Originally Posted By: tfabris

Is there a chance that they deliberately throttle the data bandwidth based on the number of bars? In other words, could the reported drop in data throughput still possibly be due to a software bug in the bars calculation?

Or maybe the stuff they're saying about the bars calculations is actually an oversimplification of the more complicated issue of baseband frequency switching that some have suspected might be causing the problem.

I still think that is a possibility, though seeming much less likely now. It seems that once you trigger the issue it takes it 30 seconds or more to recover:

- start a speed test, get 2000k
- touch the iSpot
- speed drops quite rapidly to 50k
- release the "death grip"
- speed stays pegged at 50k
- cancel speed test, run again
- speed still under 150k
- wait 20 seconds for bars to return a bit
- run speed test, back to 2000k
_________________________
Remind me to change my signature to something more interesting someday