A Cloned Bluetooth Tracker Meets Its Maker

The holidays bring us many things. Family and friends are a given, as is the grand meal in which we invariably overindulge. It’s a chance for decades old songs and movies to somehow manage to bubble back up to the surface, and occasionally a little goodwill even slips in here or there. But perhaps above all, the holidays are a time for every retailer to stock themselves to the rafters with stuff. Do you need it? No. Do they want it? No. But it’s there on display anyway, and you’re almost certainly going to buy it.

Which is precisely how I came to purchase a two pack of Bluetooth Low Energy (BLE) “trackers” for the princely sum of $10 USD. I didn’t expect much out of them for $5 each, but as this seemed an exceptionally low price for such technology in a brick and mortar store, I couldn’t resist. Plus there was something familiar about the look of the tracker that I couldn’t quite put my finger on while I was still in the store.

That vague feeling of recollection sent me digging through my parts bin as soon as I got home, convinced that I had seen something among the detritus that reminded me of my latest prize. Sure enough, I found a “Cube” Bluetooth tracker which, ironically, I had received as a Christmas gift some years ago. Putting them side by side, it was clear that the design of these “itek” trackers took more than a little inspiration from the better known (and five times as expensive) product.

The Cube was a bit thicker, but otherwise the shape, size, and even button placement on the itek was nearly identical. Reading through their respective manuals, the capabilities also seemed in perfect parity, down to being able to use the button on the device as a remote camera control for your smartphone. Which got me thinking: just how similar would these two devices be internally? Clearly they looked and functioned the same, but would they be built the same as well? They would have to cut costs somewhere.

Determined to find out how a company can put out what for all the world looks like a mirror image of a competitor’s device while undercutting them by such a large margin, I cracked both trackers open to get a bit more familiar with what makes them tick. What I found on closer inspection of these two similar gadgets is perhaps best summarized by that age old cautionary adage: “Don’t judge a book by its cover.”

Cutting Corners

Internally, both trackers look very similar at first glance. Not to say they actually share a PCB design, but that all the “landmarks” as it were are pretty much the same. You’ve got a single SMD LED, a small tactile switch, a CR2032 battery, a speaker, and of course the Bluetooth chipset. There are some readily apparent differences in the Bluetooth hardware itself, but we’ll get to that in a moment.

Ignoring for the moment the PCB layout, the most obvious difference is the case itself. The Cube case is thick and substantial, with a very nice rubberized coating that reminds me of the lid of a Thinkpad. In comparison, the itek tracker feels like it’s made out of material that’s only a step or two above those plastic Easter eggs. The PCB snaps securely into the case of the Cube, whereas on the itek it’s simply sitting lose inside the case with nothing to hold it down other than the lid itself. There’s also a very nice seal on the Cube to keep dust and moisture out, while the only thing “protecting” the itek is the fact that it comes with a backup for when the first one breaks.

The Cube’s built-in Piezo speaker

Clearly a lot of money was saved just on the case, both in terms of the engineering and the raw materials. But the cost savings go deeper than that. The Cube has a much more substantial battery holder, while the itek simply has a couple of tabs soldered to the board. The Cube also uses a much larger piezoelectric speaker built into the back of the device to make its chirps and beeps, complete with tabs on the rear of the PCB to interface with it. In comparison, the itek tracker has a small PCB mount speaker.

In terms of material cost its hard to say which one was actually the cheaper option, though clearly there’s less assembly and engineering costs associated with the direct PCB mount speaker than the two piece design used in the Cube. This same engineering “easy route” can be seen elsewhere in the cheaper device’s design as well, such as in the PCB antenna.

The itek uses the standard type of 2.4 GHz PCB antenna we’ve all seen on Bluetooth and WiFi gadgets; a small tinned “square wave” built into the board. The Cube on the other hand is using a rather unique approach, a thin strip that goes approximately 1/4 the way around the PCB which has been peppered with holes. There’s nothing inherently wrong with the antenna on the itek, and conversely I’m not completely clear on why Cube decided to go with such an elaborate antenna on their board, but clearly the former would have been the easier and cheaper option.

A Tale of Two Chips

It’s easy enough to compare between a thick waterproofed case and a flimsy one, or see how the manufacturer saved a few pennies by skipping a fully enclosed battery holder. But under the microscope we find that the changes go even deeper than that. While the Bluetooth functionality of both trackers is effectively the same from an end-user standpoint, the hardware used to deliver that experience is quite different.

nRF51822 SoC inside the Cube tracker

Inside the Cube is the well known Nordic Semiconductor nRF51822 Bluetooth system on chip (SoC), specifically the QFAB variant which features a 32-bit ARM Cortex processor at 16 MHz with 16 kB of RAM and an integrated 128 kB of flash storage. We’ve seen this chip used in dozens of projects here at Hackaday because it’s capable and very well supported. Seeing it in a commercial product like this is no surprise at all. In fact, it might have been more surprising if it wasn’t using some member of the nRF family of chips.

That said, the nRF51822 isn’t exactly cheap. A quick check of a few online suppliers shows prices of nearly $2 USD per chip, even with quantities in the thousands. Not a problem for a Bluetooth tracker that retails for $25, but far too expensive if you’re trying to sell the things for $5 and still make a profit.

Somewhat ironically, the itek actually uses two parts to get the cost down: a YiCHiP YC1034 Bluetooth chip and a 64 kB I2C EEPROM to hold its firmware. I wasn’t able to find much information about the YC1034 online, but a listing for the presumably closely related YC1033 on Alibaba comes up with a single unit cost of only $1.5, compared to the $2 volume price of the nRF51822. In volume, the price of the YiCHiP is likely to be pennies. Even with the additional cost of the EEPROM to support it, it will end up costing significantly less than the all-in-one solution from Nordic Semiconductor.

Putting the Money Where it Counts

While designing the itek tracker, it’s clear many design decisions were made on a cost and complexity basis. If it was easier or cheaper, that’s what they went with. On the other hand, with the Cube usability and reliability were more often the prime motivators. Looking at the materials and components used in the construction of these two very similar gadgets, we can very confidently say you aren’t “paying for the name” by purchasing the Cube over its fractionally priced doppelganger. It’s simply a far better made piece of gear.

That being said, not all the decisions are easily so vilified. A flimsy case with no seal has obvious real-world downsides, but if you can replicate the user experience with electronic components that cost less than half of what your competitor is using, is that really such a bad thing? Does the Cube really need such a powerful SoC to get the job done? It wouldn’t be the first time we’d seen a seemingly overpowered Bluetooth beacon.

A close examination of a cloned product and the original that inspired it can be an illuminating look at not only where the money is best spent during manufacturing, but where the design could stand to be optimized a bit. In this case it seems that a tracker with the body of the Cube and the electronics of the itek would be an ideal solution: put the expense in the physical aspects of the device where it’s needed the most, and cut the costs where the user experience won’t suffer.

28 thoughts on “A Cloned Bluetooth Tracker Meets Its Maker

        1. The story of the ‘original’ phone tracker is a bit more complicated than that, and it actually starts with TrackR, which was started by a couple of UCSB students who lost their keys while surfing. Their first product, the ‘Phone Halo’, was successfully crowdfunded in 2010 and later turned into a successful company.

          About 2 years later the Tile was developed by some Silicon Valley types who integrated newer technology (most importantly, bluetooth low energy came into existence which greatly reduced the power requirements), social capabilities (the tile could be tracked by any phone, in theory increasing the chance of you finding it) and with it the usual Silicon Valley mindset (non replaceable battery, etc). Due to the technological improvements the device was much smaller and a fair bit more convenient, but was really just the logical evolution of the Phone Halo, and TrackR was also bringing their next generation device to market right around with Tile came out. It was also successfully crowdfunded and later turned into a successful company as well.

          I stopped keeping track after Tile came out, but there are now dozens of essentially identical devices of varying quality available.

          You can argue till the cows come home as to what is the ‘knockoff’ or ‘competition’ or whatever, honestly the idea of a phone app that monitors for the presence of a bluetooth device and records the phone location and/or beeps when they are separated is not exactly a revolutionary idea. In fact the Tile and its brothers are essentially a BLE beacon dev kit ex https://rf-design.co.za/product/em-microelectronic-proximity-beacon-development-kit-includes-5-beacons/

        2. It’s hard to believe your serious here, but this itek thing is clearly made to be a replica of the CUBE. It’s nearly identical beyond being a different color, and even duplicates the features like using the button as a camera control.

          Comparing the CUBE to the Tile you can see they accomplish more or less the same goals, but neither of them is made to be a replica of the other. It’s no more a knock-off than a Ford is a knock off of a GM; they are both just cars.

          1. I’ve literally never heard of, nor seen, the cube. Meanwhile I’ve seen Tile everywhere. Clearly the iTek is a ripoff of Tile, since they look exactly the same and do exactly the same thing…

    1. Considering the name, I suspect the “iTag” is the more direct inspiration. Not sure if they pre- or post-date the Tile, I’ve gotten three (and put one to actual use – child with a a habit of forgetting either a binder or a cell phone – problem surface shrunk to “remember one or the other, and whichever you have will alert you”) and opened one. It was nRF51822 based, even had some test pads and a labeled Rx/Tx pair (which makes me suspect it may be a chineese knockoff – “brand” products rarely roll out the red carpet like that) though I didn’t get any further – lost focus an buried myself in other nRF stuff whenever trying.

      1. Yeah, I went with some TrackRs because of the replaceable battery… but Tile now has two models with replaceable batteries.

        I don’t have any hard data but I suspect the Tiles have better crowd tracking coverage. One item I have with a TrackR doesn’t get spotted very often.

  1. That piezo molded into the case is bad news!

    I’ve had 3 different generations of Tiles. They were all built that way too. The problem is that once you stick it in your pocket with a bunch of keys the case eventually deforms some microscopic amount and the fingers inside lose contact with the Piezo element. After that if you pick it up and play with it a little they will regain contact and the tune will play but when it’s time to leave, your keys are lost and you need the Tile to play you can forget about it!

    1. Does anyone knows how the piezo is driven? (on the model with the real piezo, not the magnetic sounder)
      Directly from the MCU or with any voltage elevating clever system? If so, what is the volume level?

      1. In the picture with the two open devices, on the cube (original) on the right, you can see a blue can attached to the PCB. I’m reasonably convinced that is an inductor wired in parallel with the piezo. If you apply a squarewave with for example a transistor, every time it switches off the inductor will produce a voltage spike higher than the supply voltage, increasing the volume of the piezo speaker.

    1. My exact thought. It’s nice to theory craft but he currently posses both products.

      Here are some really cheap and simple tests I would have loved to see (even if not completely scientific in determining overall fact)
      1. Range (m)
      2. Volume (dB)
      3. Setup (app user score?)

      There are so many things that could have been tested to answer the questions left at the end of the article. Is the unique antenna better and if so by how much?

      1. I don’t think the idea was to review the product here. We’ve got Amazon reviews and such for that kind of info. The point was to see what was inside of them, and how the cost of construction differed.

        Even if the electronics in the cheaper one are as good as the more expensive it doesn’t really matter since the case is crap and there won’t be enough of a userbase to make the product work as well as the competition. Remember these systems use other phones running the app as a way to locate the tags remotely.

        So the end part isn’t saying you should buy the cheaper one, but maybe the CUBE doesn’t need to be quite as expensive either. Though they aren’t making the thing anymore it sounds like, so doesn’t really matter in the long run.

  2. You can’t really assert that cutting costs on the electronics side doesn’t affect the user experience without testing whether it actually affects the user experience. For instance, it’s not great if it cuts the effective range in half, or runs the batteries down more quickly, or if it simply doesn’t work when you need it.

    1. Yes, that rubberization normally does no good. On the electric toothbrush it starts to get sticky, on a small (cheap) nose hair cutter it got so sticky, you could not put the thing down any more :-) OK, you could say: throw away a $3 gadget after some years, but why should it be like that? I covered it in talcum powder :-)

  3. From a complexity and difficulty perspective, I find it difficult to say anyone’s a knockoff of anyone else. Frankly BLE tracker design is fairly cut-and-paste, and I would primarily credit Nordic for having great example code, hardware design, and efficient chipsets for any run-of-the-mill BLE tracker (based on their chipset). If something has particularly long battery life or low cost – then maybe I would credit careful electronics design to the creator of that device (this generally involves a combination of hardware and software). Unless you have carefully measured power consumption of each tracker, I don’t think you can meaningfully determine this with a teardown.

    The TrackR is frankly not even competitive. It seems like it uses Bluetooth 4.0, which is inherently far too power hungry to perform well while serving this purpose. Before BLE it would have been more feasible to design some proprietary 2.4GHz phone case external tracker than to use Bluetooth, honestly – IMO many engineers considered using Bluetooth 4.0 for this and discounted it after looking into size, power, etc. Part of the reason these trackers became so popular after BLE came it is that it became blatantly obvious that the application was very doable. The difference between a well designed one and a poorly executed knockoff is a function of battery life, range, loudness, and latency, much of which is actually a firmware problem instead of a hardware problem, so I really don’t think a teardown would say much.

    There’s one caveat – the BLE chip on the so-called “knockoff” is not a Nordic device. It is possible that they copied parts of the silicon design, but it’s also possible it’s a purpose-built chip, only for tracking purposes, that actually outperforms the Nordic device. If it’s the latter case, I’d readily call them a competitive device rather than a knockoff. Either way, the devices are not even using the same chipset, so I’d really have trouble calling one a knockoff. I find it difficult to somehow claim that a rectangular device with a battery and buzzer somehow has enough merit to be “unique.”

    Actually, when I first saw the PCBs, I thought the “knockoff” was the older device. That random edge mounted cap sounds like it’d increase production costs with no meaningful benefit. If you’ve done any serious BLE design, you know that type of electrolytic is usually a no-no in low power design, and probably ruins the battery life. Of course, maybe it’s performing some function that’s not decoupling, or it’s used in some sort of non-battery-draining manner.

    For this particular application, cost is honestly part of the innovation. Most people would not buy a key tracker for $50-$100 each, so that doesn’t really help anyone. But once it hits <$10, the price makes the utility of the device accessible to more people. Cost is as much part of engineering as is design.

    1. That’s a lot of text to sound stupid. Bluetooth 4.0 _is_ BLE. That was the major addition to the spec. to move from 3.x, and one of its existential use-cases was indeed to do devices like this. BLE is also anything but power hungry – these devices should have lifetimes in the multiple year range on a CR2032.

      There are many manufacturers (including Nordic) who do single-mode BLE chips – as you correctly state, they pretty much all have a tracker as their “my first BLE device” example code. None of them are ‘knockoffs’ of each other (afaik).

      Any tag manufacturer bothering to do anything over and above the chip manufacturer’s reference design is wasting their time for an app like this.

    2. I think this edge mounted part is an inductor. But I have also seen electrolytic caps in older low power devices like calculators, mounted like this. Probably there are different qualities of caps.

  4. The big question is this — can you pair them to each other’s apps?

    All the knockoff devices I’ve seen, work only with one’s own devices. I can make my keys beep, but if my keys are across town, I can’t ask some other “crowd locate” user to keep an eye out for their MAC address. The Trackr and Tile apps have a gazillion users looking for each other’s marked-as-missing devices, and that’s what makes them powerful.

    But neither one makes a version in the form-factor I want. I’m handy with a soldering iron, but both the Trackr and Tile devices are expensive to make mistakes with.

    I’d much rather get some dollar-a-piece beacons and chop ’em up until I’m comfortable gluing one into the handle on my Pelican case, then pair it with the Trackr or Tile app. I haven’t dug deep enough to figure out how these things recognize their kin, but if there’s just some ID string I can change in the generic ones to make ’em work with the name-brand apps, I’m all about it!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.