Quote:
Originally Posted by esquared
I should have been more specific in my previous post, the medium power amplifier I will try out has a spec'd frequency range of 1-8GHz, but continues to exhibit reasonable gain and output power specs for another 2-300MHz below it. The 5V/m is my best guess at effective radiated power at these "unspecced" frequency ranges.
You've got an excellent point about AGC (Automatic Gain Correction, which in RF systems is intended to avoid saturating your receiver, which leads to data that all looks like the same, far too high level). To test this, I'd have to crack the case on the modem, and somehow I doubt our team leader will go for this  Anyone void their warranty and see how they implemented this year's radio?
Another possibility: I've seen front-end bandpass filters that had poorer than expected rolloff outside the band, resulting in a susceptible front-end. Since the hardware repair to the first team's radio was relatively quick, replacement of a poor quality filter would fit into this category. Additionally, it may not be an issue with all filters, so some teams could be operating just fine.
In the end, I expect this to be similar to last year's 8.2V problem where some teams had it, some didn't. Very similar in that both resulted in non-responsive or potentially destructive robots! |
Eric,
Pictures of the radio's internals are part of IFI's FCC OET, EA public file as is the emissions test report performed on a prototype unit.
Based on my own observations of the problem, an Electrical Fast Transients test, such as EN 61000-4-4, might be more revealing than an RF succeptability test, but certainly a lower frequency, 100% AM modulated signal can reveal problems that an EFT test would also expose.
Again, without motors running, we saw no indication of packet loss.
Our build location is surburban with no close field radiators, let alone any dynamic radio transmission that would be tightly coupled to motor actuations.