I could use just a little guidance

[Finicky]

Beautiful flight. It really did look like things were being controlled very smoothly. When you can modulate tilt to 1 degree you really are locked in.

 

[VTS SkunkWorks]

I did a VOS2 flight Saturday at San Angelo. It's a pretty new field that is really nice. West Texas doesn't have a lot of trees.

The purpose of the flight was to start investigating the use of the system for roll control only. When controlling yaw/pitch, the main objective is just to keep the rocket from spinning rapidly and from exceeding the gyro range. For roll-only, the objective would be for video. Since I haven't previously done roll control only, I decided to start with lower gains than we typically use, knowing that these will need to increase to actually hold at the desired bearing through the flight. So, I would say the flight was reasonably successful and there is potential for significant improvement. There is a video of the flight linked below with some data at the end of the video.

Speaking of the VOS2 systems, the second batch of six systems has been claimed. I'm not sure if it will be possible to make more - I hope so - but there is uncertainty just now. Our group should be able to do quite a few flights this year, so I'm looking forward to that.

Jim

My VTS III rocket with control surfaces on the aft fins. The first version with ply fins had structural failure of two fins, but did well until then. The followup has aluminum fins and smaller control surfaces.

 

[kjhambrick]

VTS SkunkWorks --

Very interesting idea for attitude control !

Do you have a build thread here on TRF ?

-- kjh

 

[VTS SkunkWorks]

VTS SkunkWorks --

Very interesting idea for attitude control !

Do you have a build thread here on TRF ?

-- kjh

No build thread, sorry. I have been doing VTS since 1995. The first one used Futaba rate servos with actual spinning mass inside!

I used an EagleTree MicroVector flight controller which also has memory. That was driven by a tiny zero arduino. The computer turns everything on and runs a control surface test prior to initialization.

The first version used small linear servos surface mounted on the fins. I upgraded to standard nano servos mounted through the fin. The upgraded fins were two sheets of aluminum glued together with JB Weld with a third thinner layer of aluminum inside like a sandwich. I did that so I could sandwich the hinges. The inner layer of aluminum is the same thickness as the hinge material. The hinges and linkage are large scale RC aircraft equipment. The fins are connected with a "fin can" of flanges so I can change the fins for development.

It has a 54mm motor mount. Obviously I used a plugged motor with tripple redundant recovery.

The aft control system was suggested by Ken Biba so the controled section could make a booster for a 2 stage rocket. That way stage 2 would be ignited while vertical.

I would be glad to answer any questions.

BTW, what I have seen on this forum is beyond what I have done, and much more skilled construction!

 

[JimJarvis50]

No build thread, sorry. I have been doing VTS since 1995. The first one used Futaba rate servos with actual spinning mass inside!

I used an EagleTree MicroVector flight controller which also has memory. That was driven by a tiny zero arduino. The computer turns everything on and runs a control surface test prior to initialization.

The first version used small linear servos surface mounted on the fins. I upgraded to standard nano servos mounted through the fin. The upgraded fins were two sheets of aluminum glued together with JB Weld with a third thinner layer of aluminum inside like a sandwich. I did that so I could sandwich the hinges. The inner layer of aluminum is the same thickness as the hinge material. The hinges and linkage are large scale RC aircraft equipment. The fins are connected with a "fin can" of flanges so I can change the fins for development.

It has a 54mm motor mount. Obviously I used a plugged motor with tripple redundant recovery.

The aft control system was suggested by Ken Biba so the controled section could make a booster for a 2 stage rocket. That way stage 2 would be ignited while vertical.

I would be glad to answer any questions.

BTW, what I have seen on this forum is beyond what I have done, and much more skilled construction!

I like your approach. It solves many of the problems associated with canard control and is a nice implementation of rear control. I've seen some "flat" servos that could almost be embedded completely within the fin.

Jim

 

[kjhambrick]

I like your approach. It solves many of the problems associated with canard control and is a nice implementation of rear control. I've seen some "flat" servos that could almost be embedded completely within the fin.

Jim

Hmmm ...

Flat servos embedded completely within the fin -- VERY interesting idea for just 'a little guidance' !

-- kjh

 

[OverTheTop]

Hmmm ...

Flat servos embedded completely within the fin -- VERY interesting idea for just 'a little guidance' !

-- kjh

Look up "wing servos".

 

[kjhambrick]

Thanks OverTheTop.

They're not as slim and I had imagined ( hoped ), are they ?

-- kjh

 

[VTS SkunkWorks]

I figured after I got it working properly, I might add some 3d printed aerodynamic covers over the servos.
The Nano servos can run at a higher voltage than the linear servos, so they are easier to integrate with the flight controller.
The Controller is set for 2 opposite fins to be "rudder" for yaw control and two opposite fins as "elevons" to provide pitch and roll control.

If you wanted a type of thrust vector included you can add graphite tabs to the fins extended into the exhaust flow, much like the redstone rocket.
However, that did not seem to be necessary as you can see in the video.

The purpose of the VTS system was to allow the rocket velocity to be close to the optimum and still go straight up without a gravity turn.
Check out Apogee Newsletter # 455, 456 and 457 for details on optimum velocities.

 

[JimJarvis50]

I figured after I got it working properly, I might add some 3d printed aerodynamic covers over the servos.
The Nano servos can run at a higher voltage than the linear servos, so they are easier to integrate with the flight controller.
The Controller is set for 2 opposite fins to be "rudder" for yaw control and two opposite fins as "elevons" to provide pitch and roll control.

If you wanted a type of thrust vector included you can add graphite tabs to the fins extended into the exhaust flow, much like the redstone rocket.
However, that did not seem to be necessary as you can see in the video.

The purpose of the VTS system was to allow the rocket velocity to be close to the optimum and still go straight up without a gravity turn.
Check out Apogee Newsletter # 455, 456 and 457 for details on optimum velocities.

Your system could be coupled with the flight computer and program that I use. You would need to be able to mount the board in the pic vertically in the air frame (it's 1.5" x 2.75"), and the servos would need to operate with PMW signals at a 1500 center point. Want to give it a try?

Jim

 

[VTS SkunkWorks]

Jim, very nice controller! I am 72 and retired from flying the VTS rockets. Although I would prep the rockets at home, and load the motor, it was too much to do at the pad for one person! I passed the rocket on to Tom Fetter who also flys a VTS system in N CA.

Your controller looks much cleaner than what I was using. Great work! Hve you flown your system?

 

[JimJarvis50]

Jim, very nice controller! I am 72 and retired from flying the VTS rockets. Although I would prep the rockets at home, and load the motor, it was too much to do at the pad for one person! I passed the rocket on to Tom Fetter who also flys a VTS system in N CA.

Your controller looks much cleaner than what I was using. Great work! Hve you flown your system?

Well, we're on Post 912 of the development thread, and along the way, I've flown the system roughly 35 times. Can't wait for the next flight - it's fun.

You can find many flight videos along this thread, or search YouTube for Jiminaus50.

Jim

 

[VTS SkunkWorks]

Well, we're on Post 912 of the development thread, and along the way, I've flown the system roughly 35 times. Can't wait for the next flight - it's fun.

You can find many flight videos along this thread, or search YouTube for Jiminaus50.

Jim

Great video's thanks for the connect. I liked your 15k my way video!
Are you familiar with "the Goddard problem"? I wrote some software to solve part of that iteratively for amature rocketry. The solution is different for each rocket but it computes the velocity profile that will get you the highest for a given total N-S motor. That is what is discussed in the newsletters I mentioned above.

Thank you for sharing the lessons learned and construction techniques!
steve ainsworth TRA 2454

 

[VTS SkunkWorks]

Jim
If you have a Rocksim design file for your rocket, I can use that to provide the drag info needed to get the optimum velocity by he second and then the thrust profile needed to make that happen. I use an output file generated by Rocksim as the input.

Steve

 

[kjhambrick]

Well, we're on Post 912 of the development thread, and along the way, I've flown the system roughly 35 times. Can't wait for the next flight - it's fun.

You can find many flight videos along this thread, or search YouTube for Jiminaus50.

Jim

I've been studying your thread for WEEKS and I am still only up to Page 9 of 31 !

Thank you for sharing your hard-won experience and knowledge, Jim !

-- kjh

p.s. one of my favorites -- the one that explained 'just a little guidance' for dummies ( i.e. me ) is Post # 176 -- Diagram: Atttitude with Canards

 

[JimJarvis50]

Great video's thanks for the connect. I liked your 15k my way video!
Are you familiar with "the Goddard problem"? I wrote some software to solve part of that iteratively for amature rocketry. The solution is different for each rocket but it computes the velocity profile that will get you the highest for a given total N-S motor. That is what is discussed in the newsletters I mentioned above.

Thank you for sharing the lessons learned and construction techniques!
steve ainsworth TRA 2454

Thanks Steve. Seems like I know you from somewhere?

I saw the articles a while back. I have some rockets that could get optimized for altitude, but my Test Rocket isn't really one of them. That one flies odd flight profiles, like short-burn booster motors, long-burn sustainers, etc. I'm ususally trying to make it not go high!

Thanks for mentioning the 15K video. I spent a lot of time on that, mainly to have some reference information for the college teams I mentor. Earlier this year, I did a NARCON video on the active stabilization system. I think NAR will let me link to it before too long.

Jim

 

[JimJarvis50]

New fin tips for the next flight.

Jim

 

[kjhambrick]

Jim --

When, where and what do you plan for your next flight ?

Thanks

-- kjh

 

[JimJarvis50]

Jim --

When, where and what do you plan for your next flight ?

Thanks

-- kjh

Sometime over the next three weeks or so, I'm planning a single-stage flight to look specifically at improving roll control. This is a follow-on to the San Angelo flight, but hopefully, with gains that result in better control. I mentioned it earlier in this thread, but roll control has both a rate gain (to slow down the roll rate when there is roll) and heading hold (to return the roll angle to that at the start of the flight). The two roll gains are tied together.

Conceptually, if you only have heading hold and no roll rate control, and if the spin can was frictionless (which it isn't), the effect of heading hold only would be washing machine-like oscillations. At the other end of the spectrum, if the roll rate gain was very high, it would fight the attempt of the heading hold to return to the initial roll angle. So, there is an optimum, where the heading hold can quickly return the roll angle to the initial without overshooting or oscillaing too much.

I did a simple analysis looking at how the two gains work together to correct a 90° error in the roll angle. The attached graphs show how quickly the system returns to the initial roll angle and the roll rate of the rocket during the correction. The first graph shows the default settings, and the calculations suggest that it will take 1.5 seconds to make the correction. By reducing the roll rate gain (the second graph), the time required can be reduced to 0.6 seconds without too much residual velocity. So, that's what I'm testing.

The analysis is very simple, assumes frictionless operation of the fin can, and ignores things like the response time of the servos. We are working to incorporate those things into a quantitative analysis, but for now, I can get some data flying the rocket.

Jim

 

[Rschub]

Your system could be coupled with the flight computer and program that I use. You would need to be able to mount the board in the pic vertically in the air frame (it's 1.5" x 2.75"), and the servos would need to operate with PMW signals at a 1500 center point. Want to give it a try?

Jim

Hi Jim, I have been following your thread for a while now. I am curious, do you have a supply of these controllers stashed away? Because I can't seem to find them anywhere.
There is a project that I am planning down the road that involves vertical stabilization, which I why I am asking.

 

[JimJarvis50]

Hi Jim, I have been following your thread for a while now. I am curious, do you have a supply of these controllers stashed away? Because I can't seem to find them anywhere.
There is a project that I am planning down the road that involves vertical stabilization, which I why I am asking.

No, I don't have a supply, although I have a few extras. If I had more, I would make some additional systems. I'm hoping a new version of the control board will be produced, and if not, there are some other options.

Jim

 

[JimJarvis50]

Jim, very nice controller! I am 72 and retired from flying the VTS rockets. Although I would prep the rockets at home, and load the motor, it was too much to do at the pad for one person! I passed the rocket on to Tom Fetter who also flys a VTS system in N CA.

Your controller looks much cleaner than what I was using. Great work! Hve you flown your system?

As it turns out, I'm currently working with Tom Fetter on analyzing the roll control problem. He certainly knows the subject. Small world.

I did another flight today, where I was looking at optimizing the roll gains. I think there was significant improvement.

Data later on.

Jim

 

[JimJarvis50]

So, here's some data from the most recent flight (the South Texas flight). It was done on an L-1000 for the purpose of improving roll control. As discussed in Post 919, there are two types of control that contribute to roll control. The gains for them need to be considered together to get the best response. I made a change that should have allowed a faster response. Per the graphs in Post 919, the change should have reduced the time to correct from a 90° error from 1.5 to 0.6 seconds. There was an initial error of that magnitude on this flight, and the response time was very close to the prediction (it was about 1.0 seconds, likely a little slower than 0.6 seconds because the spin can is not frictionless). I think further reduction in the response time is possible.

Overall, there was a general improvement in the quality of the flight. Less jerkyness for lack of a better term. The second graph below compares the San Angelo and South Texas flights, and shows the roll rate and the roll angle. For the South Texas flight, the roll angle was closer to the set point and roll rates were lower. There was also improvement in the yaw/pitch control, which I suspect is due to being closer to the desired roll angle for more of the time during the flight.

We now have the capability to do gain scheduling for all of the system gains. So, this can be used to improve the roll response during the early and late parts of the flight where the velocity is low. I look forward to testing that capability on future flights.

The flight video is also linked below.

Jim

 

[OverTheTop]

Looking great Jim. The gain scheduling will be a game-changer too.

 

[JimJarvis50]

Looking great Jim. The gain scheduling will be a game-changer too.

Agree. The only limitation, which we probably can't get around, is that gains can't be modified until roughly a half second into the flight. That's the typical time required for launch detect. Maybe not that big of a limitation?

I think it will also be necessary to think out of the box a little on how to use this. Just as an example, would it make sense to go with larger canards to get more correction at slower speeds, given that their movement can be limited at high speed?

It will be fun to try this out, and I'm open to suggestions on how to employ the technique.

Jim

 

[OverTheTop]

My big concern with going for larger canards is that it will be able to turn the rocket around and have it head potentially towards the flight line, under power. My system suffered a servo failure early in one flight (caused by a solder-ball inside the COTS voltage regulator ) and although it went more sideways than it should have it was all safely happening up in the air. Recovery was nominal.

My system limits canard throw to about +/- 5 degrees, by mechanical endstops. I also keep the servos centered until after boost. That way if something goes wrong it is at altitude and the recovery system will work as usual.

For an experimental system I would want some straight flight and velocity before attempting any steering. Roll control is more benign and can probably be enabled at launch safely.

 

[JimJarvis50]

The NAR posted a video of my presentation on the Vertical Orientation System at NARCON this past January. It just goes over the development and capabilities of the system and some of the interesting flights along the way. I hope the quality of the presentation itself was better than the video!

Jim

 

[kjhambrick]

Thanks for posting your NARCON presentation, Jim.

I learned a lot about the VOS that I did not know from reading here and watching your Videos on YouTube.

-- kjh

 

[VTS SkunkWorks]

The NAR posted a video of my presentation on the Vertical Orientation System at NARCON this past January. It just goes over the development and capabilities of the system and some of the interesting flights along the way. I hope the quality of the presentation itself was better than the video!

Jim

Jim
Just found time to watch your NAR presentation video. Thank you for taking the time to do the video and post it here! Outstanding presentation and quite informational. You have developed a very impressive system! As my flights got faster, I found the same issue that you did with canard supports, see picture. Did you fly at club launches? Any issues with LCO's?

 

[JimJarvis50]

Jim
Just found time to watch your NAR presentation video. Thank you for taking the time to do the video and post it here! Outstanding presentation and quite informational. You have developed a very impressive system! As my flights got faster, I found the same issue that you did with canard supports, see picture. Did you fly at club launches? Any issues with LCO's?

Thanks Steve. Yes, lots of lateral force on the canards. When we figured out how to protect the servos from that force, things started working.

I think I will have to agree - the system is impressive. This is mostly due to the efforts of Dr. Premerlani. We've been busy over the past few months. It seems like every time I think we've reached the limit of the system, we make another step in development. We now have the capability to gain-schedule the various control parameters with up to 100 changes over the duration of the flight. I'm not even sure 100 is a limit. We have had the capability of changing the tilt and bearing through the flight, again up to 100 times, and the most recent addition is the ability to control the roll angle. All of this is easily programmable, and even I can do it. Now, we just need to figure out how to use it, but it's really just a matter of imagination.

I guess I've flown the system at perhaps a dozen fields over time, and there has not been a problem with flying the system at any of them. I've had a couple of interesting flights that were due to staging issues, but only one flight perhaps five years ago where the rocket didn't go vertical due to a control system issue. I would argue it's the safest rocket at the field because I have a reasonable idea where it's going to go. The beta group has been flying the system at some new fields, and I'm not aware of any issues with LCOs at those fields either.

Jim