DIY electronics

Removed

A N-Ch mosfet will switch all the load you need (depending on the part you chose). These are readily available - see digikey/mouser/element14

I have thought about it, but the cost of losing a rocket makes me think abotu sticking to tried, tested, and purchased electronics.

I always fly home made electronics!

...
With a Raven as a back up... just to be safe

I also fly homemade electronics. The key is to have reliable backup deployment until you have thoroughly tested your electronics. If you are flying LPR, using motor ejection is generally sufficient as a back-up.

You will probably find that when doing LPR electronics your biggest difficulty is keeping the weight down on your av-bay. Don't even think about using anything other than a high-discharge R/C LiPo in the 130 - 160 mAh range. Even some 4-40 fasteners to get the current through a bulkhead adds a surprising amount of weight for LPR.

I am interested. I would like to see a good build threat.

cwbullet said:

I am interested. I would like to see a good build threat.

Click to expand...

+1

I love DIY projects, and having something custom built to your specs is just way cooler than strapping someone else's generic product into your rocket...

Although, I believe this guy was using pre-built circuits in a custom E-Bay and it looks just about as cool as I've seen for amateur rocketry... Especially love the key switch, huge display, and external USB connector!


I'd love to see a build thread and better explanation of the components in that E-Bay (I can't remember where I snagged the image from or I'd ask the guy)

For batteries, something like this will work. https://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=11857 I use them with my Raven.

These extensions can be cut up to make leads for connecting to your project and for charging. This is the easiest way to go.

Alternatively, you can roll your own with these connectors. This is what I did to make a charging harness until someone pointed out the extensions. You solder the terminals onto your wires and then press them into the housings. The battery will plug into these.
Mouser 538-51047-0200 Molex 2 position housing
Mouser 538-50125-8000 Molex terminals

You will need a charger and you need to make sure your matches or whatever you are trying to light will work on 3-4v if you go the single cell lipo route. My homemade Christmas bulb deployment charges work fine with the low voltage. Some matches/igniters may not.

Neuport said:

For batteries, something like this will work. https://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=11857 I use them with my Raven.

These extensions can be cut up to make leads for connecting to your project and for charging. This is the easiest way to go.

Alternatively, you can roll your own with these connectors. This is what I did to make a charging harness until someone pointed out the extensions. You solder the terminals onto your wires and then press them into the housings. The battery will plug into these.
Mouser 538-51047-0200 Molex 2 position housing
Mouser 538-50125-8000 Molex terminals

You will need a charger and you need to make sure your matches or whatever you are trying to light will work on 3-4v if you go the single cell lipo route. My homemade Christmas bulb deployment charges work fine with the low voltage. Some matches/igniters may not.

Click to expand...

They are pretty cool, but unless you make a lot, it is cheaper to buy precrimped connectors.

Chuck, I'm not sure I understood your response. I think we are saying the same thing. I recommended buying the precrimped extensions in my link. I just mentioned the molex terminals in case he had a reason to make his own. But maybe you are referring to something else? I am curious because I learn alot from how others do things.

I have been trying to standardize and modularize my AV connections to make it easy to move things between airframes. E.g. I use Deans 4pin connectors on my Raven and in all of my electronic deployment rockets so it is easy to just plug in the Raven and have all eight connections made (or four if its not DD). I have color coded the connectors as well with black being for apogee outputs and red for main.

Neuport said:

Chuck, I'm not sure I understood your response. I think we are saying the same thing. I recommended buying the precrimped extensions in my link. I just mentioned the molex terminals in case he had a reason to make his own. But maybe you are referring to something else? I am curious because I learn alot from how others do things.

I have been trying to standardize and modularize my AV connections to make it easy to move things between airframes. E.g. I use Deans 4pin connectors on my Raven and in all of my electronic deployment rockets so it is easy to just plug in the Raven and have all eight connections made (or four if its not DD). I have color coded the connectors as well with black being for apogee outputs and red for main.

Click to expand...

I looked at the molex connectors in the past. They requires a little tool if I am not mistaken to crimp them. To make up for buying the tools, you would to make a lot to make up for that cost. It a crimper is not needed then I am not well informed.

cwbullet said:

I looked at the molex connectors in the past. They requires a little tool if I am not mistaken to crimp them. To make up for buying the tools, you would to make a lot to make up for that cost. It a crimper is not needed then I am not well informed.

Click to expand...

I think we are on the same page. The precrimped harnesses are the way to go. However, if you just want to or have special requirements you can use the molex connectors. I do not have a crimper but used solder and folded the crimp tabs with a small pair of pliers.

Neuport said:

I think we are on the same page. The precrimped harnesses are the way to go. However, if you just want to or have special requirements you can use the molex connectors. I do not have a crimper but used solder and folded the crimp tabs with a small pair of pliers.

Click to expand...

I am very interested to see how other do it. I like to experiment with different connectors. Deans connectors work well.

Get a kit with the solder-less breadboard. Some years ago I picked up the Learning Lab from Radio Shack and did some experimentation with the 555 timer and prototyped a solid state turn signal for my street legal dirt bike. The solderless was great for experimenting before I made the final device and potted it. This was some awhile ago and I haven't done much with electronics since but I still have all my paraphernalia just in case I have the whim to do some more experiemting. Now I'm wanting and toying with the idea of experimenting with an Arduino or equivalent device.

Here's an interesting 555 based deployment timer project I stumbled across a while back:

https://www.thefintels.com/aer/protected/timer.htm

Cheers!

Go for it.

Unless Im under a tight schedule most of my stuff is DIY. DIY is much more interesting than COTS.

Watch the 555. The timing diagram might be a little different than you expect. (or maybe not) Can you add a 1/4 sec ignition debounce with a second 555 or an OpAmp??

Im a big fan of the $29 Coridium (and Arduino) for DIY. The Coridium is much nicer than a Stamp b/c it comes with free C compiler, tons of easy to access I/O pins, and working examples.

-->MCS


.

+1 for Arduino when you are ready to do some small digital projects.

I got the experimenters kit and use it for prototyping but then I often end up using a small homemade PCB, or even perfboard, and the Atmel chips to save space and cost. E.g. I have prototyped a few projects on my Arduino which has a ATMega 328 but then ported the code to an ATTINY85 if I don't need a lot of pins to save space. Also, while analog timers are certainly doable, if you have any experience with digital and programming, I would go ahead and build the timer with a small microcontroller like the ATTINY85 or possibly even one of the newer Atmel chips with built in USB support. The microcontroller allows you a lot of flexibility to change the operation of your timer, debounce switches, add indicator LEDS, failsafes, etc. all in code.

Like most of my hobby pursuits though, once I got bitten by the microcontroller bug I ended up buying lots of support equipment so the overall investment is large relative to any final useable projects I've created. It all started with a request from my brother to build him a wireless timer interface for a pneumatic actuator he had built. I started with off the shelf components cobbled together and eventually got into the microcontrollers to make it work reliably. He thought it was great that it only cost $100 dollars or so in various parts including the initial experiments. He didn't realize I had built up many times that amount in infrastructure and "jelly bean" parts.

Anyone getting into "build it yourself" should get to know SparkFun. You can create your own electronics by wiring together very advanced chips (pressure, acceleration, GPS, gyroscopes) that are already mounted on breakout boards (no chip soldering required). Also, they have batteries and chargers with connectors.
Fun!
www.sparkfun.com

Sparkfun and Adafruit are two of my favorite sites.

I breadboarded the Missileworks firing circuit from Figure 6 of the Scratch Builder's Supplement document. It's not clear to me how to detect continuity, though. My electronics skills just don't go that far. The output of the circuit to pin 6 of the processor tells the processor something that lets it know whether or not there's continuity, but I'm not sure what that something is. Okay,well, I assume it's a low current through the circuit, enough that the processor reads a "high" on that pin but not high enough to fire the ematch, right? Should I just be looking for pin 6 as a "high" input?

Anyway, the firing circuit itself works beautifully, even if it's an oldie-but-goodie.

I've messed with Arduinos as well as with quite a few Sparkfun products, and Parallax Stamp and Propeller chips. If I were going Parallax, I'd go straight to the Prop chip: it's FAR more capable than the Stamp and costs only a fraction of what a Stamp costs. You can breadboard a working Prop-based computer pretty easily from full documentation on their site. The learning curve is higher than for the Stamp, but once you're over the initial hump you can do complex things very easily. The Arduino has advantages, including easy analog input and a huge variety of form factors, plus it's programmed in C, but the fact that you can run multiple processes at the same time on the Prop makes some things easier (like when you want to monitor something you're measuring while doing other things). I used a Prop chip on my telemetry ground station to continually read altitude data while also looking for changes in the states of two pull-pins that indicate drogue and main deployment.

gwolski said:

Here's an interesting 555 based deployment timer project I stumbled across a while back:

https://www.thefintels.com/aer/protected/timer.htm

Cheers!

Click to expand...

Personally Im not crazy about that design.

It relies on a magnet to hold a reed switch open. When the magnet disappears - like at launch - the timer starts counting. If the magnet were to disengage accidently you have a problem.

It is not easy to replace the magnet-reedSwitch because the circuit requires a normally closed switch. Ie the switch is open on the pad so a break wire won't work or G switch won't work.

A long time ago before uProcs like the Coridium were cheap I designed a super simple timer using a handful of discrete components including a 1/4 second debounce. Maybe a starting point for you........

https://www.spiegl.org/rocket/timer/circuit.html


-->MCS


.

I advise against the use of a 555 timer as well. It will get the job done, but it becomes very difficult to make it failsafe.

I *strongly* suggest going the microcontroller route. These days microcontrollers are only slightly more expensive than 555 timers, and there are many cheap/free programming options. I am a *huge* fan of TI's MSP430 LaunchPad system. You get a complete development/debug system with two target chips for $4.30, and that includes free shipping.

https://estore.ti.com/MSP-EXP430G2-MSP430-LaunchPad-Value-Line-Development-kit-P2031.aspx

Awesome job. Can't wait for a launch report.

Are the copper pads on your board facing up in that photo?

Edit: I was using my phone to view it earlier. Now I see they are. Is it a double sided protoboard or is there a reason you put the copper on the component side? The reason I ask is you would normally want the copper on the bottom of the board with through hole components like these so you can make your solder connections on the bottom. You probably know that and I'm probably missing something. Just curious.

Here's one I'm working on. It's based on the 18f67j50 PIC. The first image is of a protoboard containing the microcontroller. I'm using it to breadboard the second design with is a complete altimeter system. The latter has an accelerometer, barometer, micro sdcard, flash, 4 channel pyro, usb. The firmware is somewhat functional and includes a kalman filter for noise reduction. The layout is done in eagle, firmware in mplabx. Board fab is at batchpcb. I might have it done and ready to go for a few rides as a passenger by next summer. My goal is to produce these things cheaply enough that I can have a couple on each of my rockets without breaking the bank.

Degreaser, if I may make some suggestions:
* Make sure you tent you vias. Bare electronics with untented vias scare the crap out of of me, especially in safety critical applications.
* Get rid of those 90 degree corners in your tracks.
* Your tracks around the uC are very small - BatchPCB are an extremely low end supplier and some of these may be etched away, shorted to surrounding copper or drills may just miss them completely. Consider PCBCart and making panels.
* Make your vias bigger wherever you can, or implement teardrops. BatchPCB miss small vias when drilling. LOL.
* Looks like you have pulldowns on your fire fets - where are your caps for debouncing and suppressing noise?
* How many amps are you rating your channels to? SOT-23 can't get rid of heat very well at all and really shouldn't be used for more than a second at over 1amp. Consider DFN w/ PowerPad if you want to fire high amp loads.
* Calculate how many amps your power tracks can take. If it's less than you can dump with the fet then consider a small power polygon.
* Increase the size of the tracks in your thermals, they are nowhere near as big as your power traces and may blow.
* Check your clearance between top left drill and USB. I don't know you'll get a screw head in there.
* If you make your programming header SMT and use pogopins you can save a bunch of PCB space.
* Add *lots* more ground staples around all your data and control lines.

Issus - this is a dev board. I view untented vias as a feature for blue wires should I need them.

I've ran a dozen designs through batch pcb without issue so fare.

Larger vias on the pyro circuits are probably a good idea.

No caps are needed on the fet gates. the 18f series of pics have the pins initialized at reset to a high impedance state. They have pulldowns to turn the fets off at powerup.

There are around a dozen decoupling caps scattered at appropriate places on the board, both top and bottom.

BatchPCB uses GoldPhoenixPCB in China. Not mil-spec boards obviously but they are 'decent' commercial quality.

Compared to the boards i'm used to out of china, BatchPCB are pretty bad - they also don't get the full QC check that GoldPhoenix would usually do (flying probe tests and such) if you sent them a panel directly. I've seen boards from my clients made at batchPCB where etchant has gotten under the mask and given hazy edges, or not etched through gaps between traces completely. Lack of ENIG is also pretty bad for prototypes if you come back to make a couple more a month or two later. I look at PCB's all day long, either physically or in Altium. BatchPCB have got to be some of the worst PCBs i've ever seen manufactured.

The decoupling caps are for killing off any static, or RF (or any other EMI) you might get in your system. It gives you a slightly higher rise time on your outputs but can be a good safety feature.

Yes, RF immunity is something to be considered. My v1.1 telemetrum was rebooting/crashing due to RF injection from its own transmitter. A .01uF cap on the power leads took care of the problem. The problem has been fixed in v1.2.