To have a solid chance at getting your new product developed and on the market, you need to simplify your product as much as possible.
If you try instead to develop your ideal product, including every possible useful feature, then you will find yourself working on your project for years with little progress.
In this article, you’re going to discover ways in which you can modify your product so that it drastically simplifies both the development process and the process of scaling the product from prototype to full manufacturing.
This article was originally published on PredictableDesigns.com. Download their free cheat sheet 15 Steps to Develop Your New Electronic Hardware Product.
NOTE: This is a long, very detailed article so here’s a free PDF version of it for easy reading and future reference.
The first product change to consider is to minimize the number of discrete devices. Let’s look at the example of a remote control car. In this case, you would have the car, which has a receiver in it, and then you have the remote control itself, which is a transmitter.
In your mind you’re seeing this as one product, and from the consumers perspective it is only one product. But from a development and manufacturing standpoint, the remote control car is really two separate products.
The car and the remote control will each need to have their own schematic diagram, printed circuit board, and firmware program. You will also have to develop two plastic enclosures, and each enclosure will eventually require its own custom mold for manufacturing.
The cost of your electrical certifications, such as FCC and UL certifications, will also be doubled since you have two, independent, electronic devices.
Everything is going to be multiplied by how many discrete devices that you have. This will also be significant when you go to scale your product from prototype to mass manufacturing.
Most of the scaling cost to bring a hardware product from prototype to mass manufacturing is related to the high pressure, injection molds required to manufacture a custom plastic enclosure at high volumes.
If you have two separate devices, each of those requires its own enclosure. Now you’ve doubled the number of molds that you need. This just doubles the financial barrier that injection molds present when you go to scale your product to manufacturing.
So back to our example, how can you reduce this remote control car so that it only requires one discreet device?
Instead of using a custom remote control transmitter, can you control the car using a smartphone or a tablet? That is one way to get the product down to only requiring one discrete device.
What if you are unable to minimize the number of discrete devices that your product requires as I suggested above?
The next best thing is to design a single PCB that can be used in both devices.
Once again, let’s use the example of the remote control car. The transmitter device and the receiver device have a lot of similar functions. Can you design a generic printed circuit board that can be easily modified to work for either the transmitter or the receiver?
There are three ways you can design your PCB so it can be used in either of your devices:
1) You can set up jumpers so you just flip a couple of jumpers and the board goes from being your transmitter board to your receiver board.
The advantage of this option is that both boards will be identical which simplifies development and manufacturing. The disadvantage is that each device will include unused components, thus increasing your manufacturing cost.
However, when bringing a new product to market you should always focus first on minimizing your development costs (and financial risk), not on maximizing your per unit profit. Maximizing profit should be a priority after the product proves to be a market success and your manufacturing volumes increase.
2) You can program each board differently as well. This is very similar to option one except instead of using jumpers to set the board’s function, you will use the firmware program.
This option has the same advantages and disadvantages as using jumpers, except that you of course get to eliminate the jumpers themselves.
3) You can do partial assemblies so only certain components get soldered onto the PCB for the different versions.
For example, in the case of our remote control car, your PCB may include a spot for a transmitter and another spot for a receiver. For the remote control itself you would populate the PCB to include the transmitter but not the receiver. For the car, you do just the opposite, populate the receiver, but not the transmitter.
The advantage of this option is that you don’t have unused components on your boards driving up your manufacturing cost. The disadvantage though is you will require two different assembly processes for these two boards.
Another way to modify your product to reduce its development cost is to minimize the number of custom plastic parts required.
A very minimal product, with just a simple plastic enclosure, will still require two custom pieces of plastic — a top side and bottom side. For most products, two pieces of plastic is going to be the minimum.
Your goal should be, if at all possible, to get your enclosure down to only two custom parts. That way, you’re only going to need two custom molds.
For instance, if your product consists of eight, custom-shaped plastic parts, then it will require eight different injection molds. That many molds will be prohibitively expensive.
Anything you can do to reduce the number of custom shaped, plastic components will be huge in reducing your cost to go from prototype to manufacturing.
One strategy to consider is using off-the-shelf parts when possible. The ideal situation, if appearance isn’t supercritical for your product, at least initially, is to go with a completely off the shelf enclosure. There are numerous websites and suppliers out there that sell all kinds of plastic enclosures.
This isn’t going to be ideal for a lot of products where appearance is really important, but it can be a way to simplify both the development and scaling costs for your product initially.
Eventually, you’re going to almost surely have your own custom enclosure to achieve the appearance you want. But at least consider this strategy initially.
But there are other little tricks that you can do.
For instance, you may design your product’s enclosure so the bottom side is a simple flat piece of plastic. This flat piece can possibly be purchased off-the-shelf. Also, it may be easily produced by simply cutting a sheet of plastic instead of custom molding. This would allow you to eliminate one mold.
Another example is if your product requires replaceable batteries, then that’s an extra piece of plastic because now you have a battery access door.
In this instance, one recommendation is to find a supplier that sells off-the-shelf battery access doors that have the general look and size of what you need. Then design your enclosure to work with this off-the-shelf battery access cover.
If you are trying to design a product that is as small as absolutely possible, be prepared for a ton of extra time and work, especially with the printed circuit board design.
It requires a lot more work to pack everything really tightly in a printed circuit board. It’s much easier just to loosely place things down without worrying about pushing the spacing down to the absolute minimum.
Going super small will require significantly more time to design a printed circuit board, which always equates to a higher cost.
The other, perhaps even bigger, disadvantage to making your circuit board as small as possible is that it complicates any changes you may need to make in the future. You’re almost surely going to have to tweak your product during the development process, and I’ve never seen a design that was perfect the first time.
Let’s say you need to add another microchip to your design. If everything is packed in really tightly on your PCB, this becomes a major chore because you have to reshuffle the entire design just to fit in another component.
Another disadvantage of pushing the size to the absolute minimum is that it may require a more advanced PCB manufacturing process, which adds to the board cost during the assembly process.
For instance, a PCB fabrication process that allows a minimum trace spacing of 4 mils will be more expensive than one that supports spacing only down to 8 mils.
There are even more advanced features, like blind and buried vias, which can drastically reduce the size of a printed circuit board. But they also significantly increase the cost of the board, especially in the prototype stages and at low volumes, so they are rarely recommended.
Finally, when you are designing a PCB to be as small as possible you will almost always want components placed on both sides of the board. But that increases the manufacturing cost. If you can settle for a larger board, then you can use a single-sided board which costs less.
If your product can be designed to use a microcontroller instead of a microprocessor it will greatly simplify the electronics design.
Both a microprocessor and a microcontroller require a processor core, RAM memory for temporary storage, and Flash memory for program storage.
In the case of a microcontroller, all of this memory is embedded onto a single chip. Whereas with a microprocessor, most of this memory is external to the processor chip.
The interface between processor and memory, especially high-speed RAM memory, adds a lot of complexity and potential design risk.
By having all of the memory embedded on the same processor chip, a microcontroller simplifies the PCB design process. Using a microcontroller will save you development time, which ultimately saves you money and gets your product to market quicker.
Also, most microprocessors are more expensive than microcontrollers. “Over-designing” your product to unnecessarily need a microprocessor will add significantly to your unit manufacturing cost.
Finally, consider that a microprocessor is a lot more power hungry than a microcontroller. If your product is battery-operated, then you should use a microcontroller if at all possible.
You should be able to get by using a microcontroller, unless your product requires an operating system, USB 3.0, gigabit Ethernet, 1080p video, or exceptionally complex processing capabilities.
If your product requires any wireless functionality, whether that be Bluetooth, WiFi, cellular, or LoRa, always start with a pre-certified wireless module. Rarely does it make sense to design your own, custom, chip-based wireless circuit.
There are two primary reasons why. First of all, RF (radio-frequency) design is complicated- really complicated. It tends to be one of the most complicated types of electronic circuits to design, with a lot of associated design risk and design cost.
If you buy a pre-certified, wireless module its already been tested and all the bugs have been worked out, which drastically reduces your development cost and time.
Secondly, wireless modules also reduce your certification costs. If you do a custom, chip-based, wireless design, you’re going to require (in the US) an intentional radiator FCC certification. This costs well over $10,000 dollars.
On the other hand, if you use a pre-certified module, then your product will only require unintentional radiator FCC certification which costs only a couple thousand dollars.
For most products it doesn’t make sense from a profit margin standpoint, to do a custom wireless design until you get up to producing a few hundred thousand units. Only then, consider transitioning over to a custom, wireless design that will improve your profit margins.
You have to be producing a high number of units before any cost savings will compensate for the additional certification and development costs required.
Avoid embedding any custom, high-voltage, AC electronics in your product. If possible, you don’t want any component in your product to ever “touch” an AC voltage line.
Because of safety issues, you need to have UL certification if your product connects into any household AC power. You don’t want your product plugging into an AC outlet and causing a house fire.
Fortunately, most products don’t actually require AC power, instead they simply convert AC voltage into a DC voltage. This DC voltage is then used to either power the electronics or to recharge a battery.
If you do this AC to DC conversion inside your product, then you’re going to require a UL certification that costs around $10,000.
One way around this UL certification requirement is to use a pre-certified AC-to-DC adapter. This is similar to my suggestion to use pre-certified wireless modules to avoid the higher FCC certification fee.
For instance, let’s say your battery-operated product will recharge via a USB port. In that case, I would recommend that you instead simply purchase an off-the-shelf USB charger that already has the required UL certification. Then include this charger with your product.
Pre-certified USB chargers can be purchased from various suppliers on Alibaba.com for only a couple dollars each.
If your product requires a mobile application, then start with either Android or iOS, but not both initially.
You don’t want to be in a situation where you’re trying to test and debug two different mobile apps. That’s going to be a challenge to manage. Every time you find a problem in one, you’ve got to go back and make the change in the other version.
Mobile apps are not cheap to develop, and if try to develop both together, there’s not a whole lot of that you can reuse between the two. This means you will almost double your mobile app development cost if you target both operating systems.
I recommend that you choose just one operating system initially. Start with the one you believe your user base uses the most.
For example, start with an Android app, get that out of the market, get some testing done, at least some early adopter testing.
Once you have all the bugs worked out on Android, begin to develop the mobile app for the iPhone. This will drastically simplify the development process for the mobile app and allows you to get the product to market faster.
If your product incorporates video I recommend that you start with a resolution no higher than 720p HD.
720p HD is sufficient for most products initially, and it will simplify your product design’s complexity compared to 1080p HD video.
720p HD video has a screen resolution of 1280 x 720, whereas 1080p has a resolution of 1920×1080. This means that 1080p has 2.25x as many pixels as 720p, so it requires 2.25 times the processing speed.
The main reason that 720p simplifies a design compared to 1080p is that there are microcontrollers available that can handle 720p video at up to 30 frames per second.
For example, the STM32F4 series of microcontrollers supports a camera interface up to 54 MB/second. That’s fast enough for a 720p camera, but not 1080p.
In order to incorporate 1080p video you will need to use either a fast microprocessor, or a specialized video processor chip.
If you’re on the fence about using 720p or 1080p, then definitely start with 720p since it’s going to simplify things. You can always come out later with a second version that offers 1080p, if your marketing evidence shows it to be a desired function.
These companies may offer support forums where their application engineers answer your questions. Or they may provide support contact information.
Their products are also carried by various distributors so you can easily get samples and order small quantities.
They also openly share all of the technical details for their products. For example, you can easily download component datasheets, and many times even access reference schematics and PCB layouts.
Unfortunately, not all electronic component manufacturers are so friendly with small businesses.
Qualcomm, for instance, is one of the most well-known, worst companies for startups to try to work with. You can pretty much count on getting zero support from them.
They likely won’t even allow you to view their datasheets. Everything is proprietary. To get any information, they’re going to want a sales forecast from you, and lots of other information that you likely won’t have easily on hand.
Without proving to them you are worth their effort and time, you likely won’t even be able to get samples.
Avoid these types of companies like the plague. You need to stick with companies that are more open with all their solutions, and that offer meaningful technical support.
Developing a hardware product is extremely challenging. To have a chance at success you need to spend considerable energy simplifying your product definition as much as possible.
In this article I’ve shared with you ten product simplifications that can profoundly reduce the complexity, time, and cost to develop your new electronic product.
Originally published at predictabledesigns.com.