is a decoupling capacitor too, but this lime not for the power supply voltage, but for the signal instead. Its function is to ensure that the inverting input of ICI is connected to ground, via R4, for AC voltages.

Rule 1: Decoupling components must be placed as dose as is practicable to the nodes that need to be decoupled.



Long traces hove a higher impedance |= resistance for high frequencies) and as a consequence the decoupling will not work as well. The higher the frequency, the more important it becomes to keep traces shod. Pay close attention! To clarify this a little more, take a look at Figure 2a. The traces towards and away a/ways form a loop (also when one side of the component is connected lo ground). In the figure we have shaded the area between the traces for clarity. We cannot overemphasize that it is very important to place the traces in such a way that the shaded area is us small as is possible. If you have to choose between long tracks with a smaller enclosed area or shorter traces with larger enclosed area, then choose the layout with the smallest area" Why that is so we will explain at a later time. It is a little too advanced for the present discussion.

C12 and C4 are capacitors in the signal path. In this case it is also true that short traces and small area provide less opportunity to pick up trouble or cause trouble. We repeat the above once more because it is very important that you think about this.



How does a signal connection have an area? In figure 3 you can see that the signal does not only go towards the capacitor, but also has a return path via ground! The ground trace has a certain length as well! Signal path arid ground path together form o loop! This should not come as a surprise to you. Your very first acquaintance with electronics probably involved o small lamp and a battery. And when did the lamp light up? Just when the loop is closed... So, pay close attention lo Rule 1.

We can make a few additional comments with regards to the signal path. The older youths among us, those who grew up with record players and valves, will remember how easy it was to lest an amplifier from days gone by. A finger in the vicinity of the input was enough to generate a loud hum from the speaker. This worked so well because the input to a valve amplifier is high impedance. High impedance connections pick up interference much easier. Always try to make on estimate of how high the impedance of a node is and pay extra attention to make sure that the connection is as short as possible and - if possible -run a parallel trace connected to ground as a shield (= keeping the area small*).

Rule 2 is therefore: Keep high-impedance connections shorter than low-impedance ones.

R6 ensures that the output of the circuit in the idle state is exactly zero. This prevents spurious noises from the loud* speaker when the amplifier is switched on or off.

R1 and R2 provide for the correct DC bias of the non-inverting input of the opamp. This node is decoupled with capacitor C and is therefore not that critical. After al, the capacitor creates a low-impedance connection for interference signals, making the entire network less sensitive. The resistors may be placed a little further away from the opamp, provided the decoupling capacitor is placed as close to The opamp as possible.

R4 together with R5 determine the amplification. The input of the opamp is the node with the highest impedance, the output is much lower impedance and ground is low impedance as well. The input, therefore, must have the shortest connection.

Reprinted Url Of This Article: http://www.seekic.com/blog/project_solutions/2011/08/23/Part_1__PCB_design_basics_（1）.html

Print this Page | Comments | Reading(1207)