2-stage J total impulse controlled by Blue Ravens

[Adrian A]

Tomorrow looks like it's going to be a good flying day at the Pawnee National Grasslands where Northern Colorado Rocketry is hosting a launch with 20,000 AGL standing waiver and call-ins up to 35kft. Yesterday, thinking about what I should fly, I thought, maybe I should go for a 2-stager. I am 0-for- at least 5 so far for having a clean multi-stage flight, but maybe tomorrow will be the day I get it right. 2 of the 3 stages of Stratospear, my 2022 BALLS attempt, are still flyable, so today I'm getting to work repairing, modifying, and prepping.

Based on the motors I have on hand, I settled on an AT H999 to CTI J150 combination. This is a 99% J impulse.

RASAero predicts about 23,000 feet so far, but I still need to re-weigh components. If I have a successful flight, it might qualify for the J complex record (currently sitting at 18k and change) depending on what the powers that be think about the head end ignition I'm doing.

The sustainer will fly in the configuration I was going for at BALLS last year, including the motor. The second stage has a broken fin, due to a last-minute assembly mistake last fall, but my StratoSpear 1st stage booster is still intact so I'm going to use that again as a first stage It is designed for a LOKI K, so it's hilariously oversized for the H999. I had to make a trip to the hardware store for some 5/16-18 nuts to help make a spacer between the AT 360 case and the innards of the rocket that hold up the sustainer motor during boost.

I also re-molded the tip of the nosecone, which broke off at BALLS due to a shorted e-match in the apogee deployment. So I think I have all the necessary pieces. Still to come tonight are some electronics modifications so I can turn the trackers on remotely, a bunch of Blue Raven programming, testing and simulation, lengthy prep and then packing so I can leave first thing tomorrow. It might be a long night, but I'm excited.

 

[robopup]

Good luck!

 

[manixFan]

Wow, that is quite the motor combo. I really like the J150 but have never flown any of the Warp 9 motors. And yes, the AT 360 case compared to the Loki 1200 is a pretty crazy difference – and also to the CTI 6XL case. It will be very interesting to see how it performs.

Good luck and looking forward to hearing the results.


Tony

 

[Adrian A]

Wow, that is quite the motor combo. I really like the J150 but have never flown any of the Warp 9 motors. And yes, the AT 360 case compared to the Loki 1200 is a pretty crazy difference – and also to the CTI 6XL case. It will be very interesting to see how it performs.

Good luck and looking forward to hearing the results.


Tony

An H motor pretending to be a K. I added another bulkhead in the middle to reduce the chance of this thing buckling under boost.

 

[Banzai88]

View attachment 575265

An H motor pretending to be a K. I added another bulkhead in the middle to reduce the chance of this thing buckling under boost.

Very creative, but shouldn't the thrust ring of the aft closure bearing on the airframe prevent the buckling?

 

[Adrian A]

The nosecone and the mass of the sustainer and booster airframes will bear on the thrust ring. But the sustainer motor is just friction fit so it and the booster deployments all bear on that threaded rod. So maybe about half of the 200 lbs of thrust will land on the rod.

 

[Adrian A]

The H999 will get it going over 90 mph when it leaves the tower, which is the good part, but it burns out at only about 80 feet altitude (!), and barely gets the stack to 400 feet/second. So this is one case where there's not much benefit for longer ignition delays. For a straight flight, the optimum ignition time is about 6 seconds after launch, but even if ignition immediately follows the boost the apogee isn't much lower. There's not really much time for the tilt angle to change much either, so some of the more sophisticated strategies involving future angle after ignition delay, etc., won't be very relevant for this flight.

My sustainer has the 3rd and 4th channels wired together for an "or". This is the criteria I have in mind

3rd channel (straight boost, optimum delay):

• Burnout >= 1
• Nominal ascent (tilt never more than 45 degrees since launch, even temporarily)
• Tilt angle < 20 degrees
• Velocity < 260 feet/sec (the expected driver)
• Altitude > 500 (expected about 1800 feet)

4th channel (larger angle, ignite sooner)

• Burnout >=1
• Nominal ascent
• Tilt angle < 20 degrees
• Future angle > 8 degrees (possible driver)
• Altitude > 500 (the expected driver for larger angles, about 1 second after separation)
• Velocity < 500 ft/sec (always true)

Apogee and main deployments are default

For the booster, the separation charge happens as soon as burnout is detected. Apogee detection will kick out the av-bay and the chute.

 

[Adrian A]

I just did the airstart configuration from my phone. It looked like this:



You click on the "And" to select the criteria you want to use.

I connected resistors the outputs, and found from the live data on the phone that the 3rd and 4th channels were armed regardless of the position of my airstart screw switch. Oops. When I repaired my av-bay yesterday I soldered the wire on the wrong side of the switch. Fixed that.

Now on to the onboard flight simulation. I set up the sim with 0.3 second burn of 50 Gs followed by 1.2 seconds of 20 Gs (to save time vs the larger expected motor).



I set the simulated second stage ignition to happen when either the 3rd or 4th stages fire. This is a test of the actual output criteria stored and used on the Blue Raven hardware

The sim shows the output status live at the top of the screen, some data and then a little animated rocket icon at the bottom showing the stage of flight. The settings seemed to work! Here is what the sim looked like after landing:




I downloaded the recorded data over the USB serial interface and plotted in Matlab (both features that are coming to the phone next)

Here is the simulated velocity

and acceleration. The values are negative because the coordinate system here is the sensor coordinates and during the test the av-bay had +X pointed down. You can see both simulated motor burns, the main deployment and landing.

Here are the actual voltages and currents measured during the sim. I connected 10 Ohm resistors in place of the ematches. I need to fix that offset in the current measurement, but otherwise this looks great.

 

[Adrian A]

Here are the deployment logic conditions that were used for the 3rd and 4th channels:



The first time all the conditions became true was at about launch + 6 seconds when the velocity dropped below the threshold of 260 feet/second, which is around what the optimal ignition time is for this nearly-vertical simulated flight.




The 4th channel waited to fire because the rocket hadn't tilted over much yet, so the > future angle check was false until closer to apogee.

Time to close out the sustainer and move on to the booster av-bay. At this rate it will be this afternoon before I launch.

 

[Voyhager3]

You better get some good video of this! I want to hear and see that H999 (because I want to find one)! And best of luck!

 

[Adrian A]

It looks like all the events happened when they were supposed to, but the J150 just didn’t ignite, despite burning the BP pellet and a magnelite igniter. All the deployments worked and the rocket came back with no damage. More data later after I get home.

 

[Adrian A]

Here’s the requested launch video.
View attachment trim.32919C7B-AD6B-47D7-AE2A-9BAA811B51AD.MOV

The puff was the sustainer igniter and BP pellet firing.

 

[Adrian A]

Here are some more photos:


The head end ignition wires before I taped them up. Yellow stuff is vacuum bag sealing tape I used to seal up the holes when I potted them with epoxy from the other side.


The wiring of the main and apogee charges from the av-bay through the chute cannon base

Ready to go out of my new 6' tower.


Sustainer post-landing shot. No damage, despite having a full-mass motor on the descent. The little toroidal chute did its job. 28 feet/sec descent



Here's the head end igniter and the residue from the burned BP pellet.

 

[Adrian A]

More photos from the prep:

The nosecone after re-casting the tip in the mold with tungsten-epoxy. The mottled patten on the side is where extra resin squished up and had to be sanded off. I didn't sand off quite all of it because my arms got tired.


One side of the av-bay showing the screw switch for arming the airstart




Another side showing the Blue Raven, and a Tracker where I attached a magnetic switch so that I could assemble the rocket and turn on the tracker at the pad.

 

[manixFan]

I’m sure you’ve already thought about this, but it looks like the ignitor was above and to the side of the core in the dished out section of the top grain. If so, that might help explain why the motor didn’t light. If the pellet ignited from the side rather than the center, it might have burned too slowly to ignite the core. Same with the ignitor itself – the top of the grain wasn’t sufficient to sustain ignition. I’ve flown a fair number of Mellow Yellow motors and I can’t ever recall having one not light, but that’s obviously still a small sample and not with head end placement. It concerns me because I plan to also use a Mellow Yellow motor for a sustainer, and if they are hard to light in flight, I may have to reconsider that choice.

Just a thought,


Tony

 

[StreuB1]

Adrian, I am always amazed at the unbelievably small scale that you work at with your small rockets. The amount of electronics you pack into a small space to do so many things blows my mind.

Well done on the flight and good recovery. Too bad on the sustainer ignition. Would painting the inside upper portion of the port with smokeless powder dissolved in acetone be something you would be willing to try? It will help ignition.

I have a few I1299N's in my stash, if you were closer to IL, I would loan you a few!

 

[SolarYellow]

I am 0-for- at least 5 so far for having a clean multi-stage flight, but maybe tomorrow will be the day I get it right.

This was written by the guy who makes the hardware I plan to use to run my airstart build. He's now 0-for-at least 6.

It might be more difficult than I'm imagining...

 

[Adrian A]

Adrian, I am always amazed at the unbelievably small scale that you work at with your small rockets. The amount of electronics you pack into a small space to do so many things blows my mind.

Well done on the flight and good recovery. Too bad on the sustainer ignition. Would painting the inside upper portion of the port with smokeless powder dissolved in acetone be something you would be willing to try? It will help ignition.

I have a few I1299N's in my stash, if you were closer to IL, I would loan you a few!

Thanks for the offer. I think if I returned them after the loan they wouldn't be of much use to you, though. I have an I1299 saved away also. There isn't a great upper stage motor to pair with it that keeps is under the J total impulse. And for K total impulse there are some bigger motors I'm planning to use. I may play with it as a single stage to set a personal record for Gs under boost.

In 3 weeks, weather permitting, I'll try again for the J complex record with an I540 to a J94. This will combine for 1278 out of a possible 1280 N*s. Now that I know how hard the mellow yellow is to ignite, I'll try painting the J94's upper grain core with some Magnalite and work harder at getting the igniter into the core. This combination actually sims higher than my H999/J150 combo, so fingers crossed.

Then at the end of May I'm planning to go to the National Sport Launch in Alamosa and take advantage of their 55kft waiver to go for the K complex record. This time I'll use the booster as designed, with a K-1127 followed by a J530. If it all goes to plan (hah!) I'll need the whole waiver.

This was written by the guy who makes the hardware I plan to use to run my airstart build. He's now 0-for-at least 6.

It might be more difficult than I'm imagining...

Take heart that I'm a lot more careful with other people's rockets than my own. After decades in the aerospace industry, I am attracted to rocketry because it gives me the chance to make my own mistakes and find out quickly, without hurting anything but my pride and wallet. And the more mistakes I make, the better I can make the products to help the next guy avoid them. For example, I build in a lot of tolerance for av-bay pressure leaks when it comes to the main chute deployment logic, because I often compromise on that when I'm working out a new av-bay configuration. And now with the Blue Raven, the default settings will deploy the main chute early if descent rate starts getting too high because the apogee charge is undersized or a dud or shorts (I have done all 3, sometimes more than once). I am planning to make the Blue Raven's ground test capability fun and easy to use, to take the hassle out of it, and provide more actionable data.

Now if I could just get the electronics to deploy another, hotter igniter into the sustainer if the first one didn't work...

 

[boatgeek]

I’m sure you’ve already thought about this, but it looks like the ignitor was above and to the side of the core in the dished out section of the top grain. If so, that might help explain why the motor didn’t light. If the pellet ignited from the side rather than the center, it might have burned too slowly to ignite the core. Same with the ignitor itself – the top of the grain wasn’t sufficient to sustain ignition. I’ve flown a fair number of Mellow Yellow motors and I can’t ever recall having one not light, but that’s obviously still a small sample and not with head end placement. It concerns me because I plan to also use a Mellow Yellow motor for a sustainer, and if they are hard to light in flight, I may have to reconsider that choice.

Just a thought,


Tony

Adding on to this, could the BP pellet have blocked flame progress down through the core and inhibited lighting? Under standard ignition, the igniter hits the bottom of the pellet so flame from a half-burned pellet is directed down the core. With HEI, it just hangs out in the top dish of the Bates grain until the pellet burns out, at which point it's, well, burned out. I wonder if it would be better to just dump the pellet (hey, you have one of those already made!) and depend entirely on the Magnalite.

 

[Adrian A]

I have been taking a closer look at the recorded data from the flight. Here are some highlights. First, a close-up of the axial acceleration during the boost:




The whole motor burn was over in about 0.35 seconds, as advertised. The separation charge fired about 0.5 seconds after the motor ignition. You can see the separation charge boosting the sustainer with 100 Gs and pushing back on the booster with about 30 Gs. It looks like the impulse was a little more spread out in the booster, probably because the av-bay floats within the airframe and can push on the friction fit of the launch motor.

Here's a fun one:

When I zoom way in on the vertical axis of the graph, you can see the two sections with the same G readings until the separation, and then the booster had about 7x more drag acceleration than the sustainer did, which of course is why it's nice to separate ASAP. Without a nosecone, the booster's drag is a lot rougher also. Then right about 6 seconds, you can see a small impulse on the sustainer when the BP pellet in the sustainer motor fired. The booster reached its apogee after only 10 seconds.


The graph above shows the tilt and roll angles for the two sections. This was a nice, straight boost. I angled the tower a degree or two downwind so that it would be close to vertical after weathercocking, and it worked, with the sustainer well under the maximum allowable angle at the time of ignition at 6 seconds. I'm a little surprised to see how little roll the sustainer had, too. It went back and forth a little but at apogee it was only about 30 degrees from the launch orientation. You can see that the booster was trying to roll one direction and the sustainer was trying to roll the other direction until separation.

Both sections actually got to apogee without tilting more than 60 degrees. I'm not sure I've seen one of my own flights be straight enough for back-sliding, but it happened this time. The Blue Raven detects apogee based on taking a 2-of-3 vote between gyro tilt > 90 degrees, accelerometer velocity < 0, and baro pressure increasing. For both stages it was the latter two votes that triggered the apogee deployment.


Here are the sustainer electrical measurements. Now that I have increased the current measurement frequency to 500 samples per second, you can see the current go up not only for the motor ignition, where the BP pellet ignition flames and soot enabled the current to flow for the whole 1-second firing time, but also a brief spike during the apogee and main deployments at 15 seconds and 38 seconds, respectively.

 

[Adrian A]

I’m sure you’ve already thought about this, but it looks like the ignitor was above and to the side of the core in the dished out section of the top grain. If so, that might help explain why the motor didn’t light. If the pellet ignited from the side rather than the center, it might have burned too slowly to ignite the core. Same with the ignitor itself – the top of the grain wasn’t sufficient to sustain ignition. I’ve flown a fair number of Mellow Yellow motors and I can’t ever recall having one not light, but that’s obviously still a small sample and not with head end placement. It concerns me because I plan to also use a Mellow Yellow motor for a sustainer, and if they are hard to light in flight, I may have to reconsider that choice.

Just a thought,


Tony

I agree with this.

Adding on to this, could the BP pellet have blocked flame progress down through the core and inhibited lighting? Under standard ignition, the igniter hits the bottom of the pellet so flame from a half-burned pellet is directed down the core. With HEI, it just hangs out in the top dish of the Bates grain until the pellet burns out, at which point it's, well, burned out. I wonder if it would be better to just dump the pellet (hey, you have one of those already made!) and depend entirely on the Magnalite.

I like this theory too. My next attempt I'll break out the pellet and paint a the inside of the core a little, and make sure that the igniter sits inside the core. One of my safety protocols is to remove the nozzle starting before I connected the HEI, until I'm at the pad. The grains can shift around a bit during transport that way, but I'll see if I can rig up something to hold everything in place, with the igniter down the middle, without the nozzle.

 

[boatgeek]

I agree with this.

I like this theory too. My next attempt I'll break out the pellet and paint a the inside of the core a little, and make sure that the igniter sits inside the core. One of my safety protocols is to remove the nozzle starting before I connected the HEI, until I'm at the pad. The grains can shift around a bit during transport that way, but I'll see if I can rig up something to hold everything in place, with the igniter down the middle, without the nozzle.

I like that safety protocol for HEI. Could you 3D print (or just cut) a puck that fits inside the case/liner that's the same thickness as the nozzle shoulder? With a friction fit (or tape if your setup allows), it would stay in place for transport but blow out if the motor lights.

 

[Adrian A]

Now that have more than 1 day to prep for my next 2-stage attempt, I'm looking at some more options to improve performance. I still have a smaller, lighter sustainer body that I built for an I216 single-stage flight to over 16kft in about 2011. Its fins are smaller and only 0.050 thick. And I have the first 2nd stage airframe that I built, also in about 2011, which is lighter and a lot shorter than the super-long booster I used for this last attempt. This airframe had a bad landing at BALLS 2011 and had a broken fin. I peeled off the broken tip-to-tip uni that was holding it on, and re-glued it into place. The peeled and re-glued fin is now basically straight but it has a little twist to it. but I think it's o.k. for a short first stage. If I were using it as the 2nd stage in a 3-stage flight, I'd start the fin over

I also went on a little motor buying spree this morning and ordered a Loki J474 and a CTI I125 long-burn motor. They combine for 1276 Ns. The Loki 474 has a huge initial 1200 N thrust spike, which will accelerate the stack 55 Gs out of the tower. After I finish adding a 6' extension to the tower I'm working on now, the exit velocity is predicted to be 125 mph. That should make it relatively immune to cross-winds. The optimal coast duration is around 13 seconds (!), so this will be a good use of the Blue Raven's tilt sensing airstart logic. Max altitude could be north of 26k. My RASAero doesn't have the I474 loaded into it yet, but on a J500 (a few Nsec too big but similar thrust profile) I'm getting over 28kft. I would definitely like this 2-stage J impulse flight to exceed the single-stage J record, and with this combination it has a good chance.

 

[Adrian A]

I'm prepping for a flight tomorrow with an CTI I540 going to a CTI J94 mellow yellow. They combine for 1278 Ns total impulse. This time I put the igniter on the other side of the BP pellet. The hole down the middle was just large enough for the igniter wires to pass through. It's a nice big igniter that is jammed up into the top of the top grain pretty well, so I'm pretty confident I won't run into the same issue as last time. Now if I can just avoid dumb mistakes in prepping tonight/tomorrow it could be a nice flight.



I'll go for the same 2-option approach I have tried before. One channel is set up for the optimistic case of a flight that it straight enough to let the velocity go down to the RASAero optimum (350 feet/second, assuming a 1.5 second ignition delay). The other channel is set up to fire as soon as the predicted tilt after a 3 second ignition delay gets up to 12 degrees, and as long as the tilt at decision time is less than 18 degrees. That kind of flight puts the apogee about 7000 feet away horizontally, which is in the acceptable range for our launch site. The upper level winds are going in the same direction as the lower level winds, which helps, so any weathercocking will send the rocket into the jet stream up-wind of the launch site.


There should be about 28-30 Gs in the tower, which isn't as much as some other boosts I have been looking at, so it will be a good use for my new tower extension that doubles the height, bringing it up to 12 feet. The tower exit velocity is predicted to be 74 mph.

 

[OverTheTop]

Do you keep the HEI igniter shorted out during prep? That's what I did on my two stage Nike Apache.

I also carved an E motor grain and inserted it in the core of the M2020. The igniter started that which started the M.

 

[Adrian A]

I keep it shorted until the wires are twisted and taped to the electronics, at which point the nozzle is off, the electronics are off, and the screw switch between the electronics and igniter is open. That’s how it stays until it gets to the pad.

Then the tracker is turned on and gets a good view of the sky, then the nozzle goes on and the stages are stacked in the tower. Next the altimeters in both stages are turned on and I use the Blue Raven phone app to verify the deployments have continuity, the airstart doesn’t, and the airstart is disarmed in software. Then I turn on the airstart screw switch and verify the airstart has continuity but is still disarmed in software. At this point, if I had a rocket too tall to reach the arm switch, the tower could arguably be horizontal pointed away from the flight line for hardware arming with the software remotely disarmed, and then rotated vertical. I arm the airstart in software via Bluetooth as the last step before leaving the pad.

 

[Adrian A]

Yesterday I took the time to put a magnetic switch on the sustainer tracker, verified and tested the settings of the Blue Ravens, and finished making a new harness attachment point for the booster. I didn't get to the launch site until 3. Which would have been OK because the weather was quite good. I got my new extended-length tower set up, but then found out I had left my range box at home, including the sustainer nozzle. I did the 4 1/2 hour round trip and hopefully I can send it up it when there is a break in the cloud cover.

 

[Adrian A]

I made a few more mistakes that once again prevented a successful flight. 0-for-7.

The launch conditions were pretty windy, and the boost was a little squirrely, but at the edge of acceptability for tilt. It would have been worse if I hadn't extended my tower to 12 feet. Unlike the last flight, where the booster was visible for the whole flight and the sustainer was visible for most of the flight, both stages disappeared right away. The sustainer didn't ignite the airstart, this time because I forgot to do the software arming step. I had power cycled the sustainer's Blue Raven but didn't look at it again with the app before walking away, relying on the beeps to verify that all channels had continuity after re-arming the screw switch. The beeps don't (currently) include the software arm status, so I really needed to look at the app before walking away. I may update the firmware to make a software-disarmed channel have the same low beep as it would if the charge were disconnected.

The other more major mistake was deciding to go without a tracker in the booster. My 38mm booster av-bay is more difficult to safely cram a tracker into alongside the Blue Raven, because the threaded rods are permanently attached to the av-bay tube and so everything has to slide into the tube without visibility. When I did that 3 weeks ago, I pinched some battery wires and got a smoky short. In that flight I ended up flying the tracker in place but not bothering to turn it on (it can't get a GPS fix inside the CF tube), which was fine because it stayed visible. This time, I never saw the booster again after motor burnout, despite hiking several miles along the ground track of the sustainer, upwind and downwind. A friend of mine even took me up in his plane for an aerial search yesterday, but the bright pink chute never showed itself. I had a good experience finding my small fluorescent chute of a lost rocket from the air in 2011, so either the booster went way off track somehow, or the chute didn't deploy. I have flown my ejectable av-bay design several times with 100% chute deployment success, so I was really expecting to find the booster and its chute, and its disappearance is baffling. Never again will I fly a stage that could possibly go out of sight without a GPS tracker. I do have yet another CF booster stage that had a broken fin from a previous BALLS attempt, and I have 3 weeks to fix it up and get it ready before NSL launch in Alamosa.



The sustainer got up over 11,000 feet despite the lack of airstart ignition, and I had good tracking on it. The apogee deployment went fine, and it ended up coming back right overhead for the main chute deployment. Because of the GPS tracker pointing ability, I knew when it was coming overhead, and everyone was able to see it as it passed over. Unfortunately, the main chute tangled, which is another first for me with the chute cannon, and so it came in pretty hard. Fortunately, the main chute deployment shook out the unburned motor (which we were also able to see) so it landed softly enough to avoid any permanent damage:


The nosecone tip didn't even break off. I haven't tried untangling the chute yet, but I hope to find a clue about the parawad when I do.

Next up: flight data

 

[OverTheTop]

I have flown my ejectable av-bay design several times with 100% chute deployment success, so I was really expecting to find the booster and its chute, and its disappearance is baffling. Never again will I fly a stage that could possibly go out of sight without a GPS tracker.

Lucky for you is that you know where to get some decent electronics at a good price .

 

[Adrian A]

I like how well the inertial data (based on gyros and accelerometers) matches the GPS data, after the GPS data caught up and until the apogee charge fired.


Here's the tilt data, zoomed out:


The tilt crossed 90 degrees at 24.1 seconds of flight time, which exactly matches the timing of the baro apogee, and is about a second earlier than the GPS and inertial altitude apogee.
Zoomed in for the part that determined the airstart logic:

The rocket really did those oscillations, I believe. It was spinning and coning. It's interesting that the tilt went down at first, but that could have been from ascending into stronger winds that tilted it up a bit. The angle was significant, but still within the 18 degree max angle I had set.

The rocket was spinning hard. The booster fins got it spinning one way, and then after separation the sustainer fins got it spinning the opposite direction. In fact, the roll rate exceeded the gyro rate capability of 2200 deg/second at 0.7 seconds into the flight:


That makes it a little surprising the inertial navigation estimate did as well as it did.

The accelerometer data is always interesting. Here is the boost and the booster separation charge:



These are raw sensor values, and up is in the -X direction, which is why the X axis is showing as negative during the boost. The Y and Z axes clearly show the rocket bouncing within the tower in the first 12 feet, but there didn't seem to be much effect on the axial acceleration. There was a pretty hard start of the motor. I used an augmented igniter along with the CTI black pellet. The booster separation charge happened right on time. It was a smaller charge this time.

Zooming in on other areas of interest, the apogee charge was sized pretty well:


Here is the main chute ejection charge:

To make attaching the chute cannon a lot easier, this time I didn't pre-package the deployment charge BP in a cardboard tube, but instead I had it loose in the bottom of the cannon. It looks like this may have been a mistake, since I didn't get a big initial X axis boost, and the Y and Z started tugging at it immediately. I think maybe the chute tangling may be related to getting fired out more slowly.
The hard landing was the biggest G load of the flight, by far:



The battery and output continuity voltages are pretty unremarkable except you can see when the unburned motor got flung out of the sustainer just after main chute ejection, because the continuity voltages went to zero when the igniter wires broke:



Here are the logic states that went into each of the deployment outputs:



If the third channel were armed, the airstart would have been triggered as soon as the velocity dropped below the threshold at 5 seconds into the flight, because the future angle (angle predicted 3 seconds into the future) was above 12 degrees for most of the early flight, and the flight angle was less than the 18 degree maximum during that time.


The fourth channel, which was also connected to the airstart to cover the case of a very straight boost, would not have fired because the angle was greater than FANG1 threshold when the velocity was low enough.