Mr.RC-Cam

Admin
  • Content count

    5,219
  • Joined

  • Last visited

Community Reputation

114 Excellent

2 Followers

About Mr.RC-Cam

  • Rank
    RC-Cam Mentor

Contact Methods

  • Website URL
    http://www.rc-cam.com
  • ICQ
    0

Profile Information

  • Location
    USA
  • Interests
    R/C, FPV, Embedded Programming, Electronic Design.

Recent Profile Visitors

3,127 profile views
  1. There's some information scattered on the web on tuning. Here's some basic info that I found with a quick google search: http://fpvlab.com/forums/showthread.php?35784-Naze32-altitude-hold-simulate-Naza-Attitude&p=616964&viewfull=1#post616964 Do not expect miracles. Cleanflight's Alt-hold has lower performance than DJI NAZA. That is because Cleanflight was developed for racing; Its GPS and Alt-hold functions are not fully developed. So if you are expecting it to work as good as a DJI Phantom then you will be sorely disappointed. It is important that you prevent sunlight *and* prop wash from affecting the Baro Sensor. It is very sensitive to both of those things. Mine is inside a opaque housing (blocks light) and is covered with soft breathable foam (reduces effect of prop wash). Also, all your props must be in good condition (clean, no damage, straight). They may require balancing; Any prop vibrations will affect the Alt-Hold's performance.
  2. You can tweak the ALT-Hold PIDs if you need better performance. I don't move the throttle after enabling Alt-Hold. However, it can be moved to increase/decrease height; But when Alt-Hold is disabled be prepared for a dramatic change in throttle speed (due to the throttle stick's altered position). I don't know. It works in the version of CleanFlight I installed a few months ago.
  3. All four Nav lights are wired in series and operated from one LED driver board. The Strobe/Landing light is two series connected LEDs that are operated from another LED driver. Both boards are powered from the model's 4S LiPO battery. You can use either connection method. But PPM-Sum is the modern way to do it. I suggest a 8+ channel R/C system. For example, you need 4 channels for throttle/pitch/roll/yaw, 1 channel for flight modes, 1 channel for retracts, and usually a couple more channels for OSD or Gimbal operation. Although the hardware will support R/C control of the lights, and was considered, I didn't need it so it is not included in the existing firmware. The good news is that the interrupt driven PPM decoding for a second control channel is already being done for you. So remote on/off of the LEDs is possible, but you'll need to add this functionality if you need it. Correct. Correct. The Servo Stop value is ignored when using the Motor Driver board. The 3-pin connectors on R/C receivers have Ground, 5V, Signal. That is why these signals are shown on the schematic at the J2 & J4 servo plug connections. 5V power for the R/C Receiver and Arduino is provided by the BEC feature from one of the ESCs (the other 3 ESCs' BEC outputs are unused). ESC (x4), Retract Motor Driver, and LED Drivers (x2) are connected to the PDB, which is powered by the 4S flight battery. All grounds are "STAR" connected to the PDB, which serves as a power hub. Power to the 5.8GHz FPV gear is 12V regulated and includes a LC filter to ensure clean video. The ESC's I selected had the dimensions and electrical specs that I needed. But they did not have BLHeli firmware in them, so I had to flash it myself. Some ESC's have BLHeli loaded in them by the factory; BLHeli is very popular for Helicopters & "Drones" due to its optimized performance (versus a common airplane ESC).
  4. Worked well for me. But if you run into issues then keep in mind that you can change the code to meet your needs. BTW, after using the auto retract feature for awhile I found I preferred manual control (without auto retract). But everyone is different, so the feature is there for those that want it. The config.h file allows you to choose how the retracts are handled by the assigned radio switch. It does both directions. If this is not desirable then feel free to tweak the code. The PPM Servo Out port is for model airplanes that have official Retract Servos (a special kind of R/C servo). It could also be used with a Continuous rotation servo (a standard servo with the feedback pot removed). These two servo choices were not used in my Inspire installation. As documented, I customized a R/C servo by hacking it with a H-Bridge Amp to drive the motor. A layman's build guide would be great. Thanks for the kind feedback.
  5. This project was finished back in 2015: No new activity since then. It is challenging and rewarding to build. However, it is a fragile model and not suitable as a everyday flyer. Cost is high, requires over a hundred hours of printer time, and many more hours of assembly. It is suitable for someone that has already mastered configuring/flying traditional R/C Quads. It is also helpful to have a prior 3D printer experience. If my Arduino based retract controller is used then C coding and electronic hardware experience would be needed too. Otherwise, skip the Arduino and use the simple diode based limit switch solution instead. The Arduino controller firmware is very sophisticated and should be suitable for planes too. Extra effort was made so that it would be a universal R/C retract & LED controller.
  6. Next: Come back soon for tips on using your new Vid Cal Tool.
  7. It is important that your two MinimOSD based target generators have identical video characteristics. Sloppy component tolerances and manufacturing mistakes will ruin the Video Cal Tool's effectiveness. Fortunately it's easy to test the target generators to confirm they are OK. Do not use the Video Cal Tool until you have successfully passed this test! Perform this special evaluation test while indoors; Do not test outdoors because the strong ambient light can wash out the display too much. Use a high quality monitor with large screen (mine is a 23" LCD TV/Monitor that lives on my workbench). The MinimOSDs' evaluation test is quick & easy. Follow these three steps: STEP 1. Connect the vTx generator's VID-OUT to the VID-IN of the vRx generator. Connect the VID-OUT of the vRx generator directly to the monitor. Like this: STEP 2. Turn on the target generators. Within 5 secs you should see the image below: STEP 3: While reviewing the R & T target symbols, vary the monitor's contrast and brightness controls. The test is a PASS if the targets remain a perfect match at all settings. Repair or replace the MinimOSD boards if the R & T targets do not have matching video levels (FAIL).
  8. The existing Arduino firmware on both MinimOSD's needs to be reflashed with a new set of sketch files. The internet is full of How-To's that explain the flashing process, so I won't repeat the details here. If you are new to Arduino then be prepared to watch some YouTube videos and/or visit some web sites. Note: This Arduino project will compile without any errors on Arduino IDE Version 1.8.1. To avoid compile failure frustrations I suggest you use this version too. In the Arduino Tools menu choose these two settings before flashing the MinimOSD boards: Board: Arduino Pro or Pro Mini Processor: ATmega328 (5V, 16MHz) 1. Assuming you already have the Arduino IDE installed on your PC, begin by downloading the project's zipped firmware file set. Version 1.0 (dated Mar-25-2017): Vid_Cal.zip 2. Unzip the files in a Arduino working directory named Vid_Cal. 3. Next, open the Config.h file for editing and go to the User Configurable Parameters section. 4. You'll need to configure the video mode (NTSC / PAL). Edit the text to look like this: To Select NTSC Video format: #define VIDEO_MODE VID_NTSC // Video mode is NTSC //#define VIDEO_MODE VID_PAL // Video mode is PAL To Select PAL Video format: //#define VIDEO_MODE VID_NTSC // Video mode is NTSC #define VIDEO_MODE VID_PAL // Video mode is PAL 5. Because you will be flashing two different MinimOSD boards (vRx / vTx), you must specify which target type to load on each board. Configure as follows: To Select the vRx targets: //#define TARGET TARGET_VTX #define TARGET TARGET_VRX To Select the vTx targets: #define TARGET TARGET_VTX //#define TARGET TARGET_VRX In summary, select the video type (VID_NTSC or VID_PAL) and enable TARGET_VRX before flashing your vRx MinimOSD board. Then re-edit the config.h and enable TARGET_VTX before flashing your vTx MinimOSD board. 6. Set your FTDI board's 3.3V / 5V jumper to the 5V position. This voltage will be used to power the MinimOSD boards during Flashing. 7. With battery power off, proceed to flash each board using the Arduino IDE's Upload button. Reminder: Be sure to configure the vTx/vRx target choice (see step 5) before each flashing.
  9. FWIW, video level issues, like the bad FR632 vRx in this discussion, affect other FPV systems too. Not everyone recognizes that their video image's problems are due to incorrect video levels; Instead they blame other things. The only way to know if the video level is correct is to measure it. That's easier said than done! The problem is caused mainly by the lack of respect for the composite video standards. These industry documented specifications are often ignored to save manufacturing cost, other times it's due to the ignorance of the video circuit designer. Our cheap imported FPV products have been big offenders of this. To help combat the problem I've created a DiY built tool to check the video levels. It can also be used to calibrate the video level. The project is simple to build and replaces expensive/complex test equipment. Any FPV hobbyist that cares about their video signal level can now easily check it. Details are published here: https://www.rc-cam.com/forum/index.php?/topic/4126-diy-fpv-video-calibration-tool-low-cost/
  10. Assembling the vTx and vRx video generators requires basic soldering skills. And in case I have not been clear, there are two (2) MinimOSD's used in this project. First you need to connect some ground and power pads that are on the two MinimOSD boards. Just add a blob of solder across the pads shown below (Grounding pads on bottom side, SJ2 pads on component side). Some suppliers have already done this for you, but if the pads are not bridged then you must do it. The new firmware (to be flashed later) includes an optional battery voltage monitor feature. This requires adding a jumper wire, as shown below. If your MinimOSDs are the old/original version (not "KV Team") then the voltage monitor feature will require adding some 1% 1/8W resistors, as shown below. Each boards' Power and Video connections are available on the stacked 3-Pin headers. The pins are labeled on the bottom of the MinimOSD board. The vTx OSD only uses VOut and +12V power pins. The vRx OSD needs VOut, VIn, and +12V. I used 3-pin servo plugs to connect the MinimOSD boards, but direct soldering can be used instead. The video connectors you use are up to you. I installed RCA phono panel jacks and made some simple adapter cables that connect them to the FPV system. How you do this is your choice. As a reminder, the vTx's MinimOSD does NOT use the VID-IN signal, so be sure to omit it.
  11. So it's time to gather the parts to build this cool tool. As mentioned before, you need two video target pattern generators; One for the FPV vTx and another for the vRx. The target patterns are provided by low cost Arduino based FPV OSD boards. Here's the Parts List: 2 pcs, MinimOSD (KV Team version recommended). Approx $8 USD each on eBay and AliExpress. To flash the custom firmware you'll need one (1) of these: FTDI FT232RL USB to TTL Serial Converter (6-Pin version) for Arduino. Approx $3 on eBay and AliExpress. Keep in mind that there are several similar looking Chinese clones of the MinimOSD on the market. I recommend the "full size" board that has the KV Team Mod (built-in voltage attenuation resistors). Here is what the KV Team version looks like: How to tell the difference in board versions: The KV Team version will have the JP6 8-Pin header area (8 empty solder pads across the top). See photo above. It's important that both OSD's have identical video signal characteristics. Therefore, I recommend that you purchase both boards at the same time from the same supplier. This should reduce the chance of any component variations that might cause unmatched video levels. You can use your existing 7-12VDC FPV batteries (3S LiPO is fine) to power the boards. No other components need to be purchased. But if you want a power switch, A/V connectors (I used RCA phono chassis panel jacks), or plastic enclosure, then feel free to add these things to the shopping list. Although you can simply protect your OSD board with some heatshrink or duct tape, the 3D printed plastic case gives it a professional appearance. Here's the STL files for it. Case Base: case_base1.stl Case Top: case_top1.stl Case Hole Plug: case_plug1.stl Printing recommendations: ABS filament, 35% infill, 3 layer shell, 101% size scaling (shrinkage correction). There's room for a mini toggle power switch, Alkaline 9V battery, and RCA panel jacks. The "Case Hole Plug" file is a small piece that covers an unused RCA mounting hole on the vTx pattern generator (Vid-In is not needed on the vTx side). Here's how everything fits inside my vRx target generator: The vTx target generator is built the same, but has only one RCA jack.
  12. The magic behind the Video Cal Tool relies on our eyes' ability to easily see relative differences in brightness during a side-by-side comparison. The basic test setup is as follows: (1) A monochrome video overlay board provides an image reference (comparison) target pattern on the FPV display (goggles or monitor). It is directly connected between the FPV vRx's Vid-Out jack and the display. (2) A second video overlay board provides a mating target pattern. It is directly connected to the video input of the FPV vTx (the FPV camera is removed). The picture-in-picture formatted setup allows the vTx's pattern to be superimposed on the vRx's reference pattern. The mating target sets will have matching brightness when the FPV system's video level is properly calibrated. To account for the display's dynamic range and/or gamma behavior, one target set has 80% luminance and the other has 120%. Here is the vTx test pattern: Here is the VRx reference pattern: Ok, so that's what the vTx source and reference vRx targets look like. But now you're asking, how do you use them to measure the video signal's level (amplitude)? Answer: Your eyes are the "measuring" equipment. When I say using the tool is EZ, I mean it! During the test the T-target pattern fills the empty inside area on the R-target. Properly calibrated video level appears on your display monitor as shown below. But if the two interlocking patterns do not blend together (brightness not the same) then the FPV system's video level is incorrect. Here's two examples: What do you do when the brightness does not match? Fortunately this problem can be fixed on most analog 900MHz / 1.3GHz / 2.4GHz FPV video systems since they usually have an adjustable video level POT (variable resistor) in the vTx or vRx. Unfortunately modern 5.8GHz systems don't provide a calibration POT, so if the test fails you will be out of luck. But at least you will know that your 5.8GHz system has a video level problem. Here's a short video that shows how to use the Video Cal Tool:
  13. FPV Video Calibration Tool: DiY -- Low Cost -- EZ to use. The signal level (amplitude) of a FPV system's vTx / vRx composite video matters a lot. Unfortunately many are incorrectly adjusted out of the box. There's a number of reasons for this -- marginal designs, poor manufacturing QA, compatibility problems due to brand mixing by end users, cheap component drift, and just plain bad luck. Hold on you say! You are convinced that your system's video is fine. But perhaps that's wishful thinking. Here's the cold hard reality -- Many FPV systems have marginal video levels and this invites problems that are often blamed on other things. For example, the random "weak signal" blackouts we all hate are not always directly RF signal related. Poorly calibrated video levels will contribute to those random blackouts too. Plus a host of other image quality problems (e.g., image tearing, poor brightness / contrast, random sync, color loss, etc.) that are simply victims of marginal video levels. Checking the composite video signal normally requires an oscope (oscilloscope). Ideally a test pattern generator is also used to provide the 1Vpp standard video signal that is measured with the oscope. Not many hobbyist have access to this equipment or know how to correctly perform the measurements. So I experimented with video level testing using simpler tools than a oscope. My goal was to have something that worked well, but was cheap and simple to use. After a bit of head banging a clever DiY solution was born. How does ~$20 USD and a couple hours work sound to you? Yes, really. Here's what my DiY FPV Calibration Tool looks like. Spoiler alert: Inside the small 3D printed plastic box is a $8 eBay circuit board and a 9V battery. To calibrate a FPV system you'll need two of them. Both boxes will use identical hardware but with different firmware. Oh you guessed it, Arduino is involved. Come back soon. I'll show you how to build and use it. Your FPV system will thank you. The world will be a better place.