It is pretty hot outside welcome back to The Factory today we're going to talk about the PiicoDev transceiver.
While we were working on the PiicoDev servo driver that was the project that we covered in that seven-part series design a product with us, we were working on this quietly in the background and I actually think this could be more exciting.
This PiicoDev transceiver is essentially a data radio you can send data between points and that's just such an upgrade for projects now you're talking Outdoor weather stations remote control robotics all it takes is to get messages from one point to another.
So we're going to talk about how we arrive at this solution in this week's factory let's jump into it.
PiicoDev transceiver is based on this hope RF RFN 69 radio this is a 915 megahertz module and of course, it's SPI, so it's going to be a smart module we're going to need a microcontroller on board so that we can translate commands on an I2C bus to this device.
And that's actually where a lot of the development for this project went into getting a positive user experience in python so that we can interface with this module with the translator in between, and you don't even know it.
Because we're working with a radio module the circuit design is actually pretty simple like this is a module that's designed to just be placed into an existing project like a daughter board, and it just takes care of itself.
So there's minimal circuitry to design the real challenge is in the antenna you're Probably used to seeing these beautiful little PCB antennas by now.
This antenna is for a 2.4 gigahertz radio on a micro:bit, and it is just so tiny beautiful you'll see that on ESP32 modules as well which are Wi-Fi boards we're working in the 915 megahertz band lower frequency longer antenna and the solution he had us stumped for a little while until we found this really cool document from Texas Instruments.
If you're doing a wireless project this is a interesting place to start to get some Inspiration for what you might want to use, so we started off looking at this document and there are some pretty familiar patterns that emerge.
If we focus on this middle row these are all the antennas that focus on the 915 megahertz band and there's some pretty familiar shapes if you've been working with maker modules before.
You can see here we have this helical PCB trace antenna, so this is using PCB traces to spiral through a circuit board we have this flat zigzag design and then a few others More esoteric ones. Look we've even got this coil of wire that's just soldered straight to the board very simple but probably very effective, so the requirement for this project is that it's going to fit on a PiicoDev module and when you look at a couple of these reference designs you start to see some troubling news for that design requirement. Here we've got 38 by 25 millimeters now these boards are only 25 millimeters along the short edge so yeah look maybe a helical trace antenna is looking pretty Attractive for that solution but you see that a lot of these just really break the space budget you know 43 millimeters we kind of joked around with hey maybe this is the first PiicoDev module that uses a two by two unit design, so it would be twice as big as this module this zigzag design reference is so wide maybe we have to make some kind of hammerhead shaped module just to fit the antenna in the top half.
You know honestly once we opened up these app notes and delved into the specifics of each Design where you can see you know calculations and layout guidelines is that zigzag antenna one, and you know this one requires a very thin impedance controlled board this one was just going to be way too big for PiicoDev things were starting to look a little bit bleak on the antenna front we didn't quite mind if we had to sacrifice a bit of range if it meant that on board on the module there was no need to plug in an external antenna like you're a SMA connector or a UFL connector we really wanted To just have an out of box experience with nothing dangling off.
And when things were starting to look a little bit bleak, and maybe we have to compromise that design requirement we came across a very interesting part.
This chip antenna by TDK essentially shortens the PCB trace length that you need to include on the board I know there's some RF wizards out there that'll probably drag me over the coals for simplifying it like that, but that's the simplest way that I can figure out how to put It. 
This chip antenna by some combination of internal network components like inductors capacitors seems to effectively shorten the PCB trace that you need to get to the same wavelength antenna.
In this evaluation design you can see it only needs 20 millimeters by 10 millimeters and a couple of supporting components and so that's exactly what we did we followed this design guide our board is roughly the same aspect as this reference design and that 20 millimeters is perfect for fitting across the Short edge of the PiicoDev module.
So we've got a couple of these chip antennas we put them on a prototype, and it just worked we even did some empirical testing with changing these network components essentially the small matching circuit of inductors or capacitors and so that it didn't really make a difference for values that we picked out of these tables.
Now it's very unlikely that an onboard chip antenna is going to perform quite as well as something like an external antenna I mean just look at The size difference but here the requirement is the ease of user experience out of the bag and no need for external hardware no dangly antenna that may break off or go missing if it's a detachable antenna.
So now I've got a couple of these modules together I figured it would only be appropriate to do some real in the field range testing let's head outside.
So we have the transmitter the atmospheric sensor and that goes all the way down. There's our Pico taped to the mast and then down the bottom we Have USB battery bank for power. Here's our transmitter, and you can just see the TX light blinking there every now and again those are our weather messages and there's our what the data's streaming in every couple of seconds.
And of course we're seeing that on the display as well. Here is our receiver we have just a heads up display because that's part of the project that goes down to the Pico, and we're plugged into the laptop there's the transmitter in the background got the laptop, and I'm just going to go for a walk down the road here.
Still getting data I reckon where I don't know that looks to be between 40 and 50 meters keep going down the road.
I know about halfway down the block probably getting up to 70 meters or so started getting some payload errors just as I passed these large vertical poles I wonder if they have anything to do with it.
Seems to be doing all right still okay I've gone a little further along, and now we're getting no payloads.
I started getting errors at that telegraph pole And yeah now we're getting nothing so let's head back.
And we've been down we've been down that road and that is aligned with the face of the transceiver you can see it's facing right down that road now we're going to go up this street what's that south street Adamstown and see how far we can get.
And it looks like we're getting about one in every two messages or thereabouts, so we haven't gotten quite as far in the off axis direction I wonder how far we can push this out yeah see we're still seeing a couple more errors. As you can see we're still getting full packets, so we're just working in a slightly degraded state okay let's head back.
I've only taken a few steps and where we're back on it's quite interesting I've updated the over the air speed to be quite a bit slower to see if we can get more range, and I've also decreased the loop time so that we can see data coming through more frequently.
That should help us identify when we're hitting the range limit, and so I'll just do those same two tests again see how far we can get.
Okay we are where the test stopped before, and I still haven't seen a single error, so we've made it to the same range, but it's there's just no question about it this time the data is just coming through beautifully let's continue.
Oh we just got our first error that's it interesting a marginal increase maybe an additional between 10 and 20 meters. All right now the other way.
So where are we? We're right outside the studio I've come across the road because there's a lot of crud between me and my transmitter and the signal has come really good again, so I wonder if we had done that at the higher speed test whether it would have worked but let's just see how far we can get on the low speed. 
Okay we still have data coming through, and I'm just across the road, so I guess we've got like another 20 meters out of that change it's really hard to tell in this environment there are a lot of cars there's a lot of infrastructure this is I would say a pretty fair real world Test.
As with RF it's not very surprising I'll be walking down the street, and I'll see a problem the payload will be dropped you know we'll have hit the range, but then I turn around removing myself from between the receiver and the transmitter and maybe like this phone and the technology and the signal comes good again, but I'll say that's a practical maximum.
We didn't get super far down Werris street let's go around and do the other test.
Just passing the transmitter again go on the same side Of the road that seems fair.
Oh, yeah it's still coming in we're a fair way away that's the intersection. It seems very dependent now on my orientation. You look at this no payloads coming in and if I turn we start getting some partial payloads I guess we're behind you know we're behind some pretty, pretty heavy metal shielding.
And now I've got data coming back in just as I just as I cleared this vehicle we've got data coming back in.
There it is the difference a few steps makes. RF is weird.
Back in The comfort of the AC let's take a look at how far we were able to transmit. My transmitter was under these trees here, and I got down to just past this intersection around here and that is 125 meters and when I went up Moira street I only got to about the second or third property that's about the same 125 meters so that's that's promising because there's quite a lot of obstacles now being introduced when I went around the corner here.
Likewise going down south street getting to about between 125 and 130 meters as well look at that, so it would be pretty fair to say that in a real field test 100 meters is a pretty safe range to go on.
I was saying acceptable performance well over 100 meters this is like a real world field test we're not just in an open field we're actually in the suburbs there's telegraph polls and power lines and trees and cars, so I think that's a that's a pretty fair suburban test.
What about indoor ranges well the farthest two points of this building are about 35 meters apart So I set up the transmitter right here at my desk here it is and took the laptop all the way down the other end of the building to the Makerverse studio and as you can probably see there's a lot of metal shelving there's partitions and walls
Yo as you can see I'm in the Makerverse studio, and we are still getting data how good and pacing that out that was about 35 meters.
35 meters indoors in a pretty hostile environment I'm very pleased with that, and we've just had a bunch of fun messing around With these transceivers you know a transceiver by itself isn't all that exciting it's not really it's kind of hard to explain you know what it does and why that matters but the moment you like hand make a remote control, and you can see something else spring to life with movement that's when you can really see the value in something like a transceiver they just take projects to another level.
And here's what's happening behind the scenes in that field test we're doing all that normal PiicoDev setup And then to send the weather data from the transmitter to the receiver we're just calling this send method and send will accept a tuple this is what we're calling labeled data in this way we can send a label which is so many characters long and the value itself which is in this case a floating point number and that's really useful because we can call radio.send three separate times and because we're labeling the data the receiver can catch those messages and sort the data into the appropriate spot.
For example so you don't mix up temperature and humidity when you're trying to display it on your screen, and so we can see that on the receiver end when we receive a message we can just check the first element of that tuple for the label that we care about so in this case if we get the label temperature then we store the value from that tuple into the temperature variable and in this way we just have a couple of if statements each sorting the data that's coming.
In any case that's all I have for You in this episode of The Factory keep an eye out for these on Core Electronics because they are bound to be released very soon we're just tying a bow on the educational content for these can't wait to see what you guys make with the PiicoDev transceiver and until next time thanks for watching.