Model Rocket RCS Thruster Update
ฝัง
- เผยแพร่เมื่อ 2 มิ.ย. 2024
- RCS testing live stream: • Live RCS Ground Testing
Older RCS development video: • Reaction Control Syste...
Contact: bps.space/contact/
Help support BPS.space: / bps_space
For more info:
/ joebarnard
/ bps_space
/ bps.space
/ bps.space
www.bps.space - วิทยาศาสตร์และเทคโนโลยี
![[TH] VCT Masters Shanghai Playoffs DAY 4 // 100T vs GEN | G2 vs TH](http://i.ytimg.com/vi/UkTGSxdCUQw/mqdefault.jpg)
Beware of anyone who uses the terms 'linear algebra' and 'really fun' in the same sentence.
😂😂
No! Quite the contrary! When people say that you want them to teach you! They are the people who actually understand it. LA is a great tool. It gives quick and elegant solutions to a lot of tough problems! This is why it is so addictive!
@@romainlaugier5915 I don't know about it at higher level but it was in my high school & it was the easiest of them all
This is my opinion of linear algebra. I think it is fun
I find Physics fun..
“Excuse me sir!” 1:21
Calling your rocket a Sir is the pinnacle of love for your item
It’s MA’AM
“Pshhhh! Pshhhh!” -RCS
XD lol I was searching for this comment
Hi Joe, admire your work! Used to be a controls engineer for a major valve manufacture. Suggest mounting the valves on a custom round air manifold with ports directly out the rocket to eliminate the leaks and therefore the arming valve. This will significantly reduce mass. Using valves with very fast and consistent response time will make RCS much easier to control. Basically, high speed PWM the air to make it behave "analog". The valves should be balanced so the response time does not change with air pressure.
Looking at the video, I don't think you necessarily need more Cv. More Cv will increase size and mass while quickly depleting your air supply. It may make it harder to control as each air pulse may over adjust and create instability.
You can connect a pressure transducer to the valve output and put the transducer output and valve firing signal on an oscilloscope to characterize your valves for better control.
A small super capacitor at the power supply output may improve valve response time. Finally, if you have diodes across the solenoids, this could greatly delay the time it takes for the valve to turn off. Use a switching system that can handle the voltage spike when the valve turns off. A MOSFET with high enough voltage rating should do the trick.
Here's an example valve to consider: www.macvalves.com/product/bv10-mm/
Sorry for blabbering on and on with my opinions... Let me know if you want help specifying the valves, designing the manifold, etc.
Keep up the good work! Enjoy your videos :)
We did the exact same things for automatic transmissions for hydraulic gear actuators! You can also put a dither on the solenoid signal so you get rid of the slipstick effect
Great idea, I came here to suggest that exact same thing regarding the monoblock round air manifold. That really is the ideal approach.
I don’t know any of these words😂
I used to enjoy model rockets as a kid, making a model rockets land is just next level cool. From understanding your RCS project, seems like it does take some time to tune a PID controller. Other than testing on a gimbal, have you tried to put the rocket or RCS tube in a jig suspending it on stretch cords, & or balance on a big bobblehead spring?
With that type of test, & tune jig, it seems to me that would make it possible to measure the thrust coming out of each manifold with a force gauge, possibly making it easier to tune the PID while balanced on axis, & reduce any damage on live testing. Also with the manifold design, would you also incorporate recessed thrust cones into the design to improve efficiency of the RCS?
@@supakat1835 is it possible to calculate the thrust based on the mass flow? That would be very helpful.
Hi,
happy to see that you're doing some attitude control of your rocket. To answer shortly to your question, we found a high flow AND high-speed miniature valve from Parker. I don't remember the reference but they were kind to answer us at the time, based on our requirements.
Long story short, my bachelor thesis in 2015 was actually about controlling the attitude of a model rocket on a similar scale (2.45 kg, 1.4 m length); a cold-gas thruster attitude control for a rocket model. The aim was to control the position of the rocket, especially at low speed.
And during the process, we get to the same trouble you have/had. As the weight was a limitation for us (we couldn't get more powerful motors), we used 4 proportional valves (around 4mm diameter) to control the yaw and pitch with 2 of them 5mm off the axis to control roll and the 2 other one 5mm off the axis the opposite way. So we were able to fully control the rotations. We used 78g CO2 cartridges, similar to bicycles ones but bigger because as the CO2 stays liquid until the last moments, you have a stable pressure output until it dries out and constant force output, which is a good thing when you want to simulate the capabilities of the rocket, position of the thrusters, controller,... This was the main reason we used that cheap CO2 cartridge.
We actually went to present at the 1st symposium on space education activities hold by ESA. Quite a funny experience! Here is a link to the university news about the project if you want to get a bit more information. actu.epfl.ch/news/a-step-closer-to-reusable-rockets/
Don't hesitate to contact me for more info as well.
Him: But Co2 is more expensive
Me: Just breathe in it
中二病 This must’ve what the Mythbusters guys meant when they say “failure is always an option.”
Seems kind of morbid
Imperial units are my biggest fear
I mean...almost nobody uses It but It works so I guess that this unit isn't that bad.
I always forget PSI is an imperial unit. In Australia everything is metric except for a very few specific things. One of those is pressure, which is always measured in PSI.
Deal with it, we’re in Merica.
I hate imperial except when it comes to pressure then I forget because in Sweden we use psi (for avionics) & bar (for stuff on water)
Not even the imperial power that is , Britain uses the imperial system
Hey Joe, I'm an aerospace engineering student and have been involved in the development of small experimental rockets for the last four years.
Two years ago I had the exact same problem with solenoid valves for an on-board pneumatic system: not enough mass flow rate (thus no thrust at the nozzles) for the ones that we found flight worthy (relatively low mass & size to fit in the rocket + low voltage requirement at 12V or 24V).
After quite a lot of research, we didn't find any that fitted our needs.
We chose to go another way: use a small solenoid valve that activates a bigger air operated valve. While this solution is really heavy (2 valves for each thruster + an additional small low-pressure air tank for the pilot circuit), air operated valves are way more capable than solenoids at equal mass/size. We used the VNB104B-8A from SMC. In order to actuate it in small sequences, you’ll need a 3/2 solenoid valve on the pilot circuit (to evacuate the gas and shut the air operated valve). Small tip: on the main circuit, watch for the internal sizes of your fittings/tubings in order not to choke the airflow before your nozzle.
Our system was only designed to accelerate the roll of the rocket (in order to test a yoyo-despin system), so this solution suited us because we only needed one set of these valves. But I don’t know if the weight and size penalty would be worth it for a complete RCS system like the one you’re trying to design.
I hope I helped a bit, good luck, rocket pneumatic systems are not an easy thing to design ^^
10 yr Prediction: BPS will be the first garage built, reusable launch vehicle!
You sir, are an amazing young man. I am giddy... waiting to see what your future plans entail! You tease us...
I think you are on a personal journey to see if you can make a notepad to orbital rocket project! I think you will succeed.
I love everything about your channel! ... you energy, and what you are doing is inspiring to me and my little boy! What you are doing is unique, brilliant, scientific, multi-disciplined, artistic, compelling. But more importantly, educational, inspiring, and ... well everything I want my 4yo boy to "get".
Harrison Wheeler my son is also hooked on this channel.
Use soapy water and a small paint brush on all your connections. Any leaks will show as large bubbles.
In order to counter the low mass moment of inertia in the roll axis you could split the tubing, then allocate one half of the tube to pitch and yaw and the other to roll. You could then put a secondary regulator on the roll axis tube to lower your feed pressure and make it more controllable without losing much power in the system.
You could also split (or direct) the flow slightly downwards, which would reduce the overall roll force, whilst adding vertical thrust. No extra parts required.
@@aread13 This would generate a torque force in either the pitch or yaw direction though, which would have to be countered by firing additional thrusters.
@@itemboxes8741 Good point..
Super smart dude helping the rest of us super curious and maybe not-as-smart people become super smart... Righteous hero of science right here.
Did u jus assoom my intillegent?
Why not trying some engineering or math stuff right now? I will be getting into this rocket science stuff.
@@stevato0449 same😁
Wow I'm such a better person because of watching this video. What a wonderful experience it was to sit through all roughly 13 minutes of it. I‘ll never be the same and I feel bad for everyone who didn't watch the whole thing.
Build a single block with 4 valves. There will be connections and the weight will be lower.
You should need to only have 3 2-way valves to control FB LR CWCCW. You don't need to have a valve per direction, as you'll never be pushing (let's say) left and right at the same time. By the looks of things Joe is using 6 1-way valves, which seems a little inefficient. Definitely, if he could machine (or have machined) a single block with 3 2-way valves and 3 solenoids, that would be save a ton of weight.
excuse me sir! its got a cool noise. use a dryer on the compressor or you will get corrosion in the tank and Fowl the valves.
Or he could just drain the tank completely in between uses. His scale is too small and too temporal to get this sort of problem.
yer just chicken
Hey Joe, I’ve got experience here, some thoughts:
1) You could drop the number for valves from 8 to 6, one for each control axis direction. Then some tees to join up the relevant thrusters.
This means that you won’t be able to take advantage of the cross axis coupling effects for each thruster that you could if you controlled each individually, but I’d expect the weight savings to be worth the marginal improvement in the control law.
2) If you need to scale down your roll control, easiest way is to just bring in the moment arm. The thruster layout isn’t super clear in this video, but worth taking another look at that.
3) The number to look at when sizing things is the control authority (deg/s/s (angular acceleration) for your shortest impulse) for each thruster in each axis. Some testing is in order I think to figure out how much thrust you’re getting from your setup.
4) State space control is the way to go. Make sure to have a deadband wide enough that it spans a couple firings, so you don’t bounce between the edges. Take a look at Fig. 5 of this paper - a simplified version should do you well. www.researchgate.net/publication/228869689_THE_RCS_ATTITUDE_CONTROLLER_FOR_THE_EXO-ATMOSPHERIC_AND_GUIDED_ENTRY_PHASES_OF_THE_MARS_SCIENCE_LABORATORY
For ground testing, you can use a shop air compressor which can only reach a bit over 150 psi, but delivers much greater flow while consuming less power.
I cant be the only one who understands less than half of this content but still watches as if you knew it all. Great work!
Just found your channel about a week ago. I have really been enjoying it! I love how enthusiastic and excited you are about your project! It makes me excited and I'm really enjoying watching your videos as you learn and progress!
Hilarious and smart. Best channel on TH-cam by far.
This dood and Wintergatan for me
This, wintergatan and lemmino
Suggestion: Higher pressure will give you the higher flow rate you seek. You should be able to get away with overloading the valves you currently have by just ramping up the pressure. You could go with 200PSI or 300PSI instead of 150. You will almost surely need to increase the coil driving voltage to overcome the extra force of the higher pressure, but in your application you can likely get away with it because the solenoids are rated for continuous use, and you're only firing them for short bursts, and your tank only lasts a few seconds anyhow. You can put some code in your software to insure that the valves never get held open long enough for them to burn up at the higher voltage. You'/d need to characterize the actual limits of the valves you have in hand, but I believe you could take some measurements and determine the limits and then run the rocket with a little margin below the limits.
I'm very excited for this project! Congratulations Joe
I'm not sure if there's anything off the shelf but it really needs a manifold/solenoid block type arrangement. Picture a ring of ports fed straight off the reg, the solenoid coils would be arranged within this ring also. I used to do this kind of work, it will really need to be a machined piece, i suggest shouting out to some youtube machinists (Saunders would be a good start).
With the drawing done he can pitch the machining to AvE with a video.
@@ciano5475That would be skookum as frig
Yea this would be a perfect project for AvE's 5axis Haas mill.
That's pretty darn cool Joe. I have enjoyed watching your iterations of learning and sharing, you're doing good stuff and keep it up.
I definitely see a bright future for this gentlemen. Please don’t quit no matter what.
Eyy Joe ... My Kerbal experience suggests it ought to be possible to have RCS with only 6 ports (have the roll ones also double as the perpendicular axis, when fired in pairs
Actually, if he didn't need roll, he could do it easily with only 4 if he used some kind of passive stabilizer in conjunction.
That's pretty cool! I love watching the progress of these rockets!!
Congratulations man! That's quite an accomplishment.
Dude. You are absolutely amazing, your work is incredible.
Your head must be like a Tardis. I can't imagine how you get this much thinking accomplished unless the inside is bigger than the outside. Always impressed with your work.
Absolutely awesome! Love your videos!
i have the same pump. releasing the pressure in the hose BEFORE you turn it off works wonders to improve lifespan.
That rhythm of the thrusters going off would make a neat beat
+1 for scuba tanks in the field ;)
God I love this man's videos. He is brilliant! I am about to do my own model rocket experiments very soon because of you. Thank you. 👍
Excellent job..Tim brought me here with his RCS drone video. Sadly got stuck in watching all these videos after googling some help for my rocket dug out and fixed after 15 year hibernation.. doing some check out flights tomorrow. :)
I like that loud sound warning.
Hope in future every youtubers does that.
I love that RCS quote, its very inspiring
I knew you'd get it. Good job!
Like you said in the video, those cheaper solenoid valves are not the greatest when it comes to flow. Usually in my experience you usually want small solenoid valves that are used to switch a larger pneumatic valve. You can kind of think of it like a small gate driver switching a big bjt or mosfet. Swagelock makes some pretty great pneumatic diaphragm valves. There is a pretty good supply on ebay of NOS valves because pretty much every piece of semiconductor equipment made uses them. Main downside would be all the extra weight. so it may not make too much sense on the scale you are working with. Other solenoid valve companies you may want to look into are SMC or Festo. SMC's sx10 valves do look promising though as they have a model claiming 150LPM at up to 200hz switching speed.
Hi! I think you are right: for larger flow rates, the typical solution would be to go towards staged valves. However, in the context of a closed loop control with small time constants, this is not ideal because it tends to add tons of delay (lattency) into the loop. Even the solenoids add several ms to the loop! I remember we used to advertise the transition to voice coil instead of solenoïds to remove those! To get single stage at a larger flow rate you would need larger "holes" which would mean larger force for the actuator. Or it would need to be pressure balanced, and therefore bulky.
Did you try to work on the nozzle shape? There might be some efficiency to gain there...
I am surprised by the stability of your control loop. For a highly nonlinear actuating system (gated), this is impressive. I personnally would have gone straight to PWM or PFM at the output in order to cling to some linearity. (although the valves may not like the added stress...)
Although thrusters are cool, I think a flight model would be closer at hand if you had gone in the direction of reaction wheels... no?
@@romainlaugier5915 delay greatly depends on the kind of valve Rather than system architecture. Ald valves are typically staged and are the fastest valves in the business.
Well done! Successful launches!
You're doing great job please keep it up
Watched the whole live stream super fun loved the triva
Thanks Ben! Can't wait to do more trivia soon :)
well it was to late in my country to watch the whole livestream but i watched a bit from it :)
@@henryscp318 it was midnight/2 for me
@@benadians1769 ah ok but yeah i had school and then don't wanted to go like at 3 in bed
I loved the whiteboard in the back
I don't know why I came across this video, but I'm glad I did 😊🥇
I can't believe this channel only has 57k subs. Someone reach out to smarter every day he should do a colab
Whaaaa?!! Dude! This is amazing!
you work is impressive
this video came out on my birthday!!
"there's no direct use case I'm just having fun doing it" That's the way to live!
Just a thought but in my paintball experience, CO2 experiences huge temperature drops when shooting a lot. It end up being really hard on the rubber parts of a paintball gun. This is the reason most avid players us a nitrogen tank, or just compressed atmosphere. Atmospheric air is more stable over time as well (Temp is proportional to Pressure) and CO2 tanks on an electric gun can actually freeze up after only a couple hundred (100-200) cycles of the firing solenoid and result in serious performance issues during play (shot velocity drops off severely as your working pressure drops off). I would assume, considering the fundamental workings of these two systems are very similar, using the majority of a tank of CO2 for flight control would have a similar effect on your thrust performance and components.
Love the project!
Ditch the paint ball canister and get one we use in RC flying (light weight), for air retracts. Plus we also have light weight valves.
These are SO great.
Would make a great video if you went over how you derived your plant transfer function for the regular thrust vectoring
This is soooo cooooool!!!! I like this a lot!👍
I think some guy called Gabe Newell runs a valve company
Nico Gutholz
This is by far the best comment here
Was itteraly look at RCS systems then this uploads :P
*BPS.space** would like to know your location*
That blue tube looks green to me, and I colour balance my monitor. It's so neat seeing you be a mini skunk works.
Frankly it's a kind of old Blue Tube - the color has faded a bit ¯\_(ツ)_/¯
Consider "Clippard" Air Valves, not sure of the flow rate, but what they make is usually very robust.
I really adore your dedication on your projects! keep going man! Godspeed
Really cool! I can only launch the premade rockets. The ones boy scouts use. I could never make something like this.
If you are looking for roll control on takeoff and flight then implementing movable(rotatable) fins might be your answer. They can be programmed to counteract the rotation of the rocket and will be independent of the other system since the press will likely be took high when using compressed air without an additional regulator.
you sir are a madman i love you.
Hi Joe, I really love what you do as I am a first year engineering student in New Zealand wanting to get into aerospace!
How do you document your engineering process? And do you have any examples that you could share?
Anything would be really helpful with my studies! Thank you!
Wow!!!! Incredible!!!
Inspirational quote in the background
Joe, have you looked at the current line of Robart air tanks for RC aircraft landing gear?
They also have machined blocks for air distribution. It’s just a thought.
Quote of the day is certainly inspiring! "Pshhhh! Pshhhh!" - RCS
XDreamBeam Indeed!
That system would be great for an RC glider dropped from a WX balloon for the initial descent from altitude (where aerodynamic control surfaces are not yet effective).
Steper moter might be heavey but there are smaller options that could be swaped inplace if there powerful enough to turn the valve under presser but it might not have to if your bottle valve worked fast enough
valves for vehicle air suspension (low riders) would work, ive seen 300 psi nitrogen ran through them. sizes range from 1/4" to 1" and can be purchased as a manifold with 4 to 8 valves in a block
hey Joe! would it be practical to use a reaction wheel for something like this? perhaps for the roll? Kerbal Space Program says they work very well lol
Lord Fawful I think he has actually used one, but wanted entirely RCS for this vehicle.
Have you considered doing brass inserts in 3D printer parts to mount festo connectors? Particularly for the nose cone it may help direct all your fluid out the thruster instead of having some leaking (assuming you’re still using the press insert method that was in v1) also, did you 3D print the appropriate thruster shape on your valves? Based on your pressure and your outside atmosphere by having an actual ‘bell’ it’ll help direct most of the force vectors exactly one direction instead of rapidly expanding in all directions as soon as the gas comes out of the thruster
Have you looked into micro valves used in model aircraft landing gear applications? Not sure about their max pressure or speed specs, but thought worthy enough to mention. Robart and Jet Model Products (JMP) offer product solutions. Best of luck!
Loves from Dayton, TN lol
I would think to focus on active control of the four fins on the rocket to maintain controlled flight before trying to land this thing in reverse back on a pad!
Itll have to be 20 -30 feet tall to have both.
You got ahead of yourself.
Still very cool video! - subscribed!
(Not Yet) Sponsored by Hydrazine
Hey guys this video was sponsored by the carcinogenic effects of unsymmetrical dimethylhydrazine
It will be a while before I sponsor it!
BPS.space It’s hard not to be giddy when a God replies to you!
@@BPSspace I heard Nitrogen Tetroxide wants to sponsor you too!
BPS.space
I know you are just doing this system for fun,
But,
Have you considered using the full size version that diverts rocket exhaust?
Uses an unrestricted burn and then constricts it when it needs rcs
Every time I watch one of your videos I think, "I should really subscribe..", then I look and I already am....sigh.
2:23 "It really doesn't get old..."
Great project! For the RCS jets you want a direct lift valve (like a sprinkler valve) "Cole-Parmer Manifold Direct Lift Solenoid Valve; 6 inlets, 24 VDC" However your best bet for the weight is to use a high speed servo and nylon rotary valve. Similar to the way the gimbal works at the tail. You never need a thrust vector in two opposing directions. Possibly use 2 for each 180 deg. Separate one for the roll. This way you have 3 valves for 8 ports vs 8 valves for 8 ports. Smells like 80% weight savings including all the tubing/manifolds you no longer need. This ALSO paves the way for the real solution which is to use a small parasite port on your rocket nozzle pushing hot gas up a small tube to the top of the rocket (and valve) to power the thrusters. I will make some drawings.
Hey joe, currently I’m creating a rocket engine. All the math is done but the temperature(3000 degrees) is too hot for any normal material. Do you have any idea fixing this problem?
MAC valves are balanced, quick response and varied in flow. Used them for precision robotic applications, but not sure about your application when it comes to the weight. Worth checking though.
With the rocket flying and the pressure it receives from the air it's penetrating, don't you think it will affect the power valves will produce especially with the small orifice and low CV. To sum it up, will it work during the flight considering it's not static like in your tests. Thanks for the video and good work.
Why you prefer active stability to passive? In a model like that you could storage a lot of energy pressurizating the water to use for primary trhust. And you could use another little tank to adjust the direction with your rcs system
Use spool valves.
They are found in hydraulics.
Now that you have the ability to run at high pressure, please run the full 3,000psi to the ports and run carburettor jets at the outlet.
Next, please use a 4valve manifold with double acting spool valves so you only need two on the one block.
Interesting. Is high pressure better than high flow in this application?
@@danielthrana
yes, two fold.
pressure directly translates to thrust, and the stored [ressure is stored energy and he is dumping 15 times the used energy by dropping the pressure through the regulators.
@@brucebaxter6923 Great answer, thanks! Carbourator jets is an elegant solution 👍
Have you considered replacing the manual valve with another actuator and building a test bed? It may be useful to automate a test with a timer and a pressure gauge (on a microcontroller) while trying to dial in pressures/actuator timing to enhance the efficiency of the system.
Joe, i am fairly certain you can bore out the orifice to increase flow. i would dissect one to see if it can be done.
that is AWESOME
Pneumatic tricks I learned while building competition fighting robots:
-Use Nitrogen instead of air... It's better. Air is fine for practice, but when it's go-time, go for nitrogen. (sponsored by nitrogen)
-Absurdly oversize your tubing. You want the device to perform like it's attached directly to the tank, and give it some time to warm up and expand. Big, thermally-conductive, tubes behave like an expansion tank, with less plumbing/weight.
-Use a low pressure valve, like you have there, to fire a pneumatic cylinder, that yanks open a ball valve from the high-pressure side of your regulator to the device you're powering.
-Buy/build extra parts and design for fast replacement. It goes without saying that this last technique will wear out components quickly, but spectacularly.
You may be able to get such a small ball valve from a model steam engine supplier. You can make the miniature pneumatic actuators yourself, and that would be a benefit because you will want to incorporate a stiff return spring. Placing it in the cylinder body can save space and weight. Since your application needs a faster turn-off time than a pneumatic sledgehammer, I'm thinking a double-acting piston with a spool valve may be better, though more complicated. Along that line, putting this much pneumatic/electrical/mechanical gadgetry in a tight space warrants you to start working on a monolithic structure for this assemblage. Such a structure could also satisfy most of the plumbing requirements and eliminate the bulk of individual packages and fasteners.
CO2 cartridges mounted horizontally in opposing directions. Shutter/slicer valves. Pretty sure that Myth Busters featured a project using that technology. Something about shearing off the valve of a full size air tank - if I remember correctly?
Amazing
that's awesome.
This give me a amazing idea ima over-invent the solenoid
Interesting concept. Is this the main way you plan on controlling your rocket during flight? Engine gimballing would be much more precise, but potentially harder to to engineer on such a small scale. It would definately be interesting to see these cold air thrusters in tandom with engine gimballing at this size. Using the thrusters for major corrections and gimballing for minor corrections could save a lot of air. This is the first I've seen where it's 100% thrusters! Hopefully the maiden flight goes well!
I'd think NO2 already has spec stuff plus adding in igniters and a fuel like alcohol you could save weight. You could even mount lower and heat your lines with exhaust for better burns giving stronger control while not having to take up as much space and lift as much weight.
Bigger valves/solenoids. You could thrust vector the RCS ports, which would be a smaller profile than 8 valves. That way you could get away with only 2 valves which could be much larger and provide greater flow mass.
Have you done a thrust measurement on your nozzle setup? It doesn't look like you have a throat that is smaller than the size of the valve orifice, so there is no way you could be getting choked flow, and hence supersonic flows downstream. That means that your divergent section is actually slowing the air, and the pressure in your 'chamber' might actually be below ambient. You could check the pressure between your valve and nozzle at full flow to make sure you are getting positive pressure.
A pinhole in a thin sheet of metal would probably be simplest and reasonably efficient. It would basically be a convergent nozzle, so limited to the speed of sound, but I don't see you getting much above that anyway. When your flow rate drops to the point that you get subsonic flow it will still work efficiently, where a divergent section would be detrimental. By keeping it thin you limit boundary layer viscous losses.
I would try that before deciding whether the valves are good enough or not. You really need to design the nozzle around their capabilities. You have a full round of paintball in that tank, and your valves can discharge it in less than a minute. I don't think they are your problem.
I guess you will most likely implement PWM valve control in your next iterations. Since the rocket has some length, you could place horizontal oriented accelerometers at the bottom as well as at the top for pitch and roll. This way, you would be able to read rotational acceleration directly (as a difference between top and bottom sensors) instead of differentiate it from gyro readings (Which usually natively provide rotational speed only). Your rcs outlets produce a more or less constant force, thus they directly affect rotational acceleration. With rotational acc readings and pwm valves you could build a nice lowest level control loop. On top of that, you could then put a rotational speed and ontop of that your rot. position loop. Let me know, if i missed something. I am a bit into pids on embedded controllers as well right know and would highly recommend the magnificent Teensy 3.6 boards. The Cortex M4 has an onboard DSP with easy accessible pid processing. You might check out the community at pjrc.com.
Joe have you considered/ investigated reaction wheels for this project?
Joe I'm curious, what percentage of your time considering everything goes into making the YT content versus just the Engineering side of it? Are you able to sustain yourself on it with YT and selling parts or do you have additional outside funding? I'm anxious to do my own thing already