Rocket Avionics - Building Lumineer

hey i'm joe barnard and today we're going to be talking about the avionics stack for lumineer the flight computers and electronics which have you know they've we've seen some better days uh this video is also sponsored by ren and we'll talk a little bit more about that later so let's get started [Music] flight computers and avionics for high-powered rockets can do all sorts of things and it really varies depending on the rocket what your avionics are going to do some high-powered rockets don't even need flight computers at all they use motor ejection to get the parachute out and the only computers you'd fly are for data logging some high-powered rockets rely fully on the computers to get parachutes out and some even down to the stability of the vehicle itself and to add to this complexity because of the inherent risk involved in high-powered rocketry a lot of the times you want redundant avionics and to add even further to that you often don't just want redundancy but dissimilar redundancy which means separate flight computer designs separate code bases separate manufacturers so that if you did something wrong or if you use some computer wrong there's a totally separate system to back it up for lumineer i opted to fly two computers one was an ava board manufactured by me and the other was a telemetrum computer manufactured by altus metrum the telemetrum is a dual deploy computer but it can do a few other things it has a gps on board it's able to broadcast telemetry on about 433 megahertz i selected the telemetrum because i had heard good reviews from friends who had used one and i know that it's flight proven on a lot of high-risk flights on the ava side of things when the board was you know still functional it had capabilities to fire parachutes track gps coordinates integrate accelerometers a lot of the things that the telemetrum could do as well ava's telemetry radio was on 915 megahertz so that we didn't cross with the telemetrum and ava was also set up to control a reaction wheel which like all of the electronics here has seen some better days the avionics bay was set up with a drogue deployment piston in the center ava on one sled telemetrum on another sled and the third sled for the batteries we also had two cameras in the avionics bay one looking up one looking out and both died on impact we actually have the sd cards and probably the footage from these cameras but i can't get it off three out of the four sd cards on the rocket were cracked in some way or fractured and so like the footage exists on there i know the data exists i just can't access it if you are some type of wizard who can resurrect a cracked sd card like the ceramic chip has cracked if you can do that let me know i am doubtful back to the avionics let's look at a very basic schematic to see how things are wired up these schematics are really quick and dirty the point is to get a sense for what wires go where and just basic functionality because the goal is redundancy between these two computers we want them on totally separate power buses with total electrical isolation so we'll do the schematic separately for the telemetrum we have a switch that is turned on by a screw through the wall of the airframe this turns the computer on on the apogee pyro channel we have the drogue charge and on the main channel we have the tender descender black powder charge if you watch the recovery video you'll know we actually had two redundant tender descenders and the big secret here is that to add to all of this redundancy each computer could fire both tender descenders so we wired up four total pirate charges to the two tender descenders both could be fired either by the telemetrum or ava to close this out telemetrum is powered by a 750 milliamp hour lipo battery ava's schematic is a little more complex and to the left we've got all of our flight cameras operating at 14.8 volts which are powered on by ava we direct the camera power through ava because we want the ability to turn the cameras on or off when the cameras are on we're drawing something like two plus amps to power them all this isn't terrible because we have a huge battery that's specked out for a reaction wheel but we don't want to drain that power if the vehicle is idling on the pad for two hours so we route it through ava ava can then control the same things as telemetrum which is the two pyro channels for the drogue and main parachute ava is powered on using the power hatch that i mentioned in the airframe assembly video but we arm the computer electrically using a screw switch through the wall ava also uses one of the pwm outputs to drive a speed controller for the reaction wheel motor and the esc also has a separate power line which goes directly to the lithium polymer battery so that we don't have to route crazy amounts of amperage uh through ava through the pcb itself ava broadcast telemetry using an external xp radio over 915 megahertz and to close it out we power the whole thing using a 2200 milliamp hour 4s lipo now a quick word about the reaction wheel so this was a really cool idea that i did a really not cool job executing i've said this before about lumineer but to re-emphasize it we operated on an extremely crunched schedule so i sent out step files and had a few different people manufacture reaction wheels to fly on lumineer i want to give a big thank you to those folks so the first is keith or rocket powered keith on twitter keith immediately volunteered to make one of these wheels he did a fantastic job and it came out great next up a huge thanks to masa for making several different versions of the reaction wheel with the bps logo and their own on there finally thank you to daryl yerke of solitude components um daryl i am sorry yours flew which is cool but it's a little worse for the wear so from the bottom of my heart my bad in terms of the control logic i had my buddy charlie garcia draw up an attitude controller for the reaction wheel he and a few friends took over the task of drawing up this attitude controller because i was focused on like just bringing this vehicle to launch they worked really hard they put in a bunch of effort and then we intentionally flew ava with the reaction wheel turned off before we talk about why i did that i want to talk about why the reaction wheel was a bad idea in the first place on a fundamental level the reaction wheel for a high power rocket does not make sense because you're resisting these aerodynamic torques that as you go through the boost phase are only getting stronger and you're resisting these aerodynamic torques in a non-aerodynamic method that saturates quickly in my case i was also spinning a heavy and not professionally balanced mass at a higher and higher speed and that heavy non-balanced mass was connected directly to the avionics stack with all of the inertial sensors like i don't know if you can see why that's a bad idea but just trust me it's a bad idea one of the wildest things to think about is that as you ascend you start storing momentum of a fully loaded vehicle in the reaction wheel but you're losing mass as you ascend and like the inertia of the vehicle is going down so you're storing more momentum in the wheel if you don't account for this as you go up you're storing more momentum in the you know what it's insane like the control problem is harder than it seems and then finally like all of that momentum is stored in the wheel so when you reach apogee or when you reach some point when you're ready to dump that momentum you have to put it back into the system and you're naturally just going to roll the vehicle right back up if you stop moving that reaction wheel and like what are you going to do drive this reaction wheel at 90 hertz for like a 10 minute flight it's really chaotic probably not and so then you're like okay cool well you'll dump the roll momentum at apogee right wrong i mean not wrong but like then you risk tangling up your drogue parachute anyway all of this is to say i think reaction wheels are a great idea at a small scale when your vehicle isn't going super fast and you don't have a lot of aerodynamic roll torque but for high powered rockets i think there are better ways to do it all right so on a fundamental level the reaction wheel doesn't make sense but i put in all of this effort to get this thing on the vehicle like ready to go so here is why we didn't fly it active the tldr is that we only finished building the attitude controller or the math behind the control logic in about a week before the launch that in its own is concerning but the more concerning thing is that like three days before launch charlie was like you know what we should do let's just like see if it works just hold the avionics stack and see if it'll stabilize so i loaded the code and we did that and it like oscillated a few times we do not unfortunately have footage of this event i'm like mentally losing my mind at this point i'm holding the avionics stack the wheel starts spinning up it does resist the torques effectively and i'm like okay cool let's give it a kick so i rotate it fast and it draws a crazy amount of current from the wheel and it browns out the entire avionics stack and in that moment it was like cool it's three days before launch and we're gonna fly this wheel as dead mass and then the best part about this story is you're like okay cool so just take it offline and you increase the safety of the whole flight right so like not 30 seconds after this i tell charlie cool we'll just unplug the esc from the battery so we don't drain that extra like quiescent current see how this esc connects in the diagram here we're disconnecting the red line and i forgot to disconnect the black line and you know what is inside of an electronic speed controller is a tremendous number and amount of capacitance and you know what happens when you plug into a capacitor bank um the capacitor bank charges and it does that pretty fast so i forgot to unplug the pwm cable from ava and uh favorite thing about the linear right now is the wireless video so we're going to plug in okay there's the video everyone want to take a look we're going to power on via the power hatch power hatch oh that was a loud pop [Music] oh no i smell something hot it's not too me too i burned it ava so essentially we took the reaction wheel offline to make the system safer and it killed the computer as a result really as a result of me not thinking through the schematic there's a lesson to be learned here um and the piece of information i'm not sharing is that right when i pitched the idea of unplugging the esc charlie said hey buddy that's not what your schematic says have you ever flown it or ever tested it in the state where the esc isn't plugged in and i was like what do you know man it's just positive and ground like it doesn't matter so the lesson to be learned here is when someone gently pitches the concept that like maybe this isn't an excellent idea or maybe you should think this through the correct response is not i know more than you moving on i have a few random facts that i couldn't find a good place for in the video so we'll just talk about them here this is less about the computers and more about the wiring but all of the cable runs that went through the vehicle used separate types of connectors this is really important to me and i think it should be important to you too if you're building something like this you should never ever be able to plug two different cables into each other that shouldn't go together moving on every cable that went into one of ava's or telemetrum screw terminals used solid core wire screw terminals are built for solid core wire frayed wire is not really what you want there you also want those solid cord wires to not be tinned you don't want to tin the tip of them with solder solder compresses easily and over time you can compress the wire down and you lose that grip that you get from the screw terminal screw terminals in general are just kind of a bad idea with high powered rockets i live with them with ava i'm not stoked about it they can be used but they are also susceptible to backing out in really high vibration environments and do you know what vibrates a lot it's rockets it's rockets that vibrate a lot so if you can avoid using screw terminals you probably should but if you have to use them there are a couple of ways to make them work moving on each computer has a way to electrically disconnect the pyro charges this is really important and you can't just rely on digital arming of these pyrotechnic charges physical arming is critical you should not be able to have any type of current path that could possibly fire one of the charges until you are prepared to fire one of those charges and in this case prepared means that you know that you're in a higher than normal risk scenario usually that's on the launch pad and you're prepared to like bail out if you need to and the final point here is that all batteries for the avionics stack were sized to give this thing like hours and hours of on time i think if you were to look at it from a outside perspective it might seem like overkill to have you know seven or eight hours of powered on time for the whole vehicle but when you're searching for stuff in the desert sometimes it takes a little while with all that covered here's a little bit of the footage of us prepping the avionics bay for flight we're really not how we die it's just like don't touch it and it will be fine apogee lockout what do you want to say 25 yes your apogee is at 45 seconds in your nominal simulation are you subsonic at 25 seconds oh god all right uh open rockets okay so this is the telemetry module we don't need anything else for now tight tight yep this is why we do checklists yup will it be fine with three screws almost sure do you want a fourth screw for first safe like you never feel like you need a checklist until you do a checklist and then you realize how many things that passed you past you wrote down that that present you doesn't know okay so we just we just did a bunch of checkouts the av bay is in pretty good shape we do need to put a battery in we need to wire up the tender descenders which will um allow the main shoot to come out um but like it's going well it's 10 40 a.m um we've been at this since like eight and i'm in a much better spot today than i was especially on friday last week all right you ready so we're just gonna hot glue this in because this is a slight risk um at 10 g's of acceleration i feel like it's more than a slight risk it's like can we just say it's just like a big risk fan play it off as if it's not a big deal all right ready yeah so hot glue is easy to remove isopropyl alcohol yep and then so adam's just taking a shower the next stuff is shoots i'm going to start laying those out a little bit in here um and then we integrate which is really scary because that is the last time that i see i mean launches tomorrow but that's the last time that i see my avionics stack before i either see it on the ground or i see it in the ground or i don't see it all three are likely you got this all right great what i'm going to do is um force it open and then we're going to check the length and then we're going to close it and that will be the last time that we close it okay okay so right now we're like we're like pulling vacuum to get it all the way out wow that's a good seal 14.8 volts nominal okay great 4.2 times four 16.4 volts full charge 16.8 so that you're counting it as full yeah i buy it part is before then so we're just going to sort of it's it's l minus one and we're on l minus zero checklist active which is the opposite of where we were last week where it was l minus zero and we were running l minus one checklists yeah okay so here's the best part about this build series i'm gonna open this while i do it the best part about this is that like the beginning videos would be like oh wow he's put a lot of forethought into this there are cool heads we're making good decisions and by this phase i am i am constantly on the verge of tears favorite thing about the linear right now is the wireless video so we're going to plug in okay there's the video everyone want to take a look we're going to power on via the power hatch oh that was a loud pop [Music] oh no i smell something hot it's not too me too i burned it ava how i don't know because inductance of the motor no we have less power draw than we did before something shorter um a short could do it what burned on ava it'll be the voltage regulator when we have a really big power in rush it can fry the voltage we've seen this before yes and so i thought the problem had been mitigated by connecting ground first and seemed over every test i've done to do just that but that was a pretty serious pop cannot be this the reaction wheel is unplugged this time gosh i i did we test in the reaction wheel unplugged configuration before i'm like pretty sure it's just ava and we're drawing less current than we would be can we disconnect the power leads to ava and check them for continuity and see that's a great idea let's start there um we are going to have to redo this checklist and probably in a different oh you are i got it oh bless bless you wait can you hold that pen up again okay say it with me there is a path to flight it is 4 p.m you're six hours ahead of schedule fan yes there is a path to flight yes there is i am you are calm down at this point but there is a path to fight so we're gonna have to play yeah there's 4 pm pass the flight is 4pm i must speed run this oh no oh no are there nuts on the back of this i think there might be oh it's so hard to take this off oh please just end it all here we go uh proposal i hate interning at this camera we're all wild people but like i think the the trait that we share is that we're pretty cool under pressure and oh the pressure is rising that has never been a problem in the in the setup that i have oh you know what i know where it happened where you're still connected to the signal line on the reaction aren't you oh my god charlie oh my god oh my god hold my hands oh my god you genius you genius that is it that's the path yeah we sank all of that current pacifist oh my god and it went right through oh my god that's it that's probably what fried the last one wow it works okay cool and it's got the beef okay so similar okay you got those so just keep your eyes on this yes sir and keep your fingers crossed you ready fingers crossed for you fans three two one enable cameras cameras on status on one two three that's the live feed hot glue sd card into ava dom uh hot glue sd cards and avionics cameras um toggle sd cards into vehicle side cameras check out photos of the ava face oh my god new ones yeah yeah okay so that's one that's tight um charlie do you mind this i've got the commission package this is two i want good shots those news figures out photos and torque stripes oh god yeah you're right i hate that you are but it is true we're gonna turn everything upside down we're both gonna do it yep all right now we have some alignment to do the last line uh yeah hold on this is uh normally normally chill but you boy needs the space atm oh my god it's working do we want to try a power on uh we should okay i i agree it's me joey please watch over us as we power on all right ready set it man straight and fast like somehow this is way worse than the other ones hold on yeah that's it that's that's ground idol that's uh oh yes all right the night before the flight we powered on the avionics with no booster section attached the telemetry looked clean and strong so we pushed forward to flight the next morning the first flight of that day was usc rpl with their control v vehicle 3 2 1 we had ava flying on board as a passive payload to help capture more data for their flight and once they got off the pad we set up lumineer a little further away around this time i started the live stream which i will link in the description below so you can watch us prep if you want once the vehicle launched we lost almost all telemetry at around four kilometers up we actually lost telemetry on both computers 915 and megahertz i will keep this section short but the tldr is that rf is essentially blackmagic and we should have done a lot more range testing on the ground the radios were sandwiched between two huge chunks of metal inside the vehicle the lower one was the motor forward retainer which is called an arrow pack and the upper one is this massive brass reaction wheel we were also really lacks about our data downlink budget which means that we were nearly saturating the bandwidth while on the ground and the bandwidth is only going to go down as you go up we did however get flight data that was recorded on ava and then reviewed after the fact so i figured we could take a look at that we'll start off here by looking at our barometric altitude which puts us at an apogee of about 9.7 kilometers so a little shy of 10. we can also see some interesting barometric effects as we break the speed of sound and then drop back below it when we get going fast enough if you look at this plot right here you'll see that the gps reports no satellites in view and no gps fix when we pass about 500 meters per second we can also see our gps altitude just stop updating at about that point this makes sense the gps i use in the data sheet says it will not work faster than that speed that's called a cocom limit and it's a limit within a lot of gps receivers it basically says like you can't use this on a missile if you start going too fast it's going to get angry and tell you no i won't do it we can also take a look at our accelerations here and we actually have three accelerometers on the vehicle which gives us some really cool data to correlate the three sensors the first is our high g excel or high gravity accelerometer this picks up the launch transient really well we have an average acceleration of somewhere around like 90 meters per second squared and then this little tiny peak of acceleration at about 115 meters per second squared the other two accelerometers that we have on the vehicle are configured for lower gs and they can't quite read that high of an acceleration so you can see them cap off here you can also see some things maybe possibly shifting around inside the vehicle there are some uncomfortable acceleration events that i am not a super big fan of here if we look at the off axis accelerations which are the accelerations not in the vertical axis we can see a pretty violent shock happen and it looks like this happens around when we go supersonic we all been new like transonic is a really sketchy place aerodynamically and we sort of have the data to confirm that the vehicle gets a little whipped around when we pass the speed of sound ava logs these things so we can see the command to fire pyro channel 1 right around apogee which fires our drogue charge apart from the rf stuff both computers on this flight functioned flawlessly and did their job to the best of their ability telemetrum and ava both independently sent the drogue firing command within 0.3 seconds of each other so they both had a very good agreement on what apogee was that said the telemetrum did get there first we can see in the acceleration data that the drogue charge fires just a little before ava sends the command so ava was a little bit delayed i can't quite claim that my own flight computer got the shoots out on the way down telemetrum and ava also both fired their respective tender descender charges we also got very lucky on the way down we had a period of reorientation of the vehicle where the polar patterns of the ground and flight antennas matched up well enough to get a couple of gps packets through and we were able to use those packets to ultimately find the vehicle once we got the vehicle back one of the first things that i did in the days after launch was extract the lithium polymer battery from the very crunchy avionics bay obviously you can't slide the avionics package out at this point so i had to sort of destructively disarm this essentially like lithium polymer um fire waiting to happen so i had to go back and look at the cad to find where the battery was and then be really really careful about where i dremeled through the airframe so that i wouldn't accidentally cut into that lipo both telometrum and ava ended up being in somewhat non-wonderful states which is to say they were absolutely both dead and you might be wondering to yourself hey joe how come you were able to show all that flight data if the computers were dead i am getting uncomfortably good at extracting flight data from dead computers i don't love that i have done it this many times but it is not that hard to desolder the flash chip on one computer and then like put it on another one and get the data off that way i would love if i wasn't getting so good at this skill anyway that's about it for the luminaire avionics video i learned a ton of lessons in building this avionics stack and we're going to carry those all forward into the high power rockets that i build in the future we should have at least one more coming up this year and then we have a bunch coming up next year thank you so much for watching and in the next video we're going to cover motor integration which is honestly like just just like prepare yourself for a an amount of chaos and now it is sponsor time this video is sponsored by ren ren is a website where you can calculate your carbon footprint and then offset that carbon footprint by supporting projects which plant trees protect rainforests etc the climate crisis is happening and like it's not very good but you can use ren to help offset your impact on it you can use rent's footprint calculator to evaluate your impact and learn about how you can reduce it obviously it's unlikely that you'll be able to get down to zero impact but for what you can't eliminate you can offset with monthly contributions to projects which make a positive impact it's going to take a herculean effort to end the climate crisis and the only way to do that is through collective action of lots of people so you can do your part by heading to ren dot co with ren you'll get to look at where your money goes to help understand the impact that you've had i've partnered with ren to plant 10 trees for the first 100 people who sign up using the link on screen and in the description below thank you again to wren for sponsoring today's video and thanks to you for watching my name is joe barnard may your skies be blue and your winds below

Channel: BPS.space

Views: 234,173

Keywords: BPS

Id: pZ572Rjj9vA

Length: 28min 55sec (1735 seconds)

Published: Sat Sep 18 2021