THE CONFIGURATION:
The av-bay is a LOC style 7.5". The "stiffy" is installed (my inner Beavis cackles at this every time). There are two rotary switches set flush into the wall. The sled is 1/2" ply (yes, overkill by a fair margin, but I had plenty of scrap laying around, so this was used). 1/4-20 allthread connects the top (permanent) lid to the bottom (removable). The sled measures 6.5" across. Sled guides are a pen from my former employer cut in half and epoxied to the bottom of the sled. Batteries are secured on the back side of the sled, with a small battery box for the Raven's LiPo, and the Telemetrum Lipo secured with a couple of wire ties. Raven 3 is mounted on the left, with its USB toward the sled edge. Telemetrum 2.0 to the right, with its USB similarly oriented.
THE PROBLEM:
It was noted that, with the Raven 3 running, there seemed to be excessive times to acquiring GPS lock with the Telemetrum. Once, thirty-five minutes had passed without acquiring lock. Sometimes, it would get lock in a few minutes... others, it seemed to take forever. Lock times without the Raven 3 running always seemed shorter. I'd considered that, at flight time, to get lock, THEN power up the Raven, but realized that I may lose lock during boost, and never reacquire it. I found this possibility disconcerting.
THE TESTS:
A string of tests were run to see if it was just me dismissing the short lock times mentally, and noticing only the long ones. So I did some comparisons running the Raven OFF and ON. When on, the Raven 3 was started first, and after it beeped out its ready state, the Telemetrum was turned on and the clock was started. The clock was stopped when AltOS reported "GPS ready." The test was run six times, and the averages were as follows:
OFF: 8:30
ON: 20:00+
It seemed clear that something about powering up the Raven 3 increased time to lock. The test was stopped after 20 mins, as this seemed an unreasonable amount of time to wait for initial lock, however, the clock was permitted to run until lock was achieved.
Raven 3 shares a common power. Its event triggers are grounds coming out of the altimeter to the event charges. Featherweight electronics offers an elegant little solution to wiring this with their Power Perch, which I have used in the past on other projects. I did not on this one because I want the rotary switches for off/on duties, not the magnetic switch inherent to the Power Perch. I'd considered attempting to modify the Power Perch to accept the rotary switch, and take off the magnetic switch. Then I looked at it. Ain't no frigging WAY I'm going to start trying to take those little itsy bitsy components apart. What I did, then, was use a #4 screw, and attach the wires to it. This was a simple solution at the time, and seemed to work well, until I noted the decrease in Telemetrum performance. I tried, then, to move the screw to see if it had any effect.
The same test was run as above, and the average of the six times were:
OFF: 6:19
ON: 20:00+
I found it interesting that the OFF time decreased with the movement. There were two of the six that acquired lock at 17:41 and 18:27, but these were offset by the excessive times from the other 4 tests. This position did show some benefit, though was not considered a full solution.
Moving to this position placed the power hub (for lack of a better term) close to the allthread on the Raven side. It was at this point I considered a Faraday cage. Then I realized I was considering a Faraday cage, and went into the living room to watch Family Guy and idiotic youtube videos on the couch with my girlfriend.
Next: The Solution...
Later!
--Coop