Rocket igniter + data logger

#1

Hello all,

I’m a perfect newbee in the electronic field. I started 1 week ago to play with Arduino and esp8266.
Really fun and nice but I have a lot of things to learn. No problem with code part since it’s my job since 20y but I have all to learn in the electronic field.

I’m trying to build a rocket igniter using wifiwebserver + a parachute deployment system using acceleration + atmo pressure logics + data logger.

I would like to design a custom PCB for that. Since I’m also a total newbee on Fritzing (nice Qt app!!)
I don’t know how to use the PCB view + auto routing features.

I’m also blocked with the 2 igniters. I have only 1 digital output free after pluged my sensors (MPU6050 + bmp180) and my SD card board. The first igniter must be actived when we call a link through wifi. The second one need to be activated when rocket is at apogee.

I also would like the PCB being 5cm width max (no limit for the height)

Here what I’ve thinked

can someone help me please?

#2

For range safety, you better have that wifi well secured. It would be very bad if anyone but authorized people could trigger that wifi command. A basic esp32/esp8266 with web server is not really enough. Anybody with access to the wifi network/access point could hit the link. Depending on the situation, wifi networks are minimally secure.

I know of another project that had much of the same functionality. I am checking to see if any of that is available online. I believe parachute deployment was purely mechanical (no software involved), but it included sensors, data logging, plus gps tracking. The controller was Ardunio mini, so no problem with pin counts.

Autorouting will attempt to position traces (one connections are defined in one of the other views), but the parts themselves need to be positioned on the board first. Manual routing of the traces is normally much cleaner.

#3

While I’m not all that familiar with the NodeMCUs, there look to be a bunch of digital pins available in this pinout diagram:

both s2 and s3 on the top of the board (unused on your sketch) show as being GPIO pins (and thus digital input/output presumably if configured that way.) In addition to the issues @microMerlin pointed out, you have a few more (some Fritzing, some not.) Assuming the sketch is intended to be in the rocket, the space shouldn’t be an issue. Here is your sketch with pcb set to 13cm by 4cm (you can set that by selecting the grey backgound and setting the size in Inspector (the lower right window.)

Capture
Capture1473×915 179 KB

rocket-1.fzz (102.0 KB)

There are a few Fritzing part problems though. The node mcu part is badly broken. None of the holes are drilled because they are all ellipses. There is a corrected version (which I have not checked) at the bottom of this post. Since the person complained that the part had no holes, I expect you have the first one here :slight_smile:

as well both the sd board and the bmp80 board have pins that are too small for a .1in connector (although wires should work fine.) The non Fritzing problem is that you don’t have a power source here. I assume that a battery will connect to the nodeMCU somewhere, but the connection isn’t shown in the sketch (and this won’t work without power!) Assuming you need access to the usb port on the nodeMCU for programming, the layout I used will block access to that, so would probably need to change. I expect weight will be an issue here as well, so you may want to find a smaller (and therefore lighter) SD card as well.

Peter

#4

Depends on the specific board. Multiple boards use the same chip, but some (ESP01, the smallest I know of) make only a few pins available. I did not look at the actual sketch file.

#5

@microMerlin no security problem since the esp8266 will be the wifi acces point. It will be protect by password. Also no risk to be hacked since rocket must be fired in isolated place. The module will be powered on only when the rocket is ready to be launched. :slight_smile:

@vanepp thank you. I will try this evening with the updated composant and plug a power source. I also can use same digitals pins for the MCU and BMP since they have diffrents address. I’ll update my Fritzing sketch and share it for review.

I’m also searching for 3V relay for the igniter part I will power with 9V battery. 1 for takeoff, 1 for ejection charge to deply the parachute. Any advice on 3V relay ? I find 5V everywhere but no 3V ones. I could power them with esp8266’s VIN but I’m not sure it’s a good idea.

regards,

Bruno

#6

A good point. I know of at least 3 variations of the lolin node mcu all with different pin width (0.9in, 1in and 1.1in as I recall) so you need to make sure the Fritzing part you have matches the cpu board you have :slight_smile: .

Some of the ESP boards have a connection for a 18650 3.4V lithium battery (and there is a battery holder for 18650s I made for someone here in the forums. A google search of the form “fritzing part 18650 battery holder” should find it.) That may be your best bet. As well given this is a rocket, with presumably high acceleration you likely need to arrange for straps to make sure the battery (and the boards if they are not soldered to the pcb) don’t pop out under acceleration.

As always google is your friend :slight_smile: a search for “3v relay” turns up a bunch of them. As does a search for 3V relay on Ebay. The Songle ones appear to be popular on the various modules available (I don’t have any experience with relays though.) Note you may need a transistor circuit to drive the relay, the coil probably needs too much current to be directly driven from an I/O pin (and the coil will need a snubber diode for back emf suppression!) Given the environment (for instance acceleration may cause the relay to close without a signal on launch depending on the orientation of the relay!) I would probably look at driving the igniter with a transistor or FET rather than a relay (no moving parts, not affected by acceleration, less weight.) You would need to know how much peak current the igniter draws to size the transistor that drives it though. Another thing to consider is wifi range, will the wifi connection be reliable at the distance the rocket rises to? I don’t have any experience with the height rockets get to and not much more in the range of wifi, but reliable range is always a consideration on a radio link. Especially if you are depending on the internal antenna on the cpu board. Best to try this out on the ground with the rocket a bit farther away from the base station than you expect it to make in actual flight (you may need the base station on the top of a hill to provide the same line of sight as will be present in a real flight.) As well rotate the rocket around to make sure the link doesn’t drop in some orientations (antennas are often at least somewhat directional!) All these things may already be familiar to you of course.

Peter

#7

rocket.fzz (110.9 KB)
I replaced the nodemcu with fixed one. Also removed the sd card, temporaly, and played with oled screen.
2 layered PCB looks ok with autorouting. Need to adjust now with SD card :slight_smile:

#8

Make sure it is a GOOD password.

Information is not available on line. He says though, that for higher powered rockets, 2 ejection charges are a good idea. One that fires at apogee and one fires 2-3 seconds later as a backup.

You may need to rethink which controller you are using. One with more pins is the easiest fix. You can also use a port expander. Other options are to use sensors that work with a sharable bus, like I2C. There are various ways to make the available GPIO pins go further.

Or SCS or SCR. I believe the igniter ‘burns out’ (goes open circuit) after firing, so an SCR should work as a trigger. Needs verifying.

I don’t think that is an issue. The wifi is only needed during launch. In flight, data logging is to the SD card, and on board processing is used to trigger the parachute. If in flight telemetry is a goal, something with longer range than wi-fi is definitely needed.

#9

Ah! Illumination strikes! :slight_smile: The wifi link is what causes launch. I this case I agree, an https connection with a strong password is required, as this is a human safety issue. A launch command while someone is near the rocket would be a disaster. I would be inclined to use a physical wire from the control point to the rocket rather than wifi. That is much easier to make safe by disconnecting the physical wire at the launch console when someone is going to approach the rocket. Safety first! As to the SD card, I was searching for a microSD card adapter (the one in use looks to be full size, and a micro one should be lighter but perform the same.) That brings up two issues: almost all the modules I found (other than a bare SD socket which may be the best answer) are designed for 5V systems with a 3.3V regulator on the card. Assuming a 18650 battery, you may not have 5V and with a 3.3V cpu don’t need 5V anyway. The SD card runs on 3.3V and the adapter cards have 5V to 3.3V level translators on board which you with a 3.3V CPU you don’t need. Again this would reduce weight which I expect is an issue with rockets. At the same time I saw a comment that the SD card needs more current than the 3.3V output from the MCU can provide (but a search for “sd card current draw” indicates spi mode is about 30ma (but they recommend using 100ma for safety!), so you probably need to check how much current the NodeCPU will provide to the external 3.3V pinsand how much current the particular SD card you use draws.

I’m somewhat puzzled by this. In the rocket there is no one to see the oled, and I doubt it would survive a launch (they are relatively delicate.)

Peter