5 ms·

Even with nonlinear amplifiers, you can still drive it too much, you can still have distortion, and you're still going to have to test with a scope regardless.

Even with nonlinear amplifiers, you can still drive it too much, you can still have distortion, and you're still going to have to test with a scope regardless. There's plenty of guidelines out there describing how to minimize harmonics, like this one http://www.tek.com/support/faqs/what-fft-fast-fourier-transform-math-function-oscilloscope-useful http://www.tek.com/support/faqs/what-fft-fast-fourier-transf..., or this one http://www.mpdigest.com/issue/Articles/2009/feb/CEL/Default.asp http://www.mpdigest.com/issue/Articles/2009/feb/CEL/Default.... or this one http://www.evaluationengineering.com/articles/200611/rf-measurement-basics-for-non-rf-test-engineers.php http://www.evaluationengineering.com/articles/200611/rf-meas... . This is one of the reasons the FCC is proposing rules like this, people don't realize the unintended consequences.

kazinator 11y ago
If the hardware register being in range causes the circuit to be out of its range, that's a hardware problem. It's not my problem. The FCC should take that up with the hardware designer. I don't supply the input signal to that amplifier; it's all within the router. The hardware should know the amplitude of the input signal and what maximum gain can be applied to it before these detrimental effects set in. It should not permit higher gains than that. I trust that it doesn't, and don't give a shit beyond that. I'm not going get a multi-gigahertz-capable scope costing tens of thousands of dollars just to tweak my home router.
- Sanddancer 11y ago
  Your attitude towards the problem is exactly why the FCC is making these rules. Enough people not caring about the rules is why the FCC is putting their foot down and saying that hardware makers have to make their devices a black box.
  - kazinator 11y ago
    I care about the rules and I assume that the hardware takes care of them. The hardware is FCC certified and legal for sale to consumers. I don't think that the hardware has to be a black box to be compliant; it just has to disallow combinations of parameters that are outside of the legal (or technically correct) area. If the chips themselves enforce this, then we can still have a device on which you can run OSS firmware. And anyway, that's the right way to do it! If the hardware doesn't enforce the rules, and you're relying on the black box combination of hardware and software, that is less safe. Maybe exploits can be found to get around the software, and then -- wee -- let's crank that amplifier until it glows, and use illegal channels, etc.
    - structural 11y ago
      "If the chips enforce this" -- at this point, you're talking about hardware that is only really usable in one country. Yes, the US market is large and important, but it's not the only one. This is why things are done in firmware, so the factory can configure particular runs of boards for legal sale in whatever country they're attempting to manufacture to sell in that day. Now, for the sake of it being cheaper, many devices may have just one big firmware blob: this is more common in devices like wifi routers than other things (e.g. cell phones). I am a researcher in RF-related technology: my personal views are that the radio firmware should be kept separate from the rest of the software running on the device (fairly similar to what many phones do with a separate baseband), with the ability to upgrade the radio bits being restricted to a) the manufacturer and b) users that the FCC has approved to make such modifications -- this would provide a sane way to certify groups who can make these changes, both for research purposes and for the OEMs that are integrating the RF capabilities into some larger consumer device. In addition, there's many times where we've purchased hardware and performed modifications in-house to the control software, specifically because we DO have a license from the FCC to do things outside of what would be allowed for normal consumers. Forcing special hardware runs for things that aren't locked down and can be used by the couple hundred groups (if that) in the country who do this sort of thing is going to be even more prohibitively expensive. Finally, I'm extremely aware of the thought processes that would lead manufacturers to lock everything down if they're forced to lock anything down, and this could be mostly worked around by making the manufacturer of the RF hardware (who typically isn't the company selling you a device) responsible for those protections. They won't have any desire or ability to touch the other parts of a device, so they won't really be technically able to lock them down in the best case scenario. This gives a lot of freedom for the higher-level software to be kept in a state where it's easy for users to modify.
- m4x 11y ago
  Why should the FCC take that up with the hardware designer? Should the police take it up with Ferrari when people use their cars to speed? You should operate your hardware within the limits specified by your local laws. Saying "I trust my hardware won't do anything wrong and don't give a shit beyond that" is not acceptable.
  - kazinator 11y ago
    "No distortion from transmitter that spews noise across the spectrum" is a technical limit. It is not valid operation merely forbidden by law.

kazinator 11y ago

If the hardware register being in range causes the circuit to be out of its range, that's a hardware problem. It's not my problem. The FCC should take that up with the hardware designer. I don't supply the input signal to that amplifier; it's all within the router. The hardware should know the amplitude of the input signal and what maximum gain can be applied to it before these detrimental effects set in. It should not permit higher gains than that. I trust that it doesn't, and don't give a shit beyond that. I'm not going get a multi-gigahertz-capable scope costing tens of thousands of dollars just to tweak my home router.

Sanddancer 11y ago
Your attitude towards the problem is exactly why the FCC is making these rules. Enough people not caring about the rules is why the FCC is putting their foot down and saying that hardware makers have to make their devices a black box.
- kazinator 11y ago
  I care about the rules and I assume that the hardware takes care of them. The hardware is FCC certified and legal for sale to consumers. I don't think that the hardware has to be a black box to be compliant; it just has to disallow combinations of parameters that are outside of the legal (or technically correct) area. If the chips themselves enforce this, then we can still have a device on which you can run OSS firmware. And anyway, that's the right way to do it! If the hardware doesn't enforce the rules, and you're relying on the black box combination of hardware and software, that is less safe. Maybe exploits can be found to get around the software, and then -- wee -- let's crank that amplifier until it glows, and use illegal channels, etc.
  - structural 11y ago
    "If the chips enforce this" -- at this point, you're talking about hardware that is only really usable in one country. Yes, the US market is large and important, but it's not the only one. This is why things are done in firmware, so the factory can configure particular runs of boards for legal sale in whatever country they're attempting to manufacture to sell in that day. Now, for the sake of it being cheaper, many devices may have just one big firmware blob: this is more common in devices like wifi routers than other things (e.g. cell phones). I am a researcher in RF-related technology: my personal views are that the radio firmware should be kept separate from the rest of the software running on the device (fairly similar to what many phones do with a separate baseband), with the ability to upgrade the radio bits being restricted to a) the manufacturer and b) users that the FCC has approved to make such modifications -- this would provide a sane way to certify groups who can make these changes, both for research purposes and for the OEMs that are integrating the RF capabilities into some larger consumer device. In addition, there's many times where we've purchased hardware and performed modifications in-house to the control software, specifically because we DO have a license from the FCC to do things outside of what would be allowed for normal consumers. Forcing special hardware runs for things that aren't locked down and can be used by the couple hundred groups (if that) in the country who do this sort of thing is going to be even more prohibitively expensive. Finally, I'm extremely aware of the thought processes that would lead manufacturers to lock everything down if they're forced to lock anything down, and this could be mostly worked around by making the manufacturer of the RF hardware (who typically isn't the company selling you a device) responsible for those protections. They won't have any desire or ability to touch the other parts of a device, so they won't really be technically able to lock them down in the best case scenario. This gives a lot of freedom for the higher-level software to be kept in a state where it's easy for users to modify.
m4x 11y ago
Why should the FCC take that up with the hardware designer? Should the police take it up with Ferrari when people use their cars to speed? You should operate your hardware within the limits specified by your local laws. Saying "I trust my hardware won't do anything wrong and don't give a shit beyond that" is not acceptable.
- kazinator 11y ago
  "No distortion from transmitter that spews noise across the spectrum" is a technical limit. It is not valid operation merely forbidden by law.