Jeff (or others if they know) can you take a look at these test reports and give me a better idea of which of these units will most likely bring in the most stable; best signal on a 802.11b old desktop? Have already checked OS support for these radios, but I don't have the expertise/knowledge to evaluate the test reports. I will be using a home brew parabolic reflector behind the antenna to maximize signal strength. Getting fair to good now with a good pcmcia card, but want to go to a desktop that has not ethernet or pcmicia support, hence usb. Let me know if I have left anything out. Also, for future reference aka evaluation for dummies, which stats in the test report are the best indication of quality of the radio?
No, I can't. All the FCC reports will show is TX power output and occupied bandwidth. There's nothing on RX sensitivity, dynamic range, BER/PER tests, and interference/reflection rejection. They do show the chipsets used, which will probably give the most useful info:
Yep. A better antenna will always improve range and reduce junk pickup from the sides and back.
Make sure it's USB 2.0. You didn't bother supplying the operating system, but my guess is that if you insist on 802.11b, it might be ancient.
Equipment used. Access point or wireless router used. Why only 802.11b? Operating System and revision level. Estimated range? Line of sight? Makers and model numbers of radioA thru C.
Not much. There's also no single number or graph that defines quality. It's also possible to take a perfectly good chipset, and ruin the performance with a buggy driver. This is all to common with USB, which is why I prefer ethernet. If pressed, I tend to look at the graph of RX sensitivity versus speed at a fixed BER or PER. That's not on the FCC reports.
802.11b=19.2dBm 802.11g=18.3dBm (about 80mw) Huge heat sink which makes me wonder if it get hot. RTL8187B and SiGe 2527L chips. I'm not happy with Realtek drivers and am unfamiliar with the 2527L chip.
TP-Link TL-WN422G Atheros AR2524-AQ1C chip. I can't identify the RF chip. General layout looks nice. I like the PCB RF bandpass filter near the antenna connector, but it might be lossy. Dunno.
20.8dBm max power output (about 100mw)
Rosewill RNX-G1 or RNX-G1W TX power 12-13dBm. (about 11 to 20mw). See 8.4 of test report. Realtek RTL8187B. I can't read the RF chip, but it's another Realtek, possibly an RTL8255. I have the same problems with Realtek on this unit.
This is a no brainer. The 2nd one (TL-WN422G) has the best chipset, the best layout, and highest Pout. (Incidentally, this took 45 mins).
Jeff Liebermann wrote in news: firstname.lastname@example.org:
Very sorry, had I known the reports had nothing really good in them, I would not have asked you. Unfortunately the TL-Wn422G has problems with linux, one of the OS's I will be using in addition to XP. The Rosewill unit runs hot according to reports and is cheaply built. I will try to do more of the homework on my own next time.