F3’s customers are often companies who have (and want) little to no internal RF product development expertise or overhead because they don’t do enough wireless products to justify a dedicated team. There are a few companies we’ve helped build internal engineering organizations and more that we’ve augmented with RF consulting but in general our customers focus on their core business competencies. That’s why they come to us to handle creation of such a specialized device as a wireless connected intelligent device. Or “Thing”, as we’re now calling it in the “Internet of Things”.
The last post was about IoT device size and how cellular radio performance and antenna performance are basically constrained by North American carrier performance requirements. This of course is in direct opposition to end user expectations of “everything keeps getting smaller”. Now let’s say you have spent the time and tears to sort out a size compromise that keeps the cell carriers satisfied (note, I didn’t say passing) but still leaves you with a viable business case. Congratulations, you defeated that mini-boss. Now for the next major development foe, interference.
The single biggest pain point in radio product development is receiver interference. This is most common and hardest to fix in an existing design that either was never a wireless connected device in first place or a wireless device that for whatever reason was never subjected to radio performance design verification test. Basically the design is already painted into a corner, so getting out is going to involve much more pain than if wireless performance was considered in the initial design.
The receiver interference source you should be worried about is simply radiated noise coming from your local electronics. The microcontroller, any memory bus, PLLs for various peripherals, switching power supplies, really anything that oscillates or switches, creates RF emissions. The frequency at which they switch and the harmonics of those frequencies may be close in frequency to what your receiver is listening for . You’ll hear the terms self-quieting (which refers to this situation on an FM voice radio) self-interference or self-jamming.
It’s important to note that very experienced RF engineers spend a fair amount of time on this exact issue. This isn’t something that happens because you didn’t know what you were doing. This is something that is pretty much always a problem in every radio development. An expert only has two or three sources of interference in the design on the first prototype. The inexperienced will have ten or more.
What makes this particularly painful is that the interference sources often overlap so fixing one out of ten doesn’t necessarily show up as an incremental improvement, you often have to fix all the sources before you see results. This also means that fixing these issues is very time-consuming and unpredictable. It’s not uncommon to spend a man-month addressing this sort of thing even for an expert.
A common thing we hear is “Oh that won’t be a problem, our device passed FCC or ETSI testing” or more amusing, “oh it’s a pre-certified module”. This is a common misconception. First, when a non-radio device is tested, it’s tested to ensure it doesn’t generate interference of a level that would bother another device nearby. As in 1-3 meters away. When you put a radio in an embedded device, your radio is cm’s away from the rest of your electronics. That distance makes all the difference.
So, what can one do? One simple trick to avoiding this entire problem is use an external antenna that’s mounted at least 1m away from your electronics. The next best option for a quick fix is to put all your electronics inside a shielded box and your antenna outside the box. F3 can help you pick an appropriate antenna and verify your product has performance good enough for your market.
In many cases however the device in question is portable or has other physical or business use-case limitations where an external antenna would make the product unusable or unsellable. If the antenna has to be inside with the electronics, and you want to hear those really quiet signals coming from more than a few meters away, then those electronics need to shut the hell up and be quiet. There’s a long list of RF emissions mitigation techniques and best practices. F3 incorporates these methods into any design work we do and we often end up applying it after the fact to designs someone else did.
Many developers don’t actually do radiated receiver testing so they don’t even know there is a problem. That is, they build a product that has a key performance parameter that’s critical to the device’s market acceptance and commercial success and they don’t check in Design Verification Test to know how their design performs. They also often don’t check it in production to ensure their production units actually function. Common examples are BlueTooth and WiFi based products that use off the shelf or pre-certified radio modules. It’s also not uncommon in cellular devices designed for Europe and Asia where carriers don’t enforce minimum performance requirements.
In case it’s not obvious, when you do DVT, you should measure every performance factor in the device to ensure it meets basic market requirements. Bringing a new product to market that relies on WiFi and only finding out from your Amazon reviews that the “WiFi range was really bad, the other brand’s unit I have works much better” is not sound engineering or business strategy.
F3’s consulting services can help you address what radio performance metrics should be tested in DVT, how to test it, how good is good enough for your market and what to do if it’s not good enough. Using our production tester IP we develop in-circuit, and end of line functional test solutions so you know your shipping good product. One bad placement or wrong value of a part in your RF path and that good performance you worked so hard to get in the design is gone on the production units and unless you check it, you wont know till the bad reviews and returns start.
Good radio performance doesn’t just happen it’s complicated and requires attention. You have to test it when you do your design, regardless if it’s a pre-certified radio module. You also need to verify basic RF performance in production or you risk shipping bad product. If your core business isn’t making radio products, F3 can help you sort these issues out.
By Chris Anderson 3/15/18
link the blue mention of IoT device size to the blog post on device size
link any blue mention of F3 to http://www.f3eng.com/
link F3’s consulting services to http://www.f3eng.com/consulting-services.html