Since the last time I wrote about the cheap cameras from China, locally known as CCC (Chinese Crap Camera), I’ve been tinkering to get something sensible out of the small buggers.
The footage we have obtained is OK as is, the problem lies in getting them to function properly when the action is happening.
At the engine test during Summer 2016 I had three CCCs running, and only one of them delivered footage. The other two stopped recording every time the engine was ignited, no matter which incantation (or malediction) I threw at them.
The acoustic environment was probably a bit too harsh, as the cameras was placed inside our Gaffophone-style test container alongside the BPM-5 engine.
At the launch of Nexø I, things went awry once again. I had fabricated a Skyview Rig; a fancy name for a box containing two powerpacks and two CCC looking up in the sky, one through a wide angle lens, the other looking through a zoom lens. All aimed at getting some awesome footage of the rocket roaring towards the firmament.
Everything went fine – up until one hour before launch. That’s when first one and then the other camera had a heat stroke and shut down. The last footage shows the baskets on the launch rail tower turning to allow the rocket to pass by.
A later test in a similar environment, temperaturewise, revealed a smell of overheated plastic when the box was opened after 4 hours of operation. One camera was still recording, the other had shut down. Doh!
It appears that something has to be done to ensure higher resilience to the thermal and acoustic challenges that will occur. A metal enclosure with a heat sink may be all it takes?
I started out removing the plastic enclosure to have a look at the posibilities.
The camera is build around a cast metal frame, carrying two circuit boards (one mainboard and a smaller interfaceboard) and the CCD/lens assembly.
It is possible to mount the front of the frame onto a heat sink, if the heat sink has a suitable aperture for the camera lens.
To make the camera more immune to the powerful vibrations from the environment, the SD card can be soldered to the card socket. So I simply removed the socket from the circuit bosrd, soldered the SD card inside the socket, and remounted the socket on the interfaceboard.
Now we need a metal enclosure around the camera, and a way to operate and connect to the camera has to be provided.
This will entail some wires connecting the camera to some push buttons, a couple of LEDs and some connectors. We might as well change from the standard battery to a couple of 18650 cells, which will provide power for several hours of operation.
After procuring a metal enclosure with the right size and proportions, the required push buttons and connectors was next on the agenda. I found it preferable to provide access to as many of the available signals as possible, which resulted in a demand for: a HDMI connector, a USB connector for data transfer, a BNC connector for CVBS video, and a DC input connector.
I also found it preferable to use switches and connectors capable of withstanding the outdoor environment we operate in, hence all are at least mud- and gravelproof. The ebay sellers all claimed waterproof in their specifications, something I really do not trust to be true.
The DC power input needs to be rather universal. This requires a small switchmode regulator, allowing the camera to be charged from any AC or DC voltage between 8V and 28V. Using the USB connector, the camera can also be charged from 5VDC. Although charging should and can be performed in due time before use, this may well be under field conditions, without a mains power outlet handy, hence the wide input voltage range.
A viewfinder would be rather nice too, enabling setting the camera up without using a WiFi terminal (smartphone). Despite the viewfinder feature, an external WiFi antenna will be added regardless. The original antenna is glued to the inside of the plastic casing, making it too difficult to reuse it in the new metal enclosure.
A connector for the viewfinder was thus placed on the top of the enclosure, meant to carry a small hooded TFT monitor. A power source for the monitor is needed, and using the camera battery as the power source will be preferred. For this a small step-up switching regulator is needed to boost the battery voltage to the required 6V.
To avoid a rats nest of wires inside the camera enclosure I decided to design a small circuit board to keep the wiring in check.
The board contains three connectors for the camera, and a larger amount of connectors to connect the switches, external connectors, and indicators. And the two required switchers are also placed on the board.
Some preliminary testing has shown that the camera, rather than getting too hot to touch comfortably, now heats the heat sink to a lukewarm temperature.
After the first camera was converted to the armored version, the conversion of the next camera was straightforward. This time I omitted the specially milled heat sink on the front of the enclosure. The milling was rather time consuming, and I would rather be without this process.
And camera two works just fine without the heat sink, for several hours on end.
I am now in the process of converting a third camera. This camera will be “turned on end” as can be seen on the photo with the still vacant enclosure. A bit of rearranging inside the enclosure has been needed, but that is minor. This configuration should be favorable for mounting the camera on a wall.
I will leave it up to the pictures to tell the rest of the story. Albeit I have to admit that somewhere, through the lengthy process of determining how best to implement the conversion, I have apparently succeeded in killing the radio in the first camera. Because the WiFi link doesn’t work anymore.
This means that I will have to use the camera in its current configuration, no changes can be made without WiFi. Unfortunately this means that the video output is disabled. Not a reason to scrap the camera in my opinion, it’s still in useable condition.
For this reason I have omitted the planned TFT monitor connector and external antenna on the first camera, an replaced the video BNC connector on the back with a DC connector meant to power an external HDMI monitor which can be used as a viewfinder.
I have also omitted the TFT monitor connector on the third camera, based on the expectation that a viewfinder on top of the camera would be impractical to use. Instead WiFi or HDMI can be utilized.
The HDMI monitor is a converted car review monitor (4.3″, 2 CVBS inputs with autoselect) with an added HDMI interface. There’s room for the interface board inside the original enclosure after removing some RCA connectors and remodelling with a hacksaw and a knife.
At the monitor has an autoselect feature, I have included a CVBS cable with a RCA connector. A power pack with an added 12V boost regulator acts as independent powersource for the HDMI monitor, there wasn’t room for a battery in the original enclosure.
The box containing the monitor is really a box intended for protecting a 3.5 inch hard disk, but will hold the monitor and an HDMI cable nicely.
The story about the Skyview Rig will have to wait for another time.
Make a donation
Help us continue building rockets with a one time donation
5 Comments
William Jackson · 20th November 2016 at 3:52 am
You need someone with experience in high intensity high frequency vibration. You need eyeletted boards, with all wires tightly laced and the laced bundles nylon tied to the chassis.
As it sits, wires will fail from fatigue caused by strong vibrations from 10 to 200 Hz with large excursions = fractions of an inch with g forces in the hundreds for modern high performance launch.
Potting can be used, if the extra weight can be tolerated.
Bo Braendstrup · 21st November 2016 at 10:00 pm
We would love to have someone with experience in high intensity high frequency vibration join our project.
Other than that, you should consider that what is described is a way to get dirt cheap cameras to function better, on a shoestring budget, in an environment where a small (5kN or less) engine is tested or launched. The camera is mounted on the ground structure, not the rocket. The footage is “nice to have”, not “need to have”, and should the camera fail, it will not render a test or a launch useless, all the needed data are obtained through other means. If the cameras were mission critical, I wouldn’t base my work on a 60USD chinese camera.
/Bo.
Jeff · 20th November 2016 at 4:26 am
Great work! But I’d definitely agree with William, you should have your wires held in place better or you may have failures down the line.
NASA publishes workmanship and materials standards for their wire harnesses and circuit boards, I think it would be worth a look. After all, they’ve been doing exactly what you are for decades, so they’ve figured out what works.
Again though, that’s a great fix and I love seeing what you guys are able to accomplish.
Ben · 20th November 2016 at 8:16 pm
I just read this article last week referring to the exact topic in the comments, and William / Jeff are right, the NASA documents are extremely well weitten and cover a fantastic range of topics.
http://hackaday.com/2016/11/03/specifications-you-should-read-the-nasa-workmanship-standards/
Jörg, Mannheim, South West Germany, Central Europe · 27th December 2016 at 2:50 pm
This is an unbelievable great job. May I say in modern newspeak “congrats bigly!”
Just two cents referring to the given CCC description and some more general ideas:
“The other two stopped recording every time the engine was ignited, no matter which incantation (or malediction) I threw at them.”
I know you must have great experts for the downlink and all IT. Therefore this maybe all not new, but just in case it is as you are really leaving the usual comfort zone of things!
Some decades ago I was thought (by people messing around at CERN), that “failing” is an option for all Integrated Circuits as a random function depending only on a very low probability of ambient radiation. In theory all IC could break or show some kind of miss function from a single flash from a single spark plug from a small combustion vehicle driving along in some distance – let alone a rocket engine nearby! That is why all IC’s going into outer space are somehow special hardened. Much more details see here
https://en.wikipedia.org/wiki/Radiation_hardening
If you crosscheck the “SEE” description in the link, now to start with a solution instead of a problem description I would be sure to spend for all IC based equipment that stays next to a burning rocket engine – where I’d assume all kind of electrical radiation is produced during engine operation, and strictly according to the general approach 😉 some radiation radiation shelter derived from e.g. a broken-up microwave oven (!) wrapped around all ground based IT. I assume the aluminum cage is in this aspect not dense enough. You can do some tests with mobile, WIFI and WLAN devices.
This radiation cage would just be a general precaution to reduce the failing probability! Maybe additional some other wires and dense cupper fence for other spectrum.
In addition I’d ask some radio amateur or mobile network company to conduct search for such electromagnetic concurrent signals on the test starts. You could even involve state agency as are obliged to search for emissions, have the devices and need training. Could be that the engine produces minor Alpha or Beta radiation. Besides this, the electromagnetic waste footprint could as well become an external control signal for the well functioning of the rocket engine!? In short “You see!” in long as I assume the electromagnetic emission of the burning engine is already monitored with this build in, free of cost general purpose electromagnetic sensor called “eye”. It would be worth thinking, if you try to ask some astronomers to follow the emission spectrum in the ultra red or ultra violet of the starting engine. Maybe easier the other way around: If somebody does shooting star monitoring (e.g. like http://www.amsmeteors.org/ ) and is happy to monitor a rocket start!
This would give additional and maybe cheap insight on the efficiency of the rocket engine by monitoring the buring temperature derived from the light emission spectrum?
PS: Just a hint to avoid much more likely causalities…as improvement number zero I’d prepare a fish rod with a lasso to bring in the watered rocket just to avoid losing a brave, functional diver.
Comments are closed.