This week, we sat down with Richard, our Managing Director, to discuss electronic prototyping – and the key differences between an A model and a final product.
What Is Electronic Prototyping?
Prototyping is making a low number of first-off samples for the purpose of testing. This way, you can evaluate the design to make sure it works.
If your prototype is electronics-based, you may subject it to a range of tests. However, if it’s more mechanical, then you can test it out and perform a few stress tests.
Ultimately, it’s a first-time attempt at creating a solution.
What Are the Different Stages of Electronic Prototyping?
For prototyping, we typically create an A model, a B model, and – depending on the complexity of the project – a C Model. From here, we’ll move onto production models of units.
It’s worth acknowledging that the A model will be wrong. The A model is the first time you bring all the design effort together. You are integrating different people’s pieces of IP onto the board, and each person’s IP will typically come with different data sheets.
The data sheets could be wrong, or they may differ in your application compared to others. There could be a specific footnote on page 173 that’s specifically relevant to your application but not for others.
And so, you stitch these devices together onto a printed circuit board, manufacture a few prototypes, and then test them.
We always advise to only make a handful of prototypes. Do not, under any circumstances, make hundreds of these. Even if your project is late. They will be wrong. You will end up having to modify them. There will, inevitably, be some sort of a problem.
You want those prototype boards in the hands of the engineers as fast as possible so they can test them and then work out design solutions. If there’s any strange behaviour that’s unexpected, it needs to be understood because it will come back to haunt you at some point during the development process!
When every piece of functionality has been tested on the A model, you can then move over to a B model, which incorporates all the changes that you’ve made, all the problems that you’ve identified, and gives you a far more solid base.
At that point, you might wish to make additional prototypes to potentially get into the hands of users, and certainly for your development teams.
From here, you’ll be looking at getting your prototype into scale-up and production. You may wish to produce some new product introduction models. Whilst you’ve got a B model that’s been designed for production in mind, your manufacturer might need some tweaks and adjustments to be able to make sure they can produce the product effectively.
How to Design a Prototype Demonstrator Model
For a prototype demonstrator model, you need to work out what it is you are wanting to demonstrate, and why.
Do you want to prove the functionality works? Or are you looking to go down the MVP (minimum viable product) route where you want feedback from users? Regardless of your intentions, these questions need to be considered to ensure you’re doing things for the right reason.
Ultimately, you need to nail down what the specific purpose of your demonstrator model is and do that. Don’t try to do everything in one go.
Why You Shouldn’t Put an A Model into Production
In reality, A model prototypes are highly likely to have something wrong with them.
For example, perhaps the specification has changed. You might realise this product iteration isn’t want you really want. This realisation can come after you’ve shown it to a user or made the prototype and got it on the workbench.
But you also need to consider that there will be changes in the electronics themselves.
Device values will be amended between the A model and the production model, printed circuit board tracks may also be altered, and you may even make changes to the device itself. You need to consider the ongoing global semiconductor shortage which could force you to design one chip out and another one in, depending on availability.
Because of this, ordering a high volume of A model prototypes for the purpose of production can be incredibly costly.
Key Differences Between Electronic Prototyping and Scale-Up
Prototyping proves that you can make a few models that demonstrate the right functionality and the right behaviours in a very benign environment. They’re usually tested in a lab environment where it’s likely to be air conditioned. Testing often takes place with users who know how the product should be used or have worked out what the use cases are which adds an unintended bias to the way the product is tested.
There’s a significant difference between having a handful of prototype demonstrators and scaling up to high volume manufacturing. At the manufacturing stage, you must have resolved problems associated with temperature, variations, tolerances, and usability.
Somebody, without prior knowledge of use cases, will do something that you never thought they would, and that will break your device, “That’s often me! If someone gives me a product to test, I can pick it up and break it in a few minutes!”
The key point here is that the test and validation part, the scale up stage, includes compliance and quality control considerations. Getting the peace of mind that every batch off the production line will be okay is vital to prevent a bunch of product recalls and customer complaints.
Once you’ve tested the product yourself, you’ve tested it over the range of temperatures that it’s advertised to perform at, and it has been tested with end users, then you’ve got a product that you can consider shipping.
Skipping this crucial stage will create a whole world of pain.
Testing Your Device Against Temperature and Humidity
You have a few options for testing temperature.
You can either take the device to every country and environment, throughout the seasons, that you might want the device to perform in.
Or you can hope for the best and throw it into production, looking at the specs of each component and hoping that it works in practice. Or – and this is the right option – you can do temperature control testing in an environmental chamber.
An environmental chamber allows you to test and control both the humidity and the temperature over a range. Verifying each of the essential parameters builds up a lot of confidence.
Fletcher’s Law of Prototyping
Fletcher’s Law is a little bit tongue-in-cheek, but essentially it says that the higher the number of prototypes initially made, the worse the product and those prototypes will be.
This is usually down to cultural or environmental pressures. This is usually the result of businesses compressing the product development process to get a product out of the door ASAP. In short, a culture of trying to do as much as possible, in as short a time as possible, to meet a fixed launch date.
And sure, the product might be launched on time, but it usually means that the design, the prototyping and the testing have all been rushed. Because of this, crucial steps such as a design review, may have also been skipped to save time.
Because of this, you’re saddled with poor-quality prototypes, and all these prototypes now need to be hand-modified, adding and removing components, cutting PCB tracks, soldering, and more.
This means that your engineering team ends up wasting time modifying boards that go out for test, instead of working on testing, and getting your prototype closer to a B model standard that would take your product to market in a reliable fashion faster.
Fletcher’s Law is an acknowledgement that by rushing you actually go slower. You spend more money, and that whilst you feel like you’re making progress, you’re actually just wasting time and money.
Electronics prototyping is not like writing code. You can’t just compile it and get away with it. There is a lead time to get components, and currently the lead time is extensive due to the global chip shortage.
Next you’ve got to manufacture and test them. So, every time you do a PCB prototype, you can lose up to 2, 4, 6, or even 8 weeks. You are better off focusing on getting the quality right on the first run, and then scale up from there.
Why Choose Ignys for Electronic Prototyping?
We can help with the electronic prototyping stage in a range of different ways.
Ideally, the best place to start is right at the beginning of a project, where you have an idea and considerations on how to deliver it. At this point we can help inform design choices, architectures, device choices, and then work out the risks, the feasibility, and roadmap the process.
However, we are not precious about which part of the process you’re at. Ultimately, we have experienced hardware, firmware, and software engineers with the expertise to be able to support you – no matter which stage of the journey you’re at.
With new projects, we typically run an idea generation workshop before moving onto the feasibility stage where we’ll assess challenges around power consumption, radio, and thermal performance speed. We then take you through various parts of electronics design to get you to that prototype demonstrator.
With software prototyping, you may have a software or hardware team already in place. Regardless, we can use either of these skillsets to fit around your team. And if you don’t have in-house capability in either area, we can help ensure that the necessary due diligence has been followed, including regression testing of software, and make sure things are tested as thoroughly as possible.
You can test forever with diminishing returns, but there are key parts of testing that can’t be avoided, for example there’s a level of testing against use cases that you must do such as supporting with compliance through EMC and the radio equipment directive.
Moving into production, we look at quality control using test fixtures, test plans and, of course, our SAM subscription for post-launch support. This means that you can get support for the products that you’ve designed on an ongoing basis.
Often once you get the product launched, you’ll find that there may be extra features you need, or there’s a bug that’s found, or maybe there’s a chip that you need an alternative for. SAM helps ensure that you are not waiting for our typical lead times and that you have rapid access to our engineering service.
Ultimately, a lot of customers come to us later in the process with projects that are running late or need expert support. We’re more than capable of supporting those customers, and rescuing them from difficult positions, but it’s far easier for all parties to begin at the start!
Authorship
This blog was written by Richard Fletcher, Managing Director of Ignys.
Watch Richard discuss electronic prototyping in the video version of this article is available on our YouTube channel.