Wireless Launch Control

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Well, I'm stoked. I spent some of the long weekend building the four field boxes and attending to other details such as fabricating RP-SMA extension cables. I was able to get all the circuits configured and it's all working!

Tonight I had to program the wireless radios. That was interesting because the newer radios I bought had firmware two releases ahead of what I have been using. So I had to back level the radios first, and then load the configurations for the net controller and the router nodes.

Since these are two side by side 8 pad wireless networks I had to vary the parameters a little bit. The good news is everything is working great. I previously tested all the hardware using the radios from my testbed. So I knew that all worked.

The new radios establish two all new AES encrypted mesh networks. Of course some work did go into developing those original configurations. Minor tweaks is all that was needed to create the new networks. The photos show everything up and working side by side. Though it's hard to tell. I did light a few home made ignitors to try it out.

The field boxes are not in their "boxes" yet. and they have some temporary stubby antennas attached to the radios vs the higher gain external antennas. Great for testing. Two of the field boxes have battery harnesses for the internal 14.8v LiPo batteries, or they can use external batteries. The other two are set up for external batteries.

So still some more work to do. Adding mounting brackets to the field boxes to hold it all together is yet another challenge to overcome.

IMG_0945.jpg

IMG_0947.jpg
 
So i've been noodling about how to make a wireless bank controller. My ideal specs are minimum 8, up to 16 bank switches on the front, plus 1 to 4 network select switches on the front or back.

The wireless is easy. It's all just addresses, The mesh networking will find the endpoints. But how to deal with all the inputs and outputs on the bank console itself? Talking switches and LEDs.

Discrete logic is fun but the chip count goes up fast when you start to get creative. There are some specialized serial latch options for display output but there's the chip count again. Especially when you are trying to drive LEDs.

So I'm ending up with the same conclusion as when I designed the panel and display controllers. Chained MCUs.

A 16 bit latch still has about 4 or 5 input pins, and you still need to drive the LEDs with something. An MCU instructing another MCU over a single pin can drive many LEDs. That's how the panel->display controller works.

I still have to calculate all the possible inputs and outputs. Maybe 20 and 20+, respectively. So two MCUs makes sense, overkill aside. And the MCU usually can drive LEDs directly, which is a big plus.

Theoretically, an 8 pad controller plus an 8x bank switch will control 64 pads (16 field boxes). 8 pads plus an A-B network switch doubles that. And the math keeps going.
 
Well so much for reusing the LCO master controller board. Chained MCUs for a wireless bank select panel will require a new circuit board design.

I think I'll design the circuit for up to 16 front, 4 back inputs (switches), and about the same for outputs (LEDs, sonalert, heartbeat etc).

But the prototype will be just for 8 banks. I hate wiring switches.

So 64 and up to 512 pads. I won't be the one building 128 field boxes though. Maybe 8.
 
About a month and a half since I started the 16 pad wireless build. And it is complete. Assembled and tested. Pictures soon.
 
The controller uses an operating system that employs task switching, and I have about a half dozen tasks (processes) and a small number of semaphores. With a 20mhz processor the problem is not cycles, but task logic more likely.
I'm just curious as to what processor and OS you are using?


Thanks,

Allan.
 
The processor (one of them) is a PIC18F2620. The OS is the Sourceboost Novo RTOS for C++. And the 2620 at the LCO end runs at 40mhz now.
 
The 16 pad wireless system was finished about a month ago but I am finally getting around to a picture. Though not a great one.

I'm still working on how to label the field boxes. Will probably do some kind of vinyl labels. Have to number the pads at least.

IMG_1056.jpg
 
Time to continue winding down. I have many parts and circuit boards left over so I'm putting those together. These are six relay modules and six wireless interface modules. Lot of flavors of panel modules still to build.

A wireless bank switch is still in the thought stage but I think eventually these relay modules will be the most valuable.

IMG_1058.jpg
 
I decided to do a little work toward the wireless bank controller. It will add a lot of flexibility to the current system. Though might be a little awkward for the random volunteer LCO to use.

I decided to design the panel first, so I know at least how many switch and indicator inputs and outputs I have to deal with. The attached pdf shows the layout.

The 8 banks A though H are clusters of at least 8 wireless pads each. So the range is 64 pads basically. I played around with the range select quite a bit. I wanted something to extend the initial 8 banks, but not make it too complicated. I don't know if I succeeded.

The 'range' obviously is a set of addressable 'banks'. Including the default range, this layout supports 5 ranges of 64 pads, or 320 pads. It will also work with the 16 pad controller I built, for 8 banks of 16 pads per cluster, or 640 pads. But it's about flexibility, not numbers. Two low power ranges, for example, left and right. Plus a mid power range, plus a high power range. Or something like that.

The range labels are arbitrary. I tried phonetics, ie Alpha, Baker, Charlie, Delta; and colors, ie Blue, Green, Yellow, Red; but didn't like the way that looked on the panel. This is my quasi test bed in any case. So I just defaulted it for now. A piece of gray elecrical tape and a sharpie is all one needs to custom relabel the ranges for a launch.

I have to design a new circuit board to handle all these inputs and outputs, plus the wireless. Hopefully just one MCU. Plus the usual scaffolding. That's going to take a little while so don't look for frequent updates.

Let me know what you think, or not. Enjoy.

View attachment Bank Switch Panel Layout.pdf
 
Me too! I need to stop building this cra* and start building rockets again. Or at least paint the ones I haven't finished. Or repair the broken ones.
 
Nice!

Time to continue winding down. I have many parts and circuit boards left over so I'm putting those together. These are six relay modules and six wireless interface modules. Lot of flavors of panel modules still to build.

A wireless bank switch is still in the thought stage but I think eventually these relay modules will be the most valuable.
 
A couple of custom panels for the bank/range controller arrived today. The small panel is the rear panel. It's the same as my other rear panels.

I did some work on defining and assigning all the inputs and outputs this controller needs. Lots of switches and LEDs and a few other goodies. Good news it will all fit on a single PIC 18F4620 which is the same chip I use in the relay boxes. So I can copy some of that circuit and PCB design, saving some work. Need to get started on that.

One advantage to the bank/range controller that I realized is there is now a second wireless node at the LCO end. The wireless modem can only directly associate with about 12 or so other nodes, those can associate with about 12 more, etc. So with two wireless nodes at the LCO there is about twice the available associations for the close in pad units. Of course they will all form their own indirect network associations far and wide as needed but it's nice to have as many as possible associating directly for lowest latency.

IMG_1062.jpg
 
Excellent work! The functionality is amazing, even more so than the design work. Wow!



All the best, James
 
Here's the completed schematic (finally) for the bank/range controller. Pretty unremarkable really. 12 switches in, 8 LEDs out, plus a few odds and ends. The wireless module plugs into the 5 and 10 pin headers J11 and J13. Select switches will connect to J4 & J5, and the enabled LEDs will plug into J6. Outputs for the power LED, heartbeat LED, and a small sonalert. The fixed and +12v references are so the controller can measure battery voltage in real time via a few analog to digital converters. Overkill but if they're there, why not use them.

View attachment Controller v2.1 (2.0z) Bank & Range Select Rev 0.pdf
 
Yeah, right. I did experiment with some MCU protoyping modules early on. Didn't really work for me once I got past the "blink the LEDs" stage.
 
The circuit boards will arrive Thursday. In the meantime I popped some switches and LED lenses into the panel.

IMG_1064.jpg
 
Back in the days I used the poor kid method of wireless launch controller by using a broke down RC car. :)
 
An attempt at a glamour shot while I wait for new circuit boards to arrive. If you do the math, this is a powerful combination.

IMG_1066.jpg
 
Circuit boards are here. Only need one for the build but two is the minimum order.

IMG_1068.jpg
 
Bank and range controller board is finished. Now I just need to wire up 12 switches and 10 LEDs on the panel into a few wiring harnesses.

IMG_1069.jpg
 
Starting to wrap up this new bank and range switch controller. The first picture is the panel wiring almost done. Wasn't as bad as I thought it would be. Next picture I'm getting ready to mount the circuit boards in the console. Third picture shows most of the wiring complete. You can see where I had to put some relief on the back panel with a belt sander to make it fit the console properly. Last picture is a top down view. All done except for that D connector which is used for in-circuit programming the MCU. The yellow tape on the sonalert is just to hold it while the glue dries. Just need to close it up.

Today I put a basic program into the MCU that gives me a heartbeat, and lets me check that all the switches and LEDs and the sonalert are working properly. I also programmed the wireless radio and made sure it would connect to the LCO master controller via the encrypted wireless mesh network.

I also just got my hands on a nice USB oscilloscope with built in logic analyzer. I've used digital scopes before but not with the logic analyzer. Very handy. It also does real time decoding of the inter circuit communications and the serial communications to the wireless radio. I've already seen a slight timing glitch on the inter circuit that prevents a frame from being read. Could account for some of the flakiness I've seen from time to time and trouble synchronizing drag races. Should be easy enough to fix.

IMG_1089.jpg

IMG_1073.jpg

IMG_1103.jpg

IMG_1098.jpg
 
Time for a single parting shot--for now. This is the completed Wireless Bank/Range Controller, designed to expand the reach of my 8 or 16 pad wireless LCO controllers to up to hundreds of pads. It checks out 100%, including the wireless. Just need to write some software for it.

That shouldn't be too tough. All it has to do really is send a message to the LCO controller telling it to communicate with a certain bank, and range. It will simply use the standard control frame already used by the LCO and field box controllers, which so happens to have a spare field. That control frame is AES encrypted and has multiple safety protocols embedded.

The bank/range controller will have a few protocols of its own but that's mainly logic. Can't select more than one bank at a time, for example. Except for "All In" mode. Probably need a better name. We'll have one mode where the goal is to launch everything. By range anyway.

I may post some updates now and then, if appropriate. I'd like to post all the schematics and PCB designs at some point. Nothing fancy there really, just detail. All the "juice" is in the software, and in the Zigbee multipoint wireless modems. They really saved my bacon. After a few months of fiddling with some small UHF transceiver modules I knew multi-node RF just wasn't going to fly for a larger system. Just too slow. Those little UHF radios (postage stamp size) will still work great for some future remote control deployment or tracking project though.

Anyway, got some software to write and debug, and some new rockets to build and fly. Until next time.

IMG_1111.jpg
 
I did order it. It's tool milled and engraved, not laser. I was so tired of making these panels by hand.

I know it's a PITA to be pestered with questions, but...

who do you source your panels from? I have exhausted my Google skills.


Thanks!


All the best, James
 
No PITA at all. Front Panel Express. Cost is reasonable but not cheap. Quality is outstanding. Never been disappointed.

You do have to invest some time in learning their design software.

https://www.frontpanelexpress.com
 
Last edited:
During my hiatus I've been thinking about how to protect the LiPo batteries inside the field box from over current and over discharge. Big expensive commercial type LiPos allegedly have protection circuits built in, but R/C LiPos don't. Over discharging your LiPo will cut its life significantly.

At first I just wanted to charge the battery properly, and than meant bringing the balance taps out to the panel somehow so the charger can connect to them. So I was looking at some mini-CPC connectors for that. But with mini-CPC I would have a short circuit danger with potentially exposed pins. Reverse polarity CPC would address that but only the regular CPC connectors have reverse polarity, and regular CPC connectors seem too big and $$ to me. And, more important, still no over discharge protection. So that's how I got onto the protection circuits.

They can be found all wired up with battery connectors and all that for $30 to $50 each, too much I think. I found some bare modules (no wiring) from some China sources at very low prices. They'll sometimes sell small quantities but you have to go though the whole quote process and then it takes awhile to get the parts. Then I found similar modules for pretty cheap all all-battery domestically which was great.

So that's what I have here. 7A and 12A battery protection modules for 3S or 4S LiPos. I think the 7A will serve my need and they are really small and narrow. Just 6 bucks. The 12A is the larger one and is for other testing. The idea is to wire these up and then enclose them in heat shrink just the like pre-made modules I mentioned. But much cheaper all said and done.

These are for battery protection and not balancing. Balancing modules are available too but they cost more. So the batteries will still have to come out of the box every now and then for a nice balance charge. That's a lot better than worring about if they are going to burn up from too much unbalanced over charging, or dying permanent death in the field from over discharge.

IMG_1564.jpg

IMG_1569.jpg
 
Last edited:
Back
Top