Advices for Custom Flight Computer and Telemetry

[unalfaruk]

Hi everyone! I'm newbie here, while I was researching about model rocket studies, I found this forum and signed up here Here are lots of information and I want to follow them always.

However, I want you to guide me for our student project. We are a team which is a competitor for international and national(in Turkey) organization about Model Rocket and Satellite. In the last year, we built successfully a model satellite for CanSat Competition, and this year we applied for a competition about Model Rocket.

We are electronics and mechanics engineering students and we love R&D. So we could design our flight computer(board) and telemetry. Now, we want to improve our design, can you guide me to lead my team?

Our components for the first draft:
1. Teensy 3.5 (Dev. Board)
2. BMP180, MPU9265, LM35 (Sensors and Modules)
3. UBLOX NEO 6M GPS
4. XBee 2.4 GHz RP-SMA 63mW Transciever Modules

Our questions:
1. Where is the best location for antenna?
2. Which type of antenna is the best for rocket?
3. How can we improve/amplify our telemetry signals? (in Ground or in Probe)

Our rocket must carry more than 4kg payload and its maximum altitude will be 3000+-100 meter.

Thank you.

 

[blackjack2564]

You will get more answers by putting your request in "electronics & software" instead of here.

 

[unalfaruk]

You will get more answers by putting your request in "electronics & software" instead of here.

I'm so sorry for this, how can i move it?

 

[tim cubbedge]

It has been moved per Jim's recommendation.

 

[ksaves2]

I'm not an electronics guy but 63mW is just too darned weak for telemetry purposes. The recovery of position packets from the 100mW stuff is, "Not every single position is received.". But is very sufficient to find a rocket. I've used a variety of products and am happy with them all.

Now if you are doing a micro-miniturized project for a modroc that's not expected to drift for a couple of miles,
you'll be in good shape. Want every bit of data? Just look at NASA films from the 1950's that have a shot of corkscrew antennas (circularly polarized) being pointed at just launched rockets of any kind and you'll catch my drift. Kurt

 

[UhClem]

Never build what you can buy. Do you have some requirement that can't be met by existing products?

Telemetry is almost certainly not needed as you can record far more data to the micro-SD card. Unless you are expecting the rocket to be destroyed or lost.

 

[unalfaruk]

I'm not an electronics guy but 63mW is just too darned weak for telemetry purposes. The recovery of position packets from the 100mW stuff is, "Not every single position is received.". But is very sufficient to find a rocket. I've used a variety of products and am happy with them all.

Now if you are doing a micro-miniturized project for a modroc that's not expected to drift for a couple of miles,
you'll be in good shape. Want every bit of data? Just look at NASA films from the 1950's that have a shot of corkscrew antennas (circularly polarized) being pointed at just launched rockets of any kind and you'll catch my drift. Kurt

Thanks for reply.

I prefer XBee Pro S2B 63 mW because its datasheet says "it supports 3200 meter range", and its frequency is 2.4 GHz, also we can use a 2.4 GHz signal booster(4W) on ground station side.

Which products have you used before, can you suggest me? Get every bit of data is important for us, we want to make the best we can.

 

[unalfaruk]

Never build what you can buy. Do you have some requirement that can't be met by existing products?

Telemetry is almost certainly not needed as you can record far more data to the micro-SD card. Unless you are expecting the rocket to be destroyed or lost.

You're right, but we're students and improving our development skills is vital for us. Besides that, we don't try to design a new RF circuit for communication, we buy standard commercial modules and integrate them together on our custom board.

Just we need is telemetry pair module for 3000+-500 meters. And we aren't sure about the type and location of the antenna in the rocket.

 

[SparkyVT]

I'd recommend the newer LoRa chips. I use them with a similar setup and get over 5,000 meters of range with just 30mW of power. If you want to see my setup, here's a link:

https://github.com/SparkyVT/HPR-Rocket-Flight-Computer



I'm working on the next version, but here are a few lessons learned:

1) Keep the pressure sensor and GPS receiver as far away from the RF transmitter as possible. I had LOTS of EMI issues with my latest setup
2) If you need more than 24 G's, then use the ADXL377. The digital H3LIS331DL is very noisy.
3) Ublox M8 modules work much better than the prior versions. Putting it into the right flight mode is key.

Also, why do you need the LM35? The BMP180 has an integrated temperature sensor. Also, the BMP180 is out of production. I recommend the BMP388.

Sparky

 

[unalfaruk]

I'd recommend the newer LoRa chips. I use them with a similar setup and get over 5,000 meters of range with just 30mW of power. If you want to see my setup, here's a link:

https://github.com/SparkyVT/HPR-Rocket-Flight-Computer



I'm working on the next version, but here are a few lessons learned:

1) Keep the pressure sensor and GPS receiver as far away from the RF transmitter as possible. I had LOTS of EMI issues with my latest setup
2) If you need more than 24 G's, then use the ADXL377. The digital H3LIS331DL is very noisy.
3) Ublox M8 modules work much better than the prior versions. Putting it into the right flight mode is key.

Also, why do you need the LM35? The BMP180 has an integrated temperature sensor. Also, the BMP180 is out of production. I recommend the BMP388.

Sparky

Wowww! You are more than the hero which I'm looking for Thank you for these recommendations.

Actually, we also planned to use 30 mW telemetry but we couldn't find the best solution. We will consider using LoRa RFM95W, however, I watched your video and I couldn't see any antenna or amplifier, didn't you need them?

1) I can understand the problem because of GPS but I don't understand why the pressure sensor affects the RF signal of telemetry?
2) As the calculation of the mechanical team, we need just 8G sensor. But this information is really vital for us for the next projects.
3) We use Ublox GPS models in our tests, we generally use Adafruit GPS in the launch. But its initialization time is too long for the beginning.

I will be watching your GitHub repository

Best regards.

 

[SparkyVT]

I used a whip antenna on the flight computer, and a Yagi on the ground station. With the LoRa chips, I'm pretty sure that at your distances the Yagi might not be necessary.

I don't recommend the Adafruit GPS units. I initially used them and had poor performance, losing lock immediately and only regaining it halfway back down, which it retained intermittently. Maybe they will work better if I programmed the chip with the available commands. With the Ublox M8 and the setup code on my Github site, I keep GPS lock the entire flight.

If you are flying your unit in a rocket, then I highly recommend an accelerometer that can sense more than 8 G's. That's pretty low by rocketry standards.

I don't understand the issue with the barometric pressure sensor either, but I assure you its real. I've spent months testing to thoroughly confirm this problem. The BMP388 works much better, but being so close to the transmitter it still has some issues. I will try to fly it this weekend.

 

[unalfaruk]

I used a whip antenna on the flight computer, and a Yagi on the ground station. With the LoRa chips, I'm pretty sure that at your distances the Yagi might not be necessary.

I don't recommend the Adafruit GPS units. I initially used them and had poor performance, losing lock immediately and only regaining it halfway back down, which it retained intermittently. Maybe they will work better if I programmed the chip with the available commands. With the Ublox M8 and the setup code on my Github site, I keep GPS lock the entire flight.

If you are flying your unit in a rocket, then I highly recommend an accelerometer that can sense more than 8 G's. That's pretty low by rocketry standards.

I don't understand the issue with the barometric pressure sensor either, but I assure you its real. I've spent months testing to thoroughly confirm this problem. The BMP388 works much better, but being so close to the transmitter it still has some issues. I will try to fly it this weekend.

We also used a whip antenna for our CanSat and a Yagi for GCS in the last year. However, isn't it important the location of whip antenna? Can you advise me about the location for the antenna for rockets? LoRa seems really nice when I look at the specs, however, its RF output is 20 dBm, how can it success to send data over 3 km?

You're right, we had also the same problem and the difference between Adafruit GPS and NEO6M is just sampling rate. I will consider using UBLOX M8, I live in Turkey and buying these new modules can be difficult sometimes because of money budget and shipment problems.

Now, I asked the mechanical team about G information. Our speed will be less than 0.8 mach regarding the competition requirement, so they are like sure We will see it, after mechanical tests Because, regarding me, this G value is weird.

Now, I have a new research topic, it's a really interesting thing. I want to test it, too.

Finally, what do you think about RSSI Tracker?

Good luck with your flight

 

[SparkyVT]

The RSSI Tracker looks like a neat project. However, I think its overkill for a rocket telemetry system. LoRa shouldn't need that level of sophistication. A simple system is also a thing of beauty.

 

[Stomper]

You want to join my project and support it? It is using LoRa WAN... LoRa uses chirp spread sectrum to achiev such long distances. And it workes f... great. I achieved 5000 m LOS with a good whip antenna and both sides.

https://www.rocketryforum.com/threa...-you-be-interested.148078/page-2#post-1853021

 

[SparkyVT]

I read the competition rules and now understand better. This is more of a helicopter project than a rocket project. I now understand your need for telemetry. You are welcome to copy the necessary bits from my code. I think your hardware setup will need the following:

1) An inertial measurement unit. I use the LSM9DS1. The code to interface with it is in my project and it runs very fast.
2) Ublox NEO-M8Q or Ublox SAM-M8Q. The code in my project will let you put it into the 4G navigation mode, which will help it keep the GPS fix during launch and tumbling
3) A LoRa telemetry unit. I use the Radiohead RFM95 library which is free to download.

In the next few days I'm posting updated code that optimizes the telemetry packets for GPS at 4 samples per second and barometric data at 20 samples per second.

Let me know if you want me to just send you the correct bits of code.

Stomper, I agree. The LoRa units are great. I get miles and miles of distance on only 20 mW of power with a whip antenna on the transmitter and a 5 element Yagi on the receiver.

Sparky

 

[unalfaruk]

I read the competition rules and now understand better. This is more of a helicopter project than a rocket project. I now understand your need for telemetry. You are welcome to copy the necessary bits from my code.

Which one did you read? Here is the URL for our competition. You can read the details, it is only about rockets. There shall not be anything related to helicopters.

Let me know if you want me to just send you the correct bits of code.

Thank you! Your GitHub repo is perfect, I read your code, just you need some optimization. We want to develop the code ourselves using your repo as a base, so we can contribute something to your code such as a new algorithm, a new decision point or a verification method etc.

However, if you can come to İstanbul/TURKEY for the competition day. You can join our team to develop together remotely, we have many subsystems(communication, sensor, flight software/computer, power, ground station), we can work on one of them together. I prefer working on the flight computer because of "Stomper"s idea.