Hacking the Sony Qx1 for output trigger/ external flash

Finally got the Qx1 to output a external trigger.


I had hoped for a easier solution but we ended up with two wires coming from the shutter motor. This photo shows the Qx1 being manually trigger and the oscilloscope shows the shutter motor voltage. This needs to be tuned in and will be used to signal the Reach GPS tracker. Happy surveying.

We are not done yet. Some final tuning, testing and adding a small plug to the case of the Qx1.

This is a harder hack as some disassembly is required and you could ruin your camera. The motor is buried. Once we get the details worked out my cousin will consider doing it for a cost as a professional electronic service in the range of $150. If you don’t care if your camera breaks he might do it for $100. It takes about two hours with someone highly skilled in the small details. More than me and I am a major tinker, hacker and builder. You will need to solder two small wires on the back of a small header connector without making contact to adjoining pins. A very small low powered iron is needed.

I did I discover.

  1. The multi USB ports has no accomodating output for external flash.
  2. We can get a pickup from the Qx1 flash’s primary coil. I disconnected the flash itself so the capacitor would stay charged so the camera is never delayed until the cap is recharged. This is still an alternative to someone not comfortable taking apart a camera.

Now forgive me but I have been calling this an external trigger as in a output from the camera. This is a non delayed output as the signal is being generated directly from the shutter motor being energized.

Questions about the Reach
What is the input specifications? I know we need 3.1 volts for a input .

  1. Is it optically isolated?
  2. What is the max input voltage allowed.
  3. What is the min to max duration for input

Our 1st test has the signal at 8vdc for 50ms.

When can I get one to test on the Solo. We recorded the whole tear down but have not checked the quality of it. If time frees up I could post something.

Burke

6 Likes

found the answers. Will plan to add a opto coupler and drive it with Reach’s 5vdc

2 Likes

Burke, this is very exciting!

What might be the drawbacks of this method? If it allows not to disassemble the camera it might be the best one.

Reach has a BAV99 diode clamp to 3.3V on time stamp input, so should survive 8vdc. It should be further tested though.

2 Likes

The Using the flash is slower. 1 frame per sec where no flash is 3 fps - tested with scope on motor leads. The flash that was hacked is only1 fps but inconsistent in firing which leaves me concerned. That being said the least invasive is a inductive or photo sensor over the flash with a slow fps. I’m looking for the the Qx1 high meg and speed so I shall hack away for the3fps. Again many thanks to “timvand” for saving me a few swear words.

Ordering reach today

2 Likes

I am wondering if maybe the flash controller was not completely fooled and is limiting the fps?

1 Like

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.

1 Like

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

1 Like

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.

1 Like

Will do. thanks you

1 Like

Successfully triggering the Reach with a Sony Qx1

Test harness.

  • 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.

Burke

4 Likes

Nice work Burke!

1 Like

Looks good, curious to see the flight tests! :slight_smile:

1 Like

Burke,

Have you had success when testing out the camera?
I’m very interested in getting the flash feedback working!

Best,
Steve

Hello Steve,

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.

Burke

Hi @Burke_Mitchell (@Bam_Mitch) ,

I sent a PM to you.

Thanks!
Steve

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 :frowning: