I am wondering if maybe the flash controller was not completely fooled and is limiting the fps?
It wasn’t fooled. To access the flash controller is more difficult then the shutter motor thus where our solution is heading. Adding our circuitry should be only as invasive the oscilloscope -hopefully. Testing will be needed to verify.
Update on Sony Qx1 output to trigger the Reach. I will be able to trigger the reach exactly when the shutter start to move but I might have some issues.
Please see the attached photo. I’m not sure if I have to process/filter the input signal. Currently I’m using the shutter motor voltage(YELLOW) optically isolated to a USB 5vdc for a input to the Reach(BLUE). This is a bench test of the input as I am waiting on my Reach. just prepping
Will this input signature work?
Does the Reach have a make/break delay? What is it? Once it has a input HI what is the minimum time before it will allow another input HI to be process? Is this adjustable? In short I know the Sony Qx1 fastest published shutter rate is 3 fps or once every 333 ms. In this case I’d like the Reach ignore the input for 250ms (make /break cycle) so if the input is noisy no false triggers. If so I’m ready for my Reach. Else I need advise on how to adjust.
We can see the shutter closing for about 5ms after a 50ms capture. The signal is erratic during this time. Additional I have seen a trailing voltage spike about 10 ms after closing - must be a cap discharging. If we have the delay between input then all this noise is mute.
Thanks for your input
Great progress! Reach will detect all pulses, so it is important to feed clean signal to it. Rate of time stamps is limited by the update rate. A simple RC debouncer could be used it this case.
Will do. thanks you
Successfully triggering the Reach with a Sony Qx1
- Red plug coming from hacked Qx1 into Optoisolator.
- Input has a diode/led to eliminate the negative spike from the motor.
- Opto out to a MCU to digitally clean the signal into a one shot rise on for 50ms then off for 150ms before resetting. This limits any noise from the motor and internal cap discharging.
- Cable lengths adjust to installation.
Current test have 50 camera triggers and 50 time stamps.
1st test cable to be fitted into the solo.
So now the signal is clean with no secondary spikes/triggers. The diode did also clean motor noise but still wasn’t good enough until adding the MCU.
The harness uses a LED out of a cheap flashlight, Sparkfun optoisolator pcb and a Adafruit Tinker 5v
Here the wiring
I’m not sure how far I will go but thinking of creating a 3d printed box that would plug into and mount on the Solo accessory This would house the reach and harness components.
OK onto firing the Qx1 with the Solo and the wiring connection. Any time saving inputs much appreciated.
Nice work Burke!
Looks good, curious to see the flight tests!
Have you had success when testing out the camera?
I’m very interested in getting the flash feedback working!
I cannot access my account so I had to create another one. We are modifying the Qx1 by adding wires to the shutter motor which has failed few time bricking the camera. This method would give use 3 frames per second. We are unable to access the primary flash circuit so we are looking to use an induction pickup from the flash which will only be one frame per second. We have someone that will provide a service to hack the camera. The cost will depend on which method is used, risk involved and who buys the camera if it is bricked. Fell free to pm me on this account until I get the other one working.
Update. Hello all, sorry about the slow updates. I’ve been working on a urgent client project which has suck all time and life out of me. We will be getting back on this hopefully in the next two weekends
Steve. filled the contact form on you website
Hi, I take a place here, I´m planing a quad with my qx1.
sadly I don’t have a clue about electronics, well, only basics and I can do some tests if you want, take apart is not a problem, so pm if you need it.
Have you seen this I guess http://diydrones.com/profiles/blogs/using-the-intel-edison-as-a-smart-camera-controller
Others known problems are, not signal video for fpv and the flip sensor when it faces down. If you have some ideas for those I am all ears.
I finally got back to more testing on the Qx1. Using a flash pick up will not work. The Qx1 appears to have a lower grade recharge and cap circuits. After 20 to 30 flashes at two second interval it lags and quits until you let the camera sit for a few moments. We have been unsuccessful at hacking the primary flash circuit. If the flash id set to on even with the flash removed the camera will not perform properly. We are going back to hacking the shutter motor inductively. The internal testing is very tricky. We have bricked 3 Qx1 just trying. I’m overloaded for the few two week so it will be the end of July before I have more results. Will update then
Edge 540T, The FPV is desirable but not currently probable/possible. The Qx1 is a great platform for UAV but not built for it. It doesn’t have a hot shoe nor live feed so I fly from MP blind.
@Burke_Mitchell Thanks for the update! I agree about the QX1… it’s the closest we have to a great UAV camera sensor… but missing some very important features
Hi, it’s just an idea and not mine, I’m not an expert, have you thought the output audio signal as trigger, don’t know if it has some delay or not, just seems easy to do it.
Good idea but the audio would most likely need to be processed to grab only the noise signature of the shutter action while eliminating all other noises while in flight.
Very interesting work you have done @Burke_Mitchell something I would definitely want to try to do with my Qx1 once I solve the problem I have now. Currently I have one big problem and that is to trigger the camera fast enough by a remote through the multiport. I am nowhere near the 3fps probably closer to 0.2fps. What settings on the camera do you use to trigger it fast enough? Currently I am using as a trigger the Map2 from Seagull UAV.
may be the firmware of sony umc-r10c will work in qx1 https://store.3dr.com/products/site-scan-r10c