@aleksorsist Oh hey it's TAXI. You should chat to @miek: https://www.eevblog.com/forum/thermal-imaging/getting-digital-data-out-of-my-pm695/
Martin Ling
- 40 Followers
- 62 Following
- 10 Posts
Wandering engineer (he/him)
@aleksorsist I'll take a look. Are there particular aspects you'd like feedback on?
@agowa338 @lexr @jonty For USB High Speed (480Mbps) signalling to work, the D+/D- pair needs to be close to the right differential impedance (90 ohms) all the way along the signal path.
Proper USB cables have twisted pairs of the right geometry to hit that. A random DB25 serial cable from 1993 does not. So the signal bounces around inside the cable and interferes with itself.
However, the High Speed negotiation only involves a 10kHz square wave. So that bit works, but 480Mbps packets don't.
@lexr @jonty It completes the High Speed chirp negotiation, but fails to exchange actual packets. Which is not exactly surprising, but it's cool that we can see it :)
@mjg59 Welcome to the future! It's nice here :)
@destevez This version looks much better to me.
The main thing that still stands out to me is the use of thermal reliefs on the copper pours. In general these are usually best avoided for GHz work. But they do make things easier to hand solder. Again, I don't think it's anything that will stop this board from working, just general advice.
@destevez The key benefit of 4L is not better impedance control - that can be done on any stackup. It's really that the layer pairs are so much closer together (~0.1-0.2mm). So better isolation, because fields are more tightly contained. Shorter vias, which means lower inductance. More inter-plane capacitance, with low impedance at high frequency that contributes to better decoupling. Lower losses, because you're less dependent on the properties of the dielectric. It helps in lots of ways.
@destevez In general I think you'd have better results with a 4-layer stackup and the cost over 2-layer is negligible these days, but I don't see any reason the 2-layer won't work. I'd suggest grounding the NC pins on the HMC319 though. Then you can get a much better transition from the RF output into your CPWG line. It's safe to do on that part, see e.g. the reference board: https://mm.digikey.com/Volume0/opasdata/d220001/medias/images/1073/105706-HMC391LP4.JPG
@tnt Ah, OK. I don't personally know of any integrated package for that. Getting color temperature and CRI are fundamentally curve fitting tasks though, and I have seen Fityk (https://fityk.nieto.pl) being used for viewing spectral data and curve fitting to it, in IR spectroscopy and similar techniques. Maybe a starting point, or part of a workflow.
@tnt Not sure what you're looking to do, but if it involves color then Argyll CMS (https://www.argyllcms.com/) has driver support for several spectrometers, and does the necessary math for making ICC profiles etc. It doesn't have a GUI, but there are third-party ones. I've used it for profiling screens, cameras, etc.