RevG Hardware Bringup
Dear Reader,
Sorry for the long delay. I was on vacation for the last few weeks. Before I left I assembled and brought up Revision G of the meter, which fixes the problems of Revision F (as detailed in my previous post here).
RevG came together faster than any of the previous revisions. Starting after posting the RevF report on March 26, assembled RevG units were on my desk on April 2. Not bad!
RevG Changes
RevG fixed the following issues with RevF:
- Pinout confusion caused by Altium pin swapping in our dual package FETs
- Current sense resistor self-heating causing significant current sense amplifier drift
- The sense resistor is now 10x smaller
- The current sense amplifier was redesigned to use an instrumentation amp with higher gain
- The current source was redesigned to use the new instrumentation amp.
Bench-top testing showed all the changes were successful. If this unit passes regulatory muster, the design will be locked down and volume manufacturing will begin!
Radio Pre-test
Before a radio device can be sold to consumers in the US, it must be shown by an approved lab to conform to FCC laws regarding the amount of radio power being broadcast so as not overwhelm any devices around it. This testing is quite expensive, so doing everything you can beforehand to make sure your device is within spec before submitting for an official stamp of approval is a good idea.
I don’t have an anechoic chamber, but I set up my RF explorer 6 inches away from a RevG Mooshimeter and ran the modem through its various test modes. Bluetooth low energy uses 3 channels, you can see the results of modulated and continuous wave transmission for all 3 channels below.
These waveforms show how much power is being transmitted by the meter (Y axis) at various frequencies (X axis). Given the informality of the setup, I don’t trust the amplitude reading as a good indicator of absolute transmission power, I am more interested in the overall shape of the spectrum and making sure there’s not excessive spurious transmission in the side-channels. These spectrums look roughly like we expect them to, good enough to send off for formal verification.
Ideally, I’d like to check the harmonics of our transmission frequency, 4.8GHz being the one of most concern. But my spectrum analyzer doesn’t go that high.
Thanks for reading!
Would you happen to know if there are additional regulatory obligations for a device with externally connected leads? I imagine it could be quite difficult to certify something with arbitrarily long metal tentacles.
On an unrelated note, I’ve been enjoying the idea of pronouncing ‘Mooshimeter’ with the emphasis on the second syllable, much like ‘altimeter’.
Fortunately we don’t have to test with the leads. The leads are COTS and their manufacturer is responsible for relevant testing.
Hi James,
Thanks for the great update.
I hope that when you go do the FCC/UL testings, you also test for Europe’s CE marking at same time. The tests are similar.
The open-source open-hardware project SmartScope just issued an exemplary EMC testing update here:
https://www.kickstarter.com/projects/751733865/smartscope-reinventing-the-oscilloscope/posts/828930
All the best at the EMC lats, they wrote in a great way their feeling there too.
I’m not that worried about the frequencies permitted to emit, but about the general EMC noise usually.
Hope to see a similar update from your project too :-)
Best Regards, and keep the great work and regular updates :-)
Beat
Thanks Beat!
Our devices are in the lab right now… preliminary results indicate we’re doing well for emissions. We’ll definitely post results from the official report when it comes in from the lab.