We've been getting a lot of questions about how a new wireless USB adapter from a company called Gsky stacks up to the Alfa AWUS036H 500mw USB adapter that we sell (we ran a head to head test which we will discuss further down, but if you want to get right to the test results, click here).
The Gsky adapter looks a lot like the Alfa model, it also boasts the same maximum output of 500mw. We went ahead and got one of the Gsky adapters online, after all if there is an adapter out there that can perform as well as the much heralded Alfa long range adapter, it is certainly something we would want to inventory.
The first thing we noticed when we hooked it up and looked at some of the documentation was that claims on some web sites that it is a clone of the Alfa AWUS036H 500 mw adapter are not true. The Alfa adapter has a Realtek 8187L chipset, while the Gsky has the 8187B chipset, also from Realtek but with lower receive sensitivity. We should note for Linux users that the Alfa 500mw adapter is plug and play in Backtrack 3, while the Gsky with the 8187b chipset is not- it was not recognized by Backtrack 3. We also found out that with the Gsky you can only transmit at high output when using 802.11b throughput or lower. If you want to have an 802.11g connection (54 mbps), the output drops to just 30mw (yes, that's thirty, not 300).
The biggest difference though was the actual receive sensitivity. Here is a look at what you get with the Alfa (the numbers recorded are negative, so in this case the lower the better, as such, -90 means greater sensitivity than -80):
11 mbps -91 dBm at 8% packet size
54 mbps -76 dBm at 10%
Here are the numbers for the Gsky:
11 mbps -82 dBm at 8%
54 mbps -65 dBm at 10%
Seems like a pretty big difference, especially at the 802.11g/54 mbps level.
Okay, but numbers aren't everything, so we decided to pit both adapters head to head using an indentical 8 dBi antenna to see how the above differences manifested in actual usage. The adapters were positioned in the same place, one after the other, with the same 8 dBi antenna connected to each one. The tests were run 2 minutes apart. For comparison purposes, we also tested the Alfa 50mw (fifty mw) AWUS036E model. We have put the results into this PDF file here.
The final conclusion?
Alfa 500mw AWUS036H- detected 16 APs, 3.2/5 bars average signal strength
Gsky 500mw- detected 9 APs, 1.5/5 bars average signal strength
Alfa 50mw AWUS036E- detected 8 APs, 2.9/5 bars average signal strength
As the test demonstrates, the Alfa 500mw adapter gets the best range, and the best signal strength compared to the Gsky. The Gsky gets a slightly better range than the AWUS036E model from Alfa, which we have priced at $24.97 at the time of this blog post. But the AWUS036E model is more sensitive.
27 comments:
Good article, I'm interested in purchasing an Afla usb adapter from you guys in the near future. Would it be possible for you to run similar tests with the AWUS036E vs. AWUS036EH, as the EH is supposed to have a much higher output.
Maybe you could also include the AWUS036S also, since it's using the Ralink chipset (RT73?) while the others use the the Realtek 8187L.
Thanks
Hi Dan,
Thanks for your comment. We will add that to our list. We have a testing planned shortly that will pit the AWUS036EH 300mw adapter vs. the Gsky and a similar entry level palm-sized 500mw we will be carrying in a few weeks, and we will likely be testing the EH model vs. the E and H models of the Alfa as well- just to show the differences. Adding the S model is a good idea.
-RokBlogger-
Yes do run a test ASAP since I am awaiting your testing in order to buy one.
Great news. Looking forward to your upcoming tests. Thx!
An ETA for the new testing results is Monday, 2/23.
Thanks for the update. Will check back then for the results.
We've complete some new tests as promised. For a test of the Alfa AWUS036EH 300mw USB adapter and the Gsky 500mw adapter, click here. For a test of the Alfa AWUS036E 50mw USB adapter, the Alfa AWUS036EH 500mw adapter, and the Alfa AWUS036S 80mw (Ralink) adapter, click here.
That's interesting that the Gsky (using the Realtek 8187B chipset) wasn't automatically detected in BT3. I would have (incorrectly) assumed that the rtl8187 driver would support it.
Though there is just a one-letter distinction in the chipset number, it is a completely different chipset with a different driver package. The same with some older adapters using the Realtek RT8185 driver. For Mac and Windows it is the same way, you cannot use the 8187L driver for an 8187B device. I have never tried to force install it in Windows, but am guessing it would give a code 10 error, just like installing any other incorrect driver. I'll have to remember to check it out.
I would imagine that minor difference (and thus incompatibility) cause a lot of grief for users. I wonder why Realtek simply choose to change one letter vs. giving it a different chipset model number.
Good question. I have always wondered that about a number of manufacturers. Senao/Engenius have often distinguished from products with and without antenna connectors by adding an EXT to the end of the model number, for all intents and purposes giving the versions the same model number but with one having an EXT. As a longtime seller of Engenius products, that has caused confusion, and so as you can see with the Senao B cards we sell that do not have a connector, we make sure to highlight this fact at the top in bold letters.
Linksys is another company that uses one letter distinctions. Probably someone more astute than myself about assigning model numbers to products can give some good reasons why the practice is better for businesses, but I can't think of many.
Of course the most confusing are products that have the same model numbers but multiple version numbers, often times with each version having a different chipset or different compatibility...
GSKY is better ....
please see real video at
http://www.youtube.com/watch?v=VUmVDmw7nsw
1) Youtube is full of fail.
2) All documentation I've come across regarding the Gsky is that it uses the Realtek 8187B chipset, not the 8187L found in the Alfa.
3) While the video "appears" to indicate that the Gsky has discovered 1 more wireless network (3 total) than the Alfa, Rokland is a well established company that has no reason to fabricate results about the Alfa which itself is a proven product in the wireless community.
Thank you for the link. A couple of things I would just note about the test you linked to. One would be that while the video shows a Gsky adapter on the left, the other adapter looks like an Alfa adapter but does not have the Alfa logo and appears to be white in color vs. Silver (the Alfa is the same color as the Gsky, but in your video it does not appear to be the same color). Color aside though, from speaking with Alfa we have been told genuine Alfa products have the Alfa logo prominently displayed on top, and that is certainly the case with every AWUS036H model I have ever used.
Another thing to note, the testing is done on two different laptops and the adapters are about a foot apart. The Gsky model is also hooked up to an antenna that appears to be twice as tall as the one connected to the Gsky. What is important about the spacing between the two antennas is that as wifi testers know, moving an adapter even six inches can affect the scan results. For testing conducted here at Rokland, we make sure each adapter uses the identical antenna, and also make sure the units are tested in sequence, in the same place, and that the tests are repeated to detect any anomalies. We also test on the same laptop. In this video, two different laptops are used. While that is unlikely to make a difference, it is possible depending on the power level of the USB ports on each.
We appreciate your sharing your own test results with us but would encourage folks to take all these things mentioned into consideration, and also look at the receive sensitivity of both adapters. Dan is right, the Gsky uses the less-sensitive 8187B version of a Realtek chip, which is about -11 dBm less sensitive than the Alfa.
Rokland, do you have any comment about Alfa claiming their product is 500mW when 500mW is the maximum current the USB bus can provide the product? Assuning no power consumption, this would be possible, but as we know this product - like all others - draws power from the USB bus which makes any claims of 500mW power output a bit debatable in my mind. Would appreciate an explanation in this area if you could. I'm surprised this advertising claim has been overlooked for so long.
USB power output is 500mA not 500mW.
Carlos is correct. Electric power from the USB port is not the same as the output power of the adapter. There is some misconception out there about this and about using a dual connector cable with an Alfa device vs. a single connector cable, that two connectors are required in order to get the max output power. It is not correct. If the device can be powered by the computer and function, that in can operate properly and to full capacity. Here is an article that confirms what Carlos wrote:
http://powerelectronics.com/ar/Maxim.pdf
Let me also add that because we do not regularly monitor comments on blog posts, especially as they get older, if anyone has any questions about the specifications of any product, please feel free to post a comment but also please contact us directly by going to our web site http://www.rokland.com and selecting the Contact Us link there. We can then provide you with more details and/or put you in touch directly with the manufacturer if needed.
I was just shopping on Ebay and a search for "8187L" turns up several posts showing the Gsky units with the 8187L chip on the board. Has anything changed, is the Gsky now using the 8187L chip or just a marketing ploy?
BUY AN ALFA'36H = HAPPIER THAN YOUR WILDEST DREAMS. DO NOT HESITATE - IM TELLING YOU YOU'LL LOVE IT! I'm ordering another... that's how much I love mine. No need for dumb 'cans' I got some rp-sma cable, a 12 db omni-direct antenna, and an Alfa, and I cannot be happier.
Can someone please email me the GSKY 27 USB driver for XP? Thanks! seanwilliamthomas@gmail.com
I inclination not approve on it. I think warm-hearted post. Particularly the appellation attracted me to be familiar with the whole story.
I have a Gsky and an Alfa. From my experience, the amount of "bars" you get isn't a good way to compare WiFi adapters. I have done test where the Gsky registers more bars than the Alfa, but the Alfa will connect to more APs and get better data transfer rates. The Gsky can't connect to some APs even when it registers 5 bars. When it does connect, it registers a speed of 54Mbps at first, but the rate drops to 1Mbps when something is actually downloaded. It is very deceiving. In my experience, the Alfa is a much, much better adapter.
^^^^^^^
This last comment represents EXACTLY what I experienced with Gsky 500mW. 5 bars and no stable connection. It was working 1Mbs anytime something was downloaded from net. Actually, it was much worse than TP-LINK 422 USB adapter whick cost me 2.5x less. To bad I can't test Alfa, but GSky was horrible from my knowledge
GSKY accept change MAC on windows 7 ?
Wow, that's crazy man. They should really try to do something to fix that.
Post a Comment