We finally got some testing done on the DXT. Still have not found the trouble with the board Tom populated, but Beat sent the one he had tested and found to loose calibration about 30 minutes after calibrating. Beat has a new design he is testing. The thinking was that the batteries, which are in the middle of the DXT, and much closer to the sensors were the source of the calibration not lasting. The magnetic field of batteries will change during discharge and the close proximity would amplify the problem. With the A3 Disto X, it has been my experience that calibration lasts about a month - but the batteries are further away from the sensors.
Here are the two boards. Tom's on left, Beat's on the right. The big chip is the bluetooth module.
Here is the board being installed
Here is closeup of the open DXT with board.
Tom installed Beat's board and after a few troubles with the HP IPAQ and bluetooth (nothing do to with the DXT) we calibrated. I'll spare you the details, but it took 3 tries to get a good calibration. The problem turned out to be my fault (important to not take a reading too quickly after turning on the laser during calibration) and Beat explained what was happening.
Here is DXT Disto X in test fixture with remote battery pack. We placed the battery on the floor (of the cave, about 3.5 ft away from the Disto) during testing. We did check the influence of the battery pack and wire on the Disto X. As long as the battery pack was 6 inches away we did not detect any impact. So we were pretty certain any trouble would not be coming from the batteries
After uploading the good calibration, we ran the theodolite comparison test. On the usual scale it looked pretty good.
Unfortunately, there were outliers - way far off outliers:
We are still mystified about the cause of these. We ran the cave compass test course and although about 85% of the time the numbers were good, the other 15% of the time the numbers were wildly off - like 10-22 degrees off in azimuth. Beat says it is not the same issue of taking data too quickly.
Roughly 2 hours later I repeated the Theo test:
This shows the same kind of drift that Beat experienced. And it is obviously not because of the batteries, since in our test the batteries are far away. Cause is still unknown. And yes, the magnet and magnetic switch in the DXT were removed, as were the hinge pins and spring on the bracket (entire bracket removed). There was no screw in the corner of the case nearest the magnetometer chip. Any thoughts on what could be causing a changing magnetic field within the DXT itself would be appreciated.