DX-4 Rocket Build Thread

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Joe Walsh

Well-Known Member
Joined
Mar 23, 2013
Messages
45
Reaction score
0
I've been dormant on here for a while, but I've started design work on a new finless rocket I want share with you. Some of you might remember me from my previous finless rocket project: Fly Straight Up

This new rocket will be powered by four sequenced D motors providing nearly 20 seconds of continuous thrust. Here's the design:

Overview.jpg

Stay tuned.
 
Definitely outside the box (but still inside the tube). :wink: Will be interesting.
 
Very cool project. It will be very difficult to keep the weight low enough that a single D5 or D3 will be able to lift it. Maybe kick it off with a D12 or E12 and hover with D3's?

Can't wait for video!
-Ken
 
It will be very difficult to keep the weight low enough that a single D5 or D3 will be able to lift it

That is exactly right; weight is a massive challenge for this project. Providing a "kick" was the intent of the Quest D5.

According to my bill of materials (View attachment Weight & Cost.pdf), the weight is such that the rocket will have an initial three second period of acceleration, followed by a ten second deceleration until hover (and perhaps mild backsliding), and finally a five second period of modest acceleration.

With a flight profile like that, it's difficult to say how fast or how high this thing will go. A further confounding factor is the impossibility of neatly aligning the sequenced ignitions.
 
There's been a minor setback. I designed a PCB that met the requirements of the schematic:
PCB.jpg

and yesterday the board arrived from the fab at OSH Park:
OSH Park.png

Unfortunately, I overlooked a problem in the design. The components which control the current through the ignitors are placed incorrectly. If you look at the schematic, you'll see that they are placed "upstream" of the ignitors. For correct function, they'll need to be located downstream.

I'm going to have to redesign that portion of the project, bear with me.
 
Maybe you can try using PNP transistor, like TIP125, they can switch loads on the ground side of the transistor. Not sure of if they will have pinout compatible with your board.
 
BTW, I'm very excited to follow your project along. I would like to build a rocket that looks like Delta IV heavy, but without the transparent fins modelers usually use. This project gives me hope that it's possible to build an actively stabilized rocket.

How are you going to handle the coast phase of the flight when you can no longer use the engine for active stabilization?
 
Hope that wasn't too expensive of a mistake

It wasn't too bad: OSH Park sold me three (the minimum order quantity) PCBs for $11.75.

Maybe you can try using PNP transistor, like TIP125

This would be a brilliant idea, except current on PNP transistors flows from emitter to collector. That's exactly backwards from an NPN, so I'd have to twist the collector and emitter pins around. A PNP also needs the base tied high to avoid inadvertent activation while the microcontroller is starting. My current design can't support this, but I like the way you think.
 
I would like to build a rocket that looks like Delta IV heavy, but without the transparent fins modelers usually use

Realism is the reason I got into this stabilization stuff in the first place. It's definitely possible, perhaps some things from this project will be useful for your project idea.

How are you going to handle the coast phase of the flight when you can no longer use the engine for active stabilization?



There won't be any stabilization during coast. I'm planning to blow out a chute with an ejection canister as soon as the IMU detects that the rocket is in freefall. If the rocket is moving fast enough (greater than 30 mph, I think) for significant wind drag, freefall detection won't occur until the rocket slows below this speed.

For my barely-there weight situation, freefall will probably occur as soon as the motor burns out, and there won't be much of a coast phase before the parachute gets blown.
 
The PCB has been redesigned and reordered, and I've assembled the final computer.

Computer.jpg

It weights 14.2 grams.
 
Last edited:
The motor mount and gimballing assembly have been finished:

Motor Cluster.jpg

The longer white motor is a Quest D5 that will provide the initial kick off the pad.
 
Nice! Did you design and program all the active guidance electronics yourself?
 
Did you design and program all the active guidance electronics yourself?

Yes: all the electronics have been designed and programmed by me. I'm trying to upgrade the control algorithm from my last project. That part isn't quite done yet.

I've finished the instrument unit. This will sit at the top end of the rocket. It contains the main computer, gyroscope, recovery actuator, battery, and servos. A busy little package, but lightweight (94.2g).

Instrument Unit (reduced).jpg
 
I finished installing the subassemblies into the rocket body. The gaps in the body will be sheeted with cardstock for rigidity.

Vehicle Skeleton (reduced).jpg
 
IT LIVES!

Haha, I know, I know. The project disappeared for a quite a while there.

Today I finished testing the motor sequencing code using LEDs instead of actual ignitors:

[video]https://drive.google.com/file/d/0B8B78hvl6NpySHhSZEdiazJ4UW8/view?usp=sharing[/video]
 
That's sic bro! You better not blue ball us and fly it without filming and posting! I will hunt you DOWN!
 
Haha, I know, I know. The project disappeared for a quite a while there.

Today I finished testing the motor sequencing code using LEDs instead of actual ignitors:

[video]https://drive.google.com/file/d/0B8B78hvl6NpySHhSZEdiazJ4UW8/view?usp=sharing[/video]

Hi Joe,

The video with the LEDs is really cool. One thing you might consider for your code though is to have power applied to only one igniter at a time. If for some reason the igniter leads continue to provide a current path after ignition they may bleed enough current off your battery that you get a failure to ignite on the subsequent motors. This doesn't need to complicated - you could keep it as simple as modifying your code to turn off the n-1'th igniter when you apply power to the n'th igniter, if you follow my meaning.

Very, very cool project. I'm happy to see it's still moving.
(And yes, I am still making very slow progress on my own actively stabilized project.)
 
One thing you might consider for your code though is to have power applied to only one igniter at a time. If for some reason the igniter leads continue to provide a current path after ignition they may bleed enough current off your battery that you get a failure to ignite on the subsequent motors.

Yes, I thought the same thing, i.e., all the current going to one igniter at a time. Of course, if each motor fires as it should, the previous igniter will be gone and only one igniter will be fired at a time.

The plastic parts really look good. Are they machined or something like 3-D printing? Like others I would really like to hear how this project flies. This is a really interesting concept.
 
If for some reason the igniter leads continue to provide a current path after ignition they may bleed enough current off your battery that you get a failure to ignite on the subsequent motors
This is very true. Originally I was thinking the on-time should be much shorter than simply the period between ignitions, but I think you're right: just disengaging the previous one when the next one engages should be sufficient. Best of luck to you on your project, Kevin. I'm always interested in finding other people who are doing this type of stabilization: do you have a build thread of your project somewhere?
The plastic parts really look good. Are they machined or something like 3-D printing?
The motor housing and gimbal parts were 3D printed using a Stratasys Mojo. The connectors that the ignitors (in this test LEDs) plug into are spring wire connectors like these: https://www.aliexpress.com/item/10-Pcs-2P-Spring-Connector-LED-Strip-Light-Wire-Connecting-No-Welding-No-Screws-Quick-Connector/32359478839.html.
 
Progress has been very slow lately, but I just finished the test stand and am preparing to do a couple engine runs.

Test Stand (reduced).jpg
 
Back
Top