There is still a need to increase the angle so I see the upper parts of the trees during fast proximity flight.
I am doing new tests with a higher angle next time.more ...
In the meantime I created a cased version of the illuminator as well.
I have reduced the number of LED modules to 2, and raise the current for maximum light output.
The LED modules are now driven by 1A, and instead of the old-school linear current regulator I used quality high efficiency switching regulators. The LEDs are the 850nm ones, again from Osram.
The total output light power is now ~8200mW, the IR module is just slightly warm after 1 minute of run, and as the full case is aluminium, it distributes the heat very well.
Because of the casing, the module can be hardly abused, without the fear of destroying the LEDs.
Now it can be mounted to ANY 250-sized quadcopter via a single velcro strap.
The input can be 4S only, a 3s LiPo does not provide enough voltage for the module.
For vTX I used the Aomway 200mW 5.8G device with CP antenna.more ...
After a long journey, I came to the point so I can finally state, that the project is finished and successful.
For this, I had to triple the infrared LED arrays, but finally I have enough light power to see everything that needs to be noticed while doing fast proximity flying. The diy illuminator module now contains two 940nm arrays on sides, and one 850nm array at the center. Currently there are three independent LED drivers, all set to 830mA. The estimated light power is around 8700mW, while the intake power is around 40W. Sure the module is heating as hell while on ground, but when airborne, the temps are just within the specs.
The AUW is now 660g, flight time with 1300mAh/4s LiPo is 6 minutes, and 4.5 min. if speeding.
I was so pleased by the today's test that I forgot to try out the recently arrived low light lens for the fpv camera :-)
Nevermind, here is the video about the last test:
Now I am going to find some nice places that I can flight by night. There is maybe just one more thing that can be adjusted, changing the dissipating circuits for some switching ones, to achieve better efficiency and lower weight and temperatures. And of course to test the LLL (low light lens) on the Starlight fpv camera. Maybe by the next flight...more ...
Yesterday I did the first airborne test with the Surveilzone's Starlight camera (CC1564). The result is still not satisfying, although much better than with the former HS1177 (PZ042) fpv camera. Much more details could be seen but still, for fast proximity flying the illumination level is too low.
The good news is, that the fpv camera is really able to handle the 4s LiPo, so I had not have to use additional DC converter on the quadcopter. The AUW of the bird is now 635g.
I will try to reposition the illuminator as when flying just a bit higher than ground level, there is not enough light to see the ground well. I should adjust the angle of the infrared modules, and I need to move them closer to the nose of the quad, so the Mobius does not partially shade the IR light. In addition to that, I will set the fpv camera back to B/W mode, as I hope that it will increase the sensitivity in overall.
The next test will be done today by night. Until then, I share the yesterdays video. Note that I was too tired to realize that I picked a wrong position and during the whole duration of flight, the 5.8G CP patch antenna was turned back to the copter, so I flown behind it. That is why the picture is flickering :-( Culpa mea.more ...
The Starlight camera finally arrived so I could do the ground tests.
The camera is said to be as sensitive as 0.00008 LUX and capable of operating from 4S LiPo battery. For now, I run it from a 3s battery.
The only setting I made in the OSD was to set the camera to B/W mode. Beside that, I changed the stock 3.6mm IR CUT lens for the 3.6mm IR sensitive one.
The result is quite satisfying. The camera performs very well either in very dark places or in well-lit areas. Comparing to my HS1177 (PZ042, CC1333), it is much more sensitive to ANY light it can process. However, in very dark places the picture gets quite grainy.
I have not noticed any lag different from the HS1177, even with DNR settings in AUTO, that should be around 30ms. Many says the cam is not suitable for proximity flying because of lag and blurriness but I think the opposite. Soon I will test it in the near park to prove the statement.
In the video above I compared the picture of the CC1564 to the IR sensitive Mobius, first in the darkness (just a weak street lamp), and later on when the area was illuminated by a 940nm infrared module.
Later on, I try the camera even in color mode and was surprised that the picture seemed to be a bit better during no extra light and much better when adding extra illumination. So I decided to left it in color mode finally.more ...
The additional IR module arrived this week and today I finally found time for rebuilding the DIY infrared illuminator. The original driver was simplified, now using just one resistor in the circuit, that drives the LEDs with 830mA only. I have added one more driver, for the second module. Now the module weight is 49g, and it is still mounted on top of the battery.
I am still waiting for the starlight fpv camera, it should be delivered by middle of the next week, so I hope that my night vision stealth solution will be ready soon.more ...
Yesterday, the fpv maiden was done. It was only a half success, I nearly killed the copter a few times, when I lost orientation completely, because the FPV camera could not see the trees in the dark. The onboard video, taken by the DomHD DVR unit is much brighter than the real picture, so the quad seems to be controllable but believe me it was not. I was nearly blind, during the whole flight duration.
The best flight can be seen here:
2 major things need to be adjusted:
- FPV camera (sensitivity still low with 0.01LUX, a starlight camera is desperately needed)
- +1 IR illuminator (angled at different height, so either the hover or the FFF is covered)
The project is frozen until the new starlight camera and the additional IR LED module arrives.more ...
By the evening, I powered the quad in the garden and finally get the DIY IR illuminator module airborne. A LOS flight by the evening in rate mode is not the best, sometimes I thougtht I lost the oriantation at all :-)
The quad AUW is now 600g, the flight was only 02:20 long. Charged back 400mAh into the battery.
After the landing, I checked the module temperature and it was really warm but not too hot to touch, so my assumption about airflow cooling was right. The concept seems to be working. Now I am waiting for the IR sensitive FPV camera, to be able to test the quad in the field.more ...
While I am still waiting for some fpv components to arrive, I have finished the driver for the IR illuminator and packed it together with the IR leds. For now, I decided to mount the whole pack in top of the battery, until it will be thoroughly tested. Of course the IR leds are not protected well enough this way, so I will try not to crash the copter during the tests.
The illuminator mount is ready, I have made it from an old aluminium plate. Now I am trying to find out where to fix it on the M250NV quadcopter frame. Initially, the idea was to fix it to the upper deck, in between the rubber dampeners but I will see how the fpv camera will fit when it arrives.
Yeah I forgot to mention, I decided to use the fpv camera finally. The reason is simple, at some speeds the Mobius sensitivity may not be enough to detect and avoid obstacles in time.more ...
The frame for the project was finished today. It is the same model I use for my racing quadcopter, planned from scratch and redesigned a few times. I shall call this specific model as M250NV.
It is made from black G10 glassfiber material, cut with my hobbyCNC machine.
The weight without the screws (alu spacers included) is 120g.
The frame is announced to be unbreakable. I am using 2 of these for some time, and after many crashes both are still intact.
After resizing the resistor values finally I powered the IR led array illuminator with 1A. The maximum radiant intensity at peak power should be 3740mW, according to the manufacturer specification. I assume this as the total optical power, that is radiated at 90 degrees total.
Maybe, If I could focus the light somehow the intensity in the center would be higher but still it seems to be enough for night flying.
I am not sure whether I reached the maximum power but now the garden is illuminated completely, I can see even the furthest point. I am a happy camper here :-)
The Osram IR LED array arrived today. After getting home I quickly created a test platform and went out to try it.
The result is outstanding! The array of size of 25mm x 25mm, driven only by 650mA illuminated the whole garden at a level that Mobius was able to record it. I will drive the array by 1A current later on, to get the full optical power, but anyhow, the current power level is enough for fast proximity flying.
Note that the FPV camera is much more sensitive to IR than the Mobius so night flying will not be a problem anymore. Now my only problem is to encapsulate the whole solution into a case so it can be protected from unwanted landings :-)
Check it out:
The recording was taken by the hacked Mobius camera.more ...
In the meantime, while I am waiting for the IR LED array to arrive I decided to give a try to an IR laser illuminator. I have some IR diodes harvested from old CDROMs so I wanted to try whether the laser illumination can work at all.
First question is whether the IR diode has strong enough output for area illumination. Upon information back to year 2004 I found, that the power output of these varies in between 40-60mW. I doubt that would be sufficient for my application but I wanted to test it.
The second question is whether the unfocused beam is wide enough to cover a certain area that is needed when flying with a race quadcopter. Another question is whether I can find any focusing lens that can be used to focus/unfocus the beam.
Third issue can be the beam pattern/shape, best would be to find an IR diode that has an oval or circular pattern.more ...
Recently, I found a good article about the LM317 so I decided to share it also here, as I am using this IC for current regulation with my LEDs and laser diodes.
A current source is the dual of a voltage source – it delivers a constant current, regardless of the voltage. While a resistor can be used as a rough current source, the current drawn will vary with the voltage. (Ohm’s law stipulates that I = V/R. The resistance is constant so if the voltage increases, current flow also increases. Same applies for decreases).This project is based on the common LM317T regulator, used as a constant current regulator. The circuit is ideal for driving LED arrays, laser diodes or any other circuit requiring a constant current supply. Alternatively, this could also be used as an electronic dummy load for testing small power supplies at relatively low current levels. Suggested applications and all calculations are explained in the article below.
The full article can be read here: CONSTANT CURRENT SOURCE/LOAD (LM317)
The laser IR illuminators are on a parking lane for some time, as I could not find any useful information about defocusing the beam. Unfortunately, most of IR lasers are used for burning, and using focusing lens. What I need is the opposite.
So I turned back to LEDs again and after some more searching I found a solution that can be suitable (or at least I hope it can).
Still, I give a try to the 940nm leds, to achieve the full stealthiness, although the leds at this wavelength are not as powerful as the 850 or 808nm IR ones.
The victim is the led array provided by ILS, using Osram's most powerful "black sun" IR leds. I plan to build a micro 940nm IR illuminator with this module.
- Oslon 4 PowerStar IR LED arrays contain 4 Osram IR Oslon Black Series IR LEDs
- Current range 100 to 1,000mA
- Lifetime: Up to 100,000 Hours to 70% of original brightness
- Mounting holes for M3 screws allow easy installation
- Size (L x W x H): 20mm x 20mm x 3.85mm
- Angle of half intensity: +-45 degrees
- Forward voltage: 11.2V
Not long ago I have found a video about an fpv racing by pitch black and I really loved it.
I am thinking of a similar solution some time and the time for the project has finally come.
I intend to use the same gear as for daylight flying, except the fpv camera. For that, I want to use the Surveilzone low-lux Starlight cam. For the HD video recording, the good old Mobius will be used, slightly modified.
- stealth quadcopter (no visible light emission at all)
- 4-5 minutes flight time (racing)
- IR sensitive HD recording (hacked Mobius)
- IR sensitive CCD fpv camera (low lux rating)
- additional IR illumination (IR laser)
- illuminated area at least 60 degrees and 15m distance
- ability to fly in pitch black conditions with moderate speed
- ability to mount the IR gear to any 250-sized quadcopter