Complex Launch Controller

The Rocketry Forum

Help Support The Rocketry Forum:

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

KilroySmith

Well-Known Member
Joined
Oct 28, 2015
Messages
183
Reaction score
112
<Moved from Rocketry Electronics>

Has anyone come up with a good launch controller design that can differentiate a shorted igniter from a good igniter?

I was redesigning my 8-station launcher, and was thinking that a short detector would be a great addition to it. Unfortunately, the measurement is extraordinarily difficult, especially when you're trying to not launch the rocket while you're testing it. I figure that a short will show up as roughly 1 ohm, and a good igniter will show up as about 2 ohms, and I've got about 1 ma to do the measurement with.

Any thoughts that don't involve a budget that makes me forego my next rocket build?

/frank

p.s. The pisser is that the chip we make at work would be the perfect solution - it's designed to measure microvolt changes to millivolt level signals and I can get them by the handful, but:
1. I don't know how to do a cheap chip-on-board attachment
2. I was planning on publishing the design in the NAR mag for others to use, and they certainly wouldn't be able to get the chips.
 
We had a presentation on launch systems at our meeting earlier this month: https://www.tripolimn.org/sites/test.tripolimn.org/files/MLaunch 3.0.pdf
Thanks - that looks like a great effort. Putting in the micro is Cheating - cheap, small, and solves a lot of problems. Unfortunately, I'm not that smart so I'm trying to avoid a micro. It's not obvious from the presentation, but shows up in the goals - is it capable of measuring igniter resistance? Thanks /frank
 
https://www.thrustgear.com/LCS02/

The continuity test portion of my design converts the test current into a tone. The frequency is high for a short, medium for good igniter, and low for a poor igniter. Also, intermittent tone for loose/dirty clips. I did this using a small PIC microcontroller (8 pins). Program is free.

I suspect this may be more complex than you want to build. Some clubs have built them over the years.

Here's my simple relay controller with an alarm for a stuck relay. https://thrustgear.com/relayer.pdf
It doesn't have the fancy continuity circuit. But, it uses cheap power cords, has remote continuity test, and thermal overload protection for dead shorts.
 
Back in the 90's you could buy a Verifire launch controller from Impulse Aerospace that had short detection.

kj
 

Interesting design. I like the tone-based continuity detection, and the 2-wire interface to the pad box. If I'd seen this before I got started, I probably would have just used it, although my goal was more for introductory launches than club launches, and I'd miss the ability to fire off a salvo. But now I'm in too deep to this one now....
 
Unfortunately, I'm not that smart so I'm trying to avoid a micro.

Don't count yourself out on the offset, microcontrollers (particularly with arduino libraries) are amazingly easy to work with. Most of the time they make projects easier rather than harder, even for microcontroller neophytes.
 
Don't count yourself out on the offset, microcontrollers (particularly with arduino libraries) are amazingly easy to work with. Most of the time they make projects easier rather than harder, even for microcontroller neophytes.
I have particularly high standards when it comes to working with things that might go boom. I don't want to put in the time necessary to convince myself that the code has 5 9's of reliability; I know how easy it is to write code that has 2 9's of reliability, and convince yourself that it's "good enough".
 
I have particularly high standards when it comes to working with things that might go boom. I don't want to put in the time necessary to convince myself that the code has 5 9's of reliability; I know how easy it is to write code that has 2 9's of reliability, and convince yourself that it's "good enough".

My approach is to include a hardware safety interlock. Right now I am working on a design that uses RS485 for the communications link and I use the common mode voltage for the safety interlock. Before a pad can arm it has to get both a command over the serial link and the hardware enable. That hardware enable has no software involved in any way.

As for detecting a short, that depends on what your criteria for a short is. I recall that the Kloudbuster's launch control system has a safety interlock that prevents launching if it detects a short. Some igniters (Fire In The Hole) would trigger that system. Fortunately, it had an override switch.

Detecting a short is made more difficult because the threshold is likely to be less than 1 Ohm which means that you really should have a four wire measurement system to take the wire resistance out of the picture. But I think that would be far more effort than it is worth.
 
Well, after fighting with a lot of different ideas, I've come up with a good way to measure igniter resistance, and be able to differentiate shorts (especially when using Estes Solar igniters) vs. low resistance igniters. Unfortunately, it's complex - a lot of parts, but for each channel you'd get:
1. Dedicated LED indicating a short (< 0.5 ohms, as the smallest igniter resistance I've seen is about 0.6 ohms).
2. Dedicated LED indicating a good connection (>0.5 ohms, <4 ohms).
3. Dedicated LED indicating a fault condition (i.e. Voltage at the igniter connector when there shouldn't be any).
4. Compatibility with 9.6V RC batteries, 11.1V 3S LiPo batteries, 12V Gel Cells, and probably 14.4V NiMh batteries, though the 16.8V fully-charged voltage might overstress some of the components.

This works assuming a rational choice of leads going from the controller to the rocket - I use 6', 24 AWG wires at about 0.3 ohms. It's unlikely anyone would use much thinner, nor much thicker, so that'd probably work out reasonably well.

I'm looking at a dozen resistors+diodes, three transistors, three LEDs, and a couple of $1 chips per channel. It's a triumph of brute force engineering. I just don't know that it could be laid out on a board compatible with the free version of Eagle, or that anyone else would ever have the patience to build one. All to avoid a $2 microprocessor and some programming - because, from what I've seen, requiring a micro eliminates most of the people who might build one. Sometimes my sense of engineering aesthetics keeps me from actually getting anything done.

No real questions here, just spewing after staying up way too late last night putting on the final touches...
 
Well, after fighting with a lot of different ideas, I've come up with a good way to measure igniter resistance, and be able to differentiate shorts (especially when using Estes Solar igniters) vs. low resistance igniters. Unfortunately, it's complex - a lot of parts, but for each channel you'd get:
1. Dedicated LED indicating a short (< 0.5 ohms, as the smallest igniter resistance I've seen is about 0.6 ohms).
2. Dedicated LED indicating a good connection (>0.5 ohms, <4 ohms).
3. Dedicated LED indicating a fault condition (i.e. Voltage at the igniter connector when there shouldn't be any).
4. Compatibility with 9.6V RC batteries, 11.1V 3S LiPo batteries, 12V Gel Cells, and probably 14.4V NiMh batteries, though the 16.8V fully-charged voltage might overstress some of the components.

This works assuming a rational choice of leads going from the controller to the rocket - I use 6', 24 AWG wires at about 0.3 ohms. It's unlikely anyone would use much thinner, nor much thicker, so that'd probably work out reasonably well.

I'm looking at a dozen resistors+diodes, three transistors, three LEDs, and a couple of $1 chips per channel. It's a triumph of brute force engineering. I just don't know that it could be laid out on a board compatible with the free version of Eagle, or that anyone else would ever have the patience to build one. All to avoid a $2 microprocessor and some programming - because, from what I've seen, requiring a micro eliminates most of the people who might build one. Sometimes my sense of engineering aesthetics keeps me from actually getting anything done.

No real questions here, just spewing after staying up way too late last night putting on the final touches...

How are you sensing the resistance and tripping the 3 conditions?

My 1996 circuit converts the resistance to a tone. The pitch of the tone indicates the resistance (high pitch = short; medium pitch = good igniter; low frequency = poor igniter). It uses two chips: a dual opamp and a 556 (dual 555) timer. In 2002, I simplified the circuit using an 8-pin PIC microcontroller instead of the 556 . Here's the schematic. That fits on a 1x2" board. The improved tone has a sweeping alarm sound when there's voltage at the igniter. The sounds get the flyers attention instead of having to look at LEDs.
 
Well, being as I'm planning on building this into an 8 channel controller, I'm not sure that tones are the right way to go...

I was using a precision 1 ma current source (like a PSSI2021SAY, about $0.50) to drive a known current through the lead and igniter. This gives us roughly 1-2 mv across the two. Then, we use something like an INA212 (about $1.50) as an x1000 gain to give a reasonable voltage range to work with - 1-2 volts. Then we grossly abuse a PNP transistor (used to light the LED), using a zener diode of the appropriate voltage on the emitter to set the turn on voltage relative to this voltage range - changing the voltage of the zener changes the turn on of the transistor by about 0.1V for every 0.1 ohm change in igniter resistance. There are a lot of zener voltage choices at Digikey. If you want a 1 ohm range between the short and continuity LEDs, you need a 1 volt difference in Zener diodes.

I haven't built this yet, but it simulates well in LTSpice. Just counted - six resistors, three LEDs, three transistors, two zeners, and the two chips. I guess that's not that much worse than a Micro based approach - which would have to be at least three LEDs and three resistors, assuming the micro could drive the LEDs at 20 ma without an external driver.

I do have a 555 in the design to cause a "chirp" at the pad when the key is on at the control panel. I figure you ought to be warned if you forgot to bring the key with you....
 
Sounds like a good plan for the igniter sensing and LED driving.

Why are LEDs better than tones for an 8-ch controller? I've used tones on all sizes of launch systems, from 1-pad to 32 pads.

The nice thing about the microcontroller is that you can change your thresholds without soldering components, and you can produce tones and/or blink rates for whatever logical case that may come up later on.
 
On the tone question, how do you distinguish tones from 8 different launch rods simultaneously? Or do you have a switch to select only one to listening to at a time? There are a lot of advantages to a micro - it'd be both cheaper and less PCB area than this design. But, there are three problems I have with them: 1. The number of people who would be willing to build this for their own use, in my opinion, drops drastically the moment "micro" is mentioned. Finding and installing tools, compiling code, figuring out how to reflash - it's a pain if you don't know what you're doing (and is a pain even if you do...) 2. I don't want to put in the effort to make the code reliable, even though (in this design) the micro would be purely a monitor and wouldn't be in control of launching the rocket. 3. I've spent 30 years as a professional firmware engineer, and I don't want to do that in my spare time. Think of this as my "Steampunk" approach to high-tech rocket launching.
 
Just about every club launch system I've used for 25+ years has had a continuity check button. The tones are only on when you select the pad and press the continuity check button. This is either on the main controller at the LCO table and/or at a distribution box near a cluster of pads. In addition, some club launch systems or individual launch controllers have a box at each pad that allows the flyer to check continuity there. It could be a momentary button for a quick check, or a switch that allows hearing the tone status all the time when pads are separated by some distance (HPR). The alarm tone should be active regardless of whether the flyer pushes the continuity button or flips the switch... you should get a warning approaching a pad that the relay is fried or wiring is bad before hooking up the clips.

No need for a micro, it's just an option for more functionality. But, it is dead simple software, especially for someone with your experience. For others who don't have the experience (or the desire), someone could flash the microcontrollers and send them to whoever wants to build one. It's just another 8-pin $1 chip at that point.

I'm assuming that you are posting here because you'd like some feedback, advice, and ideas. If not, you'd be building the thing instead of wasting your time on this forum. :wink:

For steampunk, I'd use a knife switch, knob&tube wiring staked along the ground, a crank-up charger from 1930's phone, big alligator clips, and horn from a Studebaker for a launch alert. And wear leather overalls and a Kaiser helmet with a spike on top to make it official.
 
For steampunk, I'd use a knife switch, knob&tube wiring staked along the ground, a crank-up charger from 1930's phone, big alligator clips, and horn from a Studebaker for a launch alert. And wear leather overalls and a Kaiser helmet with a spike on top to make it official.

Now THAT needs a build thread!!!

You could title it "My Mag-neato Launch Controller"

No...seriously, I know at least 2 of us that want to build one :p
 
Last edited:
For steampunk, I'd use a knife switch, knob&tube wiring staked along the ground, a crank-up charger from 1930's phone, big alligator clips, and horn from a Studebaker for a launch alert. And wear leather overalls and a Kaiser helmet with a spike on top to make it official.

Now that raht there was downright funny.

...you'd like some feedback, advice, and ideas...
Damned straight on that; keep'em coming. I've stolen some ideas from the presentation that thorfire linked to in the second post (which has apparently gone missing already; but "doesn't fire if the system is power cycled" is a great requirement that came from there). I've stolen one or two things from yours also.

Just like launch pads, where everybody seems to like to put their own bit of personality into their build to scratch their particular itch, that's my case here. This is mostly an intellectual exercise acting as relaxation therapy, but I'll build it at the end because I like the physical instantiation of those kinds of exercises. And I'll be pleased with it.

My goal isn't to replace the local club launch system. My focus isn't on large numbers of geographically seperated pads for serious rocketry. My focus is on a park setup for a dozen cub scouts, where with two boxes, a battery, and a length of 2x4 I can launch two dozen rockets an hour, including a salvo of everything still in one piece as the grand finale. My goal here is to make the simplest piece of gear that makes that environment run smoothest. In that environment, unfortunately, Estes solar igniters are common, and, unfortunately, shorts at the igniter are common with those. That's the original problem I came here for help with.

Of course, if I'm going to use this as my own launcher, I expect it to light anything I attach it to, so the driver section is capable of driving enough current out the jacks that you could jump start a tractor with it, but that simply requires a little bit of attention during design/layout and doesn't really increase the complexity of the project - a quarter-inch wide trace is only slightly harder to run than a 10 mil trace. I'm working on an alternate driver section that supports cluster launching - where all eight outputs are synchronized, and you have the individual short/continuity checks on each igniter that you lose with a cluster whip. It just requires designing the box for a 60 amp or so current surge - no problem!

I really do appreciate the suggestions you've made here. It's been helpful to me.
 
Back
Top