SparkFun Forums 

Where electronics enthusiasts find answers.

General project discussion / help
Did you make a robotic coffee pot which implements HTCPCP and decafs unauthorized users? Show it off here!
By tlancer
Hey everyone, this is my first post. I'm sorry to start out with a question, but I hope to become a regular contributor on this board as I gain more experience with electronics.

I'm about to graduate college with an Arts and Sciences degree, but I have always been interested in DIY projects. I'm experienced with web design and a few programming languages. It is only recently, however, that I've convinced myself that I am capable of building useful electronic devices (not so much now, maybe, but in the future). This brings me to my dilemma.

I want to build my own battery powered digital game camera for viewing wildlife. I know there are commercial options available, but I'm really interested in this DIY project. I feel like I will learn an enormous amount if I am able to complete this project. From reading around the net for around a week, I've realized this will be very difficult and time consuming to do the way I want to. Basically, I want to be able to pick a camera sensor, a PIR sensor, infrared LEDs, and the other components I would need, and design and make my own PCB board for all the components to go on. Presumably, I would then make the OS.

There are many things I don't understand about the process, though, generally speaking. I've been reading about I2C busses, uART, SPI, as well as FPGAs, SoCs, and development boards. However, I do not understand the entire process (broadly speaking), of how people pick what they need to be on a board, and then get to the point so that they have an embedded system--in hand--that they can code on.

For what it is worth, I am fully confident in my ability to create such a system if I used a Beaglebone for device interaction and processing. I feel like I understand the basic principles and how to proceed with my learning in order to attain such a goal...but I can't power a Beaglebone in the woods for 10 weeks on 12 AA batteries (a decent average for trail cam battery life). This is what has motivated me to figure out how to build my own, more power-efficient board.

I'm sorry if this doesn't make sense. And I realize that I have much to learn. I plan on doing this over a long period of time, but I would appreciate any reading suggestions, or options I may have of speeding up the development process. My main goal here is to figure out the path I need to take in order to be able to build my project--what I need to learn besides basic electronic theory.
By tlancer
Thanks for your response, Mark!

I thought about that (I looked at that particular product too), but there were a few problems I found with serial controlled cameras in general. They seem to be limited in their resolution. I should have mentioned that I'm hoping to add a 3 or more mega pixel sensor to the board.

Also, the Raspberry PI or Beaglebone both draw too much power on standby. If I turn them off and boot them with a PIR to save energy, then I have to add the boot time to the trigger time. Unless I am able to get either of their boot times to less than a quarter of a second, I have no chance of getting the pictures I want. After hearing about the PI's boot time being more than 10 seconds, I realized this was probably unrealistic.

In order for the camera to last long enough on batteries, I'm going to have to get the standby current down to around 0.5mA I think. If there is a way to do this with a PI or some other readily available platform (and if it's also able to come out of standby in less than half a second) then that would be ideal.
By Mee_n_Mac
tlancer wrote:In order for the camera to last long enough on batteries, I'm going to have to get the standby current down to around 0.5mA I think.
I think that's a bit conservative but in the ballpark. What series/parallel arrangement of AA cells are you intending to use ? Why AAs ? Is your "game" nocturnal, daytime or both ? Could you use a light sensor to further gate the PIR sensor ? Could you use a solar panel to charge the batts ?
By tlancer
First, I'm attempting to observe deer for the most part. They are active at night as well as the day. I'm going to have to either use flash LEDs or infrared LEDs to observe at night. So, I definitely will need the PIR sensor to be active 24 hours a day. I have wondered if it will be possible to power it sporadically, perhaps once every other second, if it could still work and would greatly improve battery life. Although, this would mean the real trigger time of my camera could be a second longer than it would otherwise be...which I've decided is undesirable if avoidable.

As for the power system...that really depends on the V requirement of my board and the flash system. I would imagine I would need somewhere in the vicinity of 4V for the unit. With AA's, that would mean I would need 3 AA's connected in series arrangement. If I used 12 total AA's. I believe that would yield ~3.6V and 4600mA hours. My other option is to use 6 D cell batteries which I could arrange to give me ~3.6V and 34,000mA hours. Correct me if I am blatantly wrong about either battery arrangement. Like I said, I'm just getting into all of this. I'm currently going through a great deal of information, and could easily have misunderstood battery arrangements when I read over them.

Since I'm at a very early stage, I'm basing my battery choice off commercial products. They typically run on either D cell or AA batteries. I do think I will include a way to install an optional solar panel (much like commercial products do), but I don't want the system to be reliant on it.
Last edited by tlancer on Thu Feb 27, 2014 11:00 pm, edited 1 time in total.
By tlancer
Oh and thanks for your response Mee_n_Mac! All help is greatly appreciated.
By Mee_n_Mac
What kind of pic to pic timing are you looking for ? IMO this drives the choice of processor to be used. Moving lots of data quickly from sensor to storage means MCU HP and corresponding battery power. Otherwise what's the MCU doing ? Polling the PIR, monitoring the battery (?charging?), keeping date & time, turning on/off LEDs, ??? Not exactly tasks requiring a quad core, hyper threaded, 4GHz I7 CPU.
By tlancer
I would like the capability to take about a picture per second if desired.
By tlancer
That's an interesting product.

Perhaps I've been confusing, but I'm really wanting a broad overview of what I will be getting myself into...much moreso than a list of parts that will help me build the camera system.

So I guess my question should have been something along the lines of: how can I learn to create finished products that are comparable to commercial products. So assume I have an operational trail camera system that utilizes a Raspberry Pi, Beaglebone, or what have you. What I am wanting to know is: How do you go from something like that, to fully customized PCBs that have _only_ the required components needed for that application (the minimum amount of processing needed, minimum amount of power needed, etc.)

I know one can design PCBs and order them. I also know you can give schematics to another source and they can actually assemble your product. What I don't know is how one is supposed to--broadly speaking--make the components interact. I assume there is something big I am missing and/or misunderstanding. Is all of this information supposed to be in a component's datasheet? Does the manufacturer only release this information to certain groups?

Remember, too, that I am not really looking for specific answers--just broad/point me in the right direction type of answers preferably with great reading recommendations.
By IdahoMan
Hey, mind if I bump this Thread?

I'm in the same boat. Trail/Game cameras are wonderful devices: They are compact, brick-sized, PIR-activated, hi-res video/cameras with IR-illumination (and sometimes audio) that can last for months on a set of batteries. "Camera Traps", they catch what moves. It's great to be able to set a "claymore" camera out and see what trips it, however the problem that arises is that it isn't an easy device to hide. It's pretty obvious and can/will be stolen/vandalized.

That's what got me into wanting to just build my own out of parts like mini-DVR boards, bullet-cams, sensors, batt.-packs, cases, etc.. Imagine designing a Pelican case with an SD-based recorder and a 12V Deep-Cycle or Lithium Pack rechargeable battery, outfit it with several outlets, and then roll out your own line of cable-ran cameras, sensors, illumination, mics, etc to your specification. (Ex: ... eet-sc.pdf)

The problem I am having is the same as tlancer's: Power consumption and trigger-time. For example, there is a mini-DVR out there by FoxTech. It can record on start-up, but I believe the boot-time 10-12seconds which wouldn't work as an event can be over and done in under 5secounds. When on and in standby it says it should start recording within a second, but its standby current is around 80mA and recording current is 90mA. Along with the other components a 70AH battery wouldn't last a week, let alone a year.


Where do we look for fast-trigger, low-power recording boards?
I don't supposed a TrailCam company would be willing to put out their electrical diagrams for study?