I'd thought I share my progress of setting up a Sony Qx1 onto the Solo whilst using the Reach RTK GPS tracker. Not done finished. No need to fire it from the Solo- MP only - should be easier.
1st. This method has a priority of getting the most accurate time-stamp for you images thus I chose Emlid Reach RTK GPS. We will record the GPS position when the camera act instead of when it is command. By this means I'm not concerned if the Solo is accurately positioned in the sky but what was the position when the camera is taking the picture. Flying a survey at 8ms with a 50 ms delay would cause a .4m delay error. So I ask myself why would spend the $ for a RTK GPS on a sUAV while having any-kind of measurable delay? This method here is can only be faster if we speed up electrons.
2nd. The Sony Qx1 has no hot-shoe or wiring for a external flash. We had to open up and wire directly into the shutter motor. This is a difficult and sensitive hack. Not for the fat fingered or faint hearten as you can easily break a brand new camera. Try to hack the flash control module was another layer or two into the camera so I backed out.
The shutter motor has a lot of noise in yellow (feeding a sparkfun optoisolator). The blue is the opto out. This method caused several error/extra reach time-stamps. It varied from 0 to 60% more reach time-stamp then camera images. I also got a intermittent capacitor discharging after the shutter was finished which was causing havoc.Adding filter circuits had poor results.
I added a Adafruit Tinker 5v MCU and created a one-shot signal to hold for 50 ms then force it off for another 200 ms before resetting the one-shot. The result would only allow one input every 250 ms where as the Qx1 is only capable of shooting once every 333 ms.
This is the test harness with the red plug(hacked) into the Qx1 feeding a LED/Diode to the optoisolator, into the one-shot MCU and to the reach data logger
This is the wiring.
Feedback welcomed. reference only as it may contain error.
1st. This method has a priority of getting the most accurate time-stamp for you images thus I chose Emlid Reach RTK GPS. We will record the GPS position when the camera act instead of when it is command. By this means I'm not concerned if the Solo is accurately positioned in the sky but what was the position when the camera is taking the picture. Flying a survey at 8ms with a 50 ms delay would cause a .4m delay error. So I ask myself why would spend the $ for a RTK GPS on a sUAV while having any-kind of measurable delay? This method here is can only be faster if we speed up electrons.
2nd. The Sony Qx1 has no hot-shoe or wiring for a external flash. We had to open up and wire directly into the shutter motor. This is a difficult and sensitive hack. Not for the fat fingered or faint hearten as you can easily break a brand new camera. Try to hack the flash control module was another layer or two into the camera so I backed out.
The shutter motor has a lot of noise in yellow (feeding a sparkfun optoisolator). The blue is the opto out. This method caused several error/extra reach time-stamps. It varied from 0 to 60% more reach time-stamp then camera images. I also got a intermittent capacitor discharging after the shutter was finished which was causing havoc.Adding filter circuits had poor results.
I added a Adafruit Tinker 5v MCU and created a one-shot signal to hold for 50 ms then force it off for another 200 ms before resetting the one-shot. The result would only allow one input every 250 ms where as the Qx1 is only capable of shooting once every 333 ms.
This is the test harness with the red plug(hacked) into the Qx1 feeding a LED/Diode to the optoisolator, into the one-shot MCU and to the reach data logger
This is the wiring.
Feedback welcomed. reference only as it may contain error.
