Possible Hardware changes? CAT Version

[crazybanane]

Hello, I got the idea to create a tracker for my Cats, and now i am found your LOKO Devices. I think it is a great device but i would do some changes to get it to work for Cats. So my questions are:

-Is it possible to change the GPS module without further modifications to software? Because the GPS module that is actually intregrated in Your LOKO AIR is hardly to find and expensive. I wanted to run an Quectel L86-M33m module.

• For charging i see you used MCP73832T and Protection FS8205A, to save space i would change to TP4054 charging IC and DW06D for Protection

• in my Version that i did until i found your Device i tried to get the PCB only 1 side with components equipped

• Do we can work on an Groundstation Version that can used Stationary? Or is it possible to use the Air Version with external LoraWan Gateway in connection with the Mobile Apps or is it only possible to get the Data for Position only Via Bluetooth? found for part2 of question answere on the website

-Would you help to make a LOKO-Cat version that can use your Software?

[tomipiriyev]

Hi @crazybanane ,

The GPS module is the most challenging component to source at this size. If the module outputs standard NMEA data, it should work without requiring any software modifications.
You can source a suitable module in Europe here .
On the PCB, most of the space is actually taken up by the USB connector rather than the charging and protection ICs.
I’ve spent considerable time looking for alternative GPS modules. One concern is the increasing geopolitical tension between Taiwan and China, especially since many GPS manufacturers are based in Taiwan. Additionally, Quectel implement geofencing at the firmware level, which can restrict operation in certain sanctioned countries. For this reason, I prefer not to use their chips.

[crazybanane]

Hi @tomipiriyev ,

thanks for answer. With the link you helped me a lot. I'm actually at a state on my PCB Version where i can say i could save much space with other components. And it would be easy to integrate different GPS modules so there would be no problem to choose the one that is better available in the using Country.

The question is are you interested in a CAT Version and did you accept it if i would work on this Version with your Software?

[tomipiriyev]

@crazybanane, I am not sure what do you mean by "cat version " , I think already you can use loko on cats. need better plastic casing for cat collar. meanwhile feel free to use Loko software in your version .

[crazybanane]

My concern was that if I attach it to the cat, the transmitter would always slip down. Therefore, I'm unsure whether the GPS would still work properly, and I also don't know how to neatly stow the antenna. So, my idea was to attach the main hardware to the lower part of the collar and mount the GPS module externally, so that the GPS module sits at the top of the collar. I would also try to integrate a flexible PCB antenna directly into one side of the collar instead of the wired antenna, so that it's still adjustable. I've seen the pictures on the website (dog collar) and also the teardown video. Personally, it just seems a bit too big for the cat. That's why I wanted to try to distribute the components better. If you say I don't need to worry about its functionality with a cat using your version, I don't see any reason to continue with mine.

[tomipiriyev]

The electronic part of Loko is extremely compact, while the plastic enclosure appears relatively large. This is mainly due to the manufacturing method. At the current stage, we are producing in small batches using 3D printing. With this approach, it is not feasible to create very small plastic enclosures due to the limitations of 3D printing technology. also we cannot make it waterproof because of this limitation .

We are still in the experimentation phase, observing how users interact with Loko. Transitioning to injection molding would significantly increase costs, and even small design changes to a mold are expensive and difficult to implement.
Regarding the antenna, we have successfully tested a flexible wire solution. A high-quality silver-coated audio wire performs very well. It only needs to be soldered horizontally to the PCB, with no design changes required. When used on a collar, the wire naturally aligns along the belt, making it practical and unobtrusive. I will share more photos next week.

As for the components used in Loko, most are sourced from Chinese manufacturers. Based on my previous experience, relying solely on Western components can be risky. During COVID, we experienced major chip shortages, and currently, prices are increasing again due to rising energy costs.

alos forgot to mention , Loko Works with LoraWAN gateway and we also added to Loko App.

[crazybanane]

Yeah your Product is very small, i like the way your steps will go. If there is interest, this is my version as I envisioned it. Actually without any case only PCB

[tomipiriyev]

it looks nice, coin cell looks too small for cat tracking , I uses 200mah li-ion , with 1 sec interval it can run 2 hours , I am planning to raise the battery capacity to 500mah , just slightly increasing the size . with punched li-ion battery . Loko at deep sleep uses 9-11 uA . after wake up 36 mA. on open area takes about 8-9 seconds to find a new coordinates and go back to sleep . these infos might be helpful to estimate required battery capacity .

[crazybanane]

I would perfectly fine with it if I get updates around all 5 to 10 minutes. I don't need this high Update rate. When the Battery hits 3 days it more than good.

[marcindulak]

I don't want to hijack the discussion, only to present some probems that may affect the possible design of a cat tracker.
They made me drop the idea of a cat tracker. I wish solutions to these problems existed.

The question is what the tracker is protecting the cat against?

A. Assuming the cat lives in a city, when a cat is missing, it could be due to the cat wandering.
This takes up to 2 days regularly, like twice a year the cat may be missing, because it decides to explore the environment or due to intensive social life.
A tracker that sends the signal for 3 days is insufficient in that case.
It will lose the signal at the time when it's needed most.

B. Another common reason for a cat missing is when it gets trapped in a basement or garage.
A determined cat can wander several hundreds meters in 10 minutes.
In this case it's the GPS that fails its purpose.
When indoors, it most likely won't be able to acquire the position, and the uncertainty of the area will be too large to find the cat.

C. Probably the largest risk to the tracker is the collar itself.
Even breakaway collars have a risk of entrapment and injury of a cat https://pmc.ncbi.nlm.nih.gov/articles/PMC8465301.
Moreover, collars that fall off easily, may be lost by the cat at the moment when tracking is most needed, like during the fight or escape.

[crazybanane]

For me, it's not about completely protecting the cat from being lost or stolen. My cats are silly, yet quite intelligent. We live right on a main road, and they use the neighbor's to get to safer areas without crossing the main road, which is a good thing. Sometimes, though, one of them wanders off and stays away for the night. Then I'm curious where they sleep and where they roam during the day. I know cats can cover a considerable distance in just a few minutes, but this is relatively rare, as they tend to hide if attacked by a larger animal. If my cats aren't home in the evening, they're usually there the next morning, so the charging issue isn't a problem. I assume that if I can trust tomipiriyev calculations and scale them down to my own version, I'll get at least three days of protection. Of course, more is always better. And I think I can get it better. Yes, I know that a collar can be fatal for a cat, which is why I want to develop the collar myself and try a different approach than most manufacturers. Our cats are also microchipped and could still be returned if their collars were lost. My intention isn't to track a cat only to collect it after three days; if anything, it should happen within a maximum of 48 hours. I simply think that an update interval of 5 or 10 minutes is perfectly adequate for cats. It's a completely different story with dogs. If a cat is involved in a fight, it shouldn't matter whether it loses its collar or not. In 90% of cases, the cat stays within a 250-meter radius and usually responds to its owner's calls or comes home on its own. So, you'd only need to find its last known location and locate the collar. That's at least my experience here in the village. Theoretically, you could also implement an alarm if the cat hasn't moved outside for, say, five measurements. This would allow you to assess whether something is wrong or not, because normally, at least our cats are always on the move. They sleep at home. And I don't believe that the GPS receivers that are now readily available, like the ones used here, are any worse than a smartphone. Even with my smartphone, I now have GPS reception most of the time indoors. I've already tried something else (Moto Tag) hoping it would be sufficient, but unfortunately, these trackers didn't work at all as advertised. That's why I'm looking for my own solution.

[taprootmx]

This is my custom casing for the Loko Air for mounting to a dog collar, maybe it will spark some ideas. I put the battery and the PCB side by side to make it lower profile. Version 2 will have the antenna resoldered to the bottom side of the PCB and coming out flush with the collar. It’s printed in 95A TPU so it is tough and can flex and conform with the collar. I melt the top and bottom halves together with a soldering iron and am designing a custom TPU plug for the USB port so it should be able to be more waterproof than the original design (which would have never survived my big dogs!).

It could be made a fair bit narrower for smaller width collars, or if you were willing to make modifications to the collar iteself. Aside from ruggedness and flexibility, this design would also make it easy to add a larger battery. For dogs, the height is often much more significant than the length as to how durable the device will be. Making sure that the battery doesn’t get crushed or punctured when dogs are playing is another big challenge. I’m going to design a sort of “crumple zone” around the battery to protect it from stray teeth and chewing.

The accuracy of the Loko GPS is quite poor. It does well outdoors and while moving, but indoors it is not usable—you won’t be able to tell whether your cat is in your house or 5 blocks away. Cell phones are able to achieve their accuracy in large part through sensor fusion, where they internally run a Kalman filter and combine GPS measurements with accelerometer, gyro, and magnetometer data to get a much more accurate position measurement. I’m still hoping that @tomipiriyev will release the PCB source files so we can make modifications more easily like adding an accelerometer.

[tomipiriyev]

@taprootmx
Loko was originally designed in Altium, but it should be redesigned using open-source software such as EasyEDA or KiCad, as the ASCII file is not fully compatible.

[taprootmx]

KiCad can import Altium ASCII files, have you tried it?

[tomipiriyev]

I haven't used KiCad , tried on EasyEDA, I will upload ASCII files here to test

[martinl]

We use Loko for cats as is, ordered the airtag cat collars from Temu plus some optional needlework to sew the antenna to the collar. Temu also has threaded usb-c plugs, the original ones without threads were scratched off in a few minutes. Small usability issue is remembering to charge the Lokos every night so we sometimes use airtag collars as backup (airtag battery lasts about a year even when left outside over winter).