Techniques for Designing Compact PCBs
With technology continuously getting smaller, there can be no reason to assume that the trend will stop any time soon. Therefore, designers are under constant pressure to reduce their board size. Here, Rush PCB Inc. shares some common methods we use to shrink PCB design. During the design, it may be necessary to review it several times to weed out errors before final submission. Iterations may be necessary to achieve a greater degree of compactness. Therefore, we have divided the process of designing compact PCBs into three parts and will examine each part in detail.
Plan the Design
How Many Layers?
Typically, designers begin their prototypes as a single- or, at most, a two-layer board. They have the advantage of surface access to all components and traces for easy repair, rework, and testing of the circuit. Once they have proven the concept, the designer is ready to shrink the board. Increasing the component density is the first step until there is no more room to route the connections.
Most designers use modern EDA tools with auto-routing capability. This allows them to check if the tool can route the board automatically within a reasonable amount of time with a certain number of layers.
Use Combo Modules?
Although the prototype may have used discrete solutions for cost-effectiveness, it may be taking up too much of the board’s real estate. Using combo modules could be a solution to reducing the footprint and the size of the board. For instance, using an LDO may be less expensive, but may require filtering with additional components. Using a DC-DC converter, on the other hand, may reduce the overall footprint while providing additional isolation as well.
Use Heavy Copper?
If the design requires considerable thermal management due to high current traces, using heavy copper may be a solution for a smaller board. Although minimum trace widths are useful in reducing the board size, they tend to warm up with the passage of higher currents and require more space for cooling. Using heavier copper reduces the resistance of the trace, thereby keeping it from heating up.
Use Custom Size?
Sometimes, the height of the board may be more of a problem than its width or length, for instance, due to the presence of a large capacitor. The solution may be to split the capacitor into a larger number of smaller types placed close together. Although this may take up greater board space, it provides the clearance necessary to fit the board into a tight space.
Use Resistor Networks?
Most microcontroller-based designs require many series pass resistors for preventing high current flow or a series of pull-up or pull-down resistors for IO pins. Using a resistor network rather than multiple discrete resistors can save a lot of space on the board. For instance, a simple 1210 package of resistor network can save space for about 4 to 6 resistors.
Use Stacked Packages?
Designs often require multiple transistors. Using stacked packages can save the space taken up by individual transistors. For instance, dual MOSFETs or quad MOSFETs are available that take up the space of only a single MOSFET, thereby saving a huge amount of space.
Use Compact Connectors?
Connectors can be rather space-hungry. The bigger ones also need higher-diameter pads to anchor them safely. Changing them to smaller types can be very useful, provided their voltage and current ranting permits. Manufacturers offer many types of miniature connectors, and selecting the right size can not only save board space but also cost.
It may save board space to use microcontrollers with built-in resistor networks for pull-up or pull-down for IO buses. Not only will this avoid using a number of discrete resistors, but also avoids using resistor networks, thereby saving a lot of board space.
The above points are some possible ways out for the reduction of PCB size for the designer to consider before starting the actual design. However, there will always be some crucial decision-related tradeoffs related to the cost and complexity versus the PCB size. Designers will need to select an independent path on the targeted application or for that specific targeted circuit design.
Considerations During Design
During product development, one of the major issues is design cost management. For designers, it is easy to manage the cost of the PCB by optimizing its size. Reducing the PCB size not only reduces the size of a product but also brings down production costs in most cases.
Component Package Selection
One of the major decisions during design is the selection of packages for components. The same component is available in different package sizes. For instance, an SMD resistor of 125 mW is available in sizes 0402, 0603, 0805, 1210, and more.
During prototyping, designers prefer using larger packages like 0802 or 1210 for the simple reason they are easier to handle, solder, replace, or test. However, this can end taking up a huge amount of board space. Therefore, replacing the larger package with a smaller one can help the designer shrink the board size.
One limitation to the smallest size to use will be the capability of the assembly machines, especially of the pick-and-place machine. Selecting a package size that the pick-and-place machine cannot handle is not very useful. But it is possible to edit the footprint of the component, within limits, to further shrink the space the component needs.
There can be other considerations as well. Selecting an ultra-small package, one the pick-and-place machine can handle, may allow the designer to increase the component density to a significant level. However, this may drastically reduce the space available for routing traces, and the designer may have to increase the number of layers.
Ultimately, the component package selection will be a compromise between the board size and the number of layers.
It may be necessary to use multiple layers to reduce the PCB size. This is because routing requires a huge amount of board space. Once the prototype is tested satisfactorily, the designer can shrink the size by using, say four layers, of which two inner layers are the ground and power layers. High current paths require wider traces and for thermal management, they must occupy the outermost layers.
The designer may add more layers to further reduce the board size, and these can be low current or signal-carrying traces in the internal layers. Beyond a certain level, however, there is not much benefit of size reduction by increasing the number of layers.
The number of layers is usually a tradeoff for the designer. Multilayer PCBs are generally more expensive compared to the prices of single- or two-layer boards. Typically, the decrease in the PCB cost due to size reduction is enough to offset the increase in its cost due to the addition of layers.
One way of handling traces carrying higher current is to place them on the outermost surfaces, as this helps them improve heat dissipation. Current flow through copper traces can heat them up due to their resistance, which increases when they are of smaller cross-section. As the copper thickness all over the board is the same, say 1 oz (35 µm), the only way to increase the cross-section of a current carrying track is to make it wider. But that may increase the size of the board.
An alternative way could be to use thicker copper of say, 2 oz (70 µm), for the track, which can allow it to reduce or retain the same width as earlier, helping to reduce the PCB size.
Using thicker copper also helps if it is necessary to add a heat sink to further augment the transfer and dissipation of heat. Designers typically use multiple thermal vias to thermally connect the heat sink on one side of the board to a hot component on the other side.
Most PCB manufacturers can easily handle tracks as wide as 4 mils. Designers can take advantage of this (after confirming with their PCB manufacturer) and make tracks, spaces, and vias as small as possible.
Designers can use the HDI or High-Density Interconnect technique to make their PCB as compact as possible, provided their board manufacturer has this capability. Furthermore, HDI techniques permit the use of microvias. These are minuscule vias drilled using lasers, and they take up very little board space.
Use Blind and Buried Vias
Routing a compact PCB with multiple layers can result in a high number of vias. This also happens if the board has multiple layers of signal. However, rather than use through-hole vias, the designer can resort to using blind and buried vias.
While through-hole vias cross the entire thickness of the board, they are not always necessary. Blind vias connect signals between two or more internal layers and are not visible beyond these layers. Buried vias can interconnect signals from the outermost layers to one or more inner layers.
The advantage of using blind and buried vias is they take up only as much board space as is necessary, and the balance is available for routing. HDI techniques allow the use of microvias, and designers can stack them or stagger them to make suitable blind and buried vias.
Fine-Tuning the Design
As reiterated earlier, it may be necessary to review the design several times to optimize and fine-tune it, with a view to reducing the PCB dimensions as far as possible. During these reviews, the designer has these options:
Always selecting the smallest footprint may not ultimately result in a smaller board size. This is because in high-density layouts, there is a limited amount of space available for routing. A small footprint may not allow routing space underneath it, and it may be worthwhile to change over to a more open type of footprint.
Remove Unnecessary Components
Component manufacturers may suggest using additional components for optimum performance. Designers must use their own judgment to retain these additional components. For instance, an MCU with a built-in ADC may have recommendations for external filter components. However, the filter components may not be necessary if the ADC is being used for a non-critical function, such as for monitoring the charge on a battery.
Removing such unnecessary components from the schematic can significantly reduce the board space requirement.
Start Routing with Ground Connections
When placing components, starting with the ground connections first ensures the designer always has the shortest path of least resistance to the ground.
Connectors mostly require strict and specific placements, as they need to mate with other connectors. Placing them first will allow the designer to see the space left over for placing other components.
Route Critical Signals Simultaneously
Designers must route all critical signals when they place the components. This will ensure all the vital connections that require special attention are routed the proper way.
When starting with manual routing, designers should preferably start with routing the wider traces and larger vias before others. This will make the routing simpler, and the designer can always make other traces and vias fit in by making them thinner and smaller.
Group Components by Function
Grouping components together by their function in the schematic makes it easier to route their connecting traces, which will also be much shorter. This not only results in better EMI/EMC, it also saves a lot of PCB space.
As most designers use EDA tools, it is recommended that the designer set up DRC or Design Rule Checking. This is a process to verify that the PCB layout meets the manufacturing and electrical rules the designer has defined. The software will compare the PCB layout with the defined design rules and flag any violations.
It is essential that designers review their designs several times to weed out any errors and to make improvements that could lead to a more compact PCB. This should include not only the layout but also the schematics.
Following the techniques suggested by Rush PCB Inc. should allow designers to create compact PCBs easily. Making changes during the design and prototype phase is far easier and cheaper than doing it once the device has been assembled.