George Hardesty

Mar 242010

Customer’s question:   We’ve started using the PicoStation2HP in our network in station mode and have been getting a lot of complaints.  Specifically my techs are telling me they are seeing very poor performance when the client has strong signal.  I decided to run some tests of my own in the lab and have confirmed this to be the case.  Before I started digging into it too much further I wanted to drop you a line to see if you were aware of any issues like this.  Here is what I’ve seen and done thus far:

When the receive signal is at around -55 or higher I see both the RX and TX data rates drop to 1 to 5.5 mb.  Speed tests are obviously slow, 400kb/sec or so.  If I attenuate the signal, for example stick the antenna in a coffee can, the rates jump right up to 54mb.  Speed tests also recover (as expected) and return 18+mb/sec.  The coffee can will take the signal from -55 down to around -70.  As a 2nd test I decreased the TX power on my AP from 27dbm to 10dbm.   That had a similar effect as the coffee can.  54mb rates and 18+mb/sec speed tests.

PicoStation is in station mode, tried using both open/un-secured and WPA2.  Tried G-only and B-only modes.  Tried disabling dynamic-ack timeout and hard set to 100 and 30.  Tried disabling extra reporting.  Nothing has helped.  The unit is running the latest 3.4 firmware, downloaded from your site today.  I’ve tested using a Linksys AP and our Belair AP, both with similar results.

ANSWER:   For the most part this is normal.  With any high powered device, if you are too close, or have too high of a signal, you will oversaturate the receiver of the unit (think yelling vs talking in someones ear).  The best advice is to have a signal that is -50is or less, with best signal in the 60’s.

 Posted by at 12:57 pm Ubiquiti Gear Tagged with:  Comments Off on PicoStation2HP: Performance issue when the client has strong signal
Feb 102010

The problem is somehow connected to personal firewall installed in the system. It’s a Kaspersky Internet Security 2010. It adds “Kaspersky Anti-Virus NDIS 6 Filter” service to every network connection, see attached SC1.jpg.

So we have found this workaround:
1. Plug AWUS036NH and install IS_AP_STA_RT2870_D-
2. Open properties of the newly installed wireless connection
3. Disable (uncheck) the Kaspersky Anti-Virus NDIS 6 Filter, see SC2.jpg
4. Reboot PC with plugged AWUS036NH
5. Open properties of the wireless connection and enable the Kaspersky Anti-Virus NDIS 6 Filter
6. Open Ralink Utility – it works!

We don’t know whether it’s an only Kaspersky related issue or the same problem exists with all personal firewalls like Symantec Internet Security, McAfee Firewall…

 Posted by at 11:28 pm Alfa AWUS036NH, Troubleshooting Tagged with:  Comments Off on Alfa AWUS036NH on Windows Vista: Problem with firewall
Jan 202010

Data Alliance’s testing of these two devices side-by-side indicates that AWUS036NH has better range as indicated by its ability to reach more access points than AWUS036H:  AndAWUS036NH is able to reach individual access points with stronger signal.

ALFA ENGINEER agrees that this real-world style testing is more indicative of actual specifications than to use a tool such as NetStumbler to gauge the Receive Sensitivity:  HIS EXPLANATION:  NetStumber is not able to compare them equally, because NetStumber gets the sensitivity information from the device which is AWUS036H and AWUS036NH:   However, AWUS036H and AWUS036NH are using different chipset ~ and different chipset vendor has different way to identify the sensitivity ~ So, if you want to test the units by using NetStumber, it is better to use same chipset model to compare ~

If compare different chipset, my suggestion is to do field, the real environment and try to see the throughtput ~
That will be the fair testing for different chipset device.

Email which prompted the above response from engineer with Alfa:  Email is from Mark Kilty (friend and customer of Data Alliance):

Obviously the prospect of a better USB WiFi unit for us, especially one that also supports N, is huge news.  We have been waiting for an adapter to come on the market that is as good at b/g as the AWUS036H, but with N capability also.  So, we decided to put the two units (AWUS036H AND AWUS036NH) next to each other and do some testing.  Now, we aren’t scientists, and the environment is far from “controlled”, but I think you will find the results interesting.  We also realize that WiFi, although it should be a science, does not seem to be one, and so many factors play in the signals that are received and the quality of those signals.  All this testing was done in only one location, so things might be slightly different in different locations.

We used a tool called NetStumbler, which I assume you are familiar with.  It’s a nice tool to look at the performance of WiFi adapters and compare results.

In this test, we were purely looking at Receive Sensitivity, as we know that the 2W output is an improvement, but if the receive sensitivity drops much at all, the unit is not going to perform as well in the real-world in trying to pick up weak or distant signals.  Based on the specs we have, the AWUS036NH (-92/-76) and the AWUS036H (-93/-76) have almost identical receive sensitivity specs in b/g.

In the following screen shots, the first part of the graph is the AWUS035HN (Transmit: Auto), the second section is the AWUS036NH (Transmit: 50%), the third is the AWUS036H (1watt current model), and the fourth is the AWUS036H (500mw older model).

The first and second sections in the graphs match fine, which we would expect since we are looking at the receive sensitivity, not the transmit power.  You will see a slightly better performance in then 1W vs 500mw versions of the AWUS036H, which we also expect, as you improved the receive sensitivity some with the newer 1W version.  What the surprise is, the very large difference between the AWUS036NH and the AWUS036H.  I am not sure if you folks have done any testing like this before, but would be curious on your thoughts and comments on the performance difference.  I am not an expert at this stuff, but know enough, but perhaps there is an explanation for what we are seeing.  I am hoping so, as I want to continue to be very excited about this new product.

Jan 182010

Alfa AWUS036NH CD’s Windows buttons on the autorun menu in Windows are broken and do not load the setup file.  Customers will have to manually go on the CD and run the installer program.

We are not sure if this affects the entire batch that we have in stock right now.   Alfa is aware of the problem.


 Posted by at 11:48 pm Troubleshooting, USB WiFi adapters: Alfa Tagged with:  Comments Off on Alfa AWUS036NH – issue w/ Windows autorun
Jan 112010

Re. 1000mW and 2000mW USB wireless adapter specs for USB compatibility:

In the past we have stated that each of our offerings is USB 1.1/2.0 compatible and now realize that there may be an issue with the electricity flow from USB 1.1 ports (i.e., USB ports on older PCs):

Regarding data-flow:  If an item is 2.0 compatible, it is automatically backward compatible with 1.1:    But this does not apply to the electricity-flow:   The issue is the actual electricity that can flow out of a 1.1 vs 2.0 USB port.  A 2.0 port allows more, and we are presuming that the 2000Mw alfa AWUS036NH needs more than what can flow out of a USB 1.1 port:  It will probably have some issues:  Our customers have had some issues with the 1000mW Alfa AWUS036H on 1.1 usb ports.

Windows will actually give you a warning that the USB port is drawing too much power and simply shuts off the port: at least that is what we have seen when it happened on older computers.   If you don’t get a warning, all should be OK, we assume 🙂

Much of the content of this post was contributed by Mark Kilty, based on his experiences.

 Posted by at 2:07 pm USB cables & extenders, USB WiFi adapters: Alfa Comments Off on USB 1.1 ports (on older PCs) may not pass enough electricity for 1000mW & 2000mW adapters
Aug 032009

Name of software:  MacAdressChange
Full file name:  MacAdressChange.rar
Works in windows and will change the MAC address, but only temporarily:  When the user disables and then re-enables the card/adapter, it will revert to its original MAC address.
Name of software:  Etherchange
Full file name:  etherchange.rar
This software works in DOS shell mode and you can use it to change the MAC address:
Etherchange only accepts single digit adapter numbers sequences: on all Windows PCs there are two digit sequences.

We searched around for a way to remove all the old adapter numbers (XP Pro and W7), or reassign adapter numbers, but without success.

 Posted by at 11:50 am USB WiFi adapters: Alfa Comments Off on Change MAC address of wireless device: Software apps
Jul 302009

Receive sensitivity (RX) spec for our 1-watt card is -85dBm for 802.11b mode:  This is 3x the transmit power (TX) figure for the Alfa AWUS036H (500mw G-version):  That TX figure is 28dBi.  I’m trying to verify that TX is the important figure in which the range limitation would occur:  that range limitation would not occur with RX:  Because the RX of the new realtek card is triple the TX figure of the Alfa card.

So far I have not been able to find any “official” technical papers that would disprove this:  I have found supporting information in a document from one of our customers who is very tech savvy:  He says that what I’m stating above is correctre. TX being the limiting factor.
Jul 222009

A wireless USB adapter is a Transceiver:  It is the combined radio transmitter and receiver and hence its name transceiver. Alternative names for the transceiver is simply “radio”. Modern WIFI transceivers tend to be very small. Typically they are half an inch, by 2 inch, by 1/5 of an inch. Because of their small size, they can be packaged into small places: Most modern laptop PC’s contain an internal WIFI radio. Usually it is located somewhere under the keyboard. When there is internal WIFI radio, an external radio can be connected in several ways such that we will describe later.

Note: It is possible to have both an internal radio and an external radio.


Two characteristics of the transceiver are important:


  1. The power of the transmitter, and
  2. The sensitivity of the receiver.  

The power of the transmitter can be thought of as the loudness with which the radio can “talk” to a WIFI Access Point. A louder voice enables one to have a conversation over a greater distance. Similarly, the sensitivity of the receiver can be thought of as the ability of the radio to “hear” a remote WIFI Access Point.


In summary, to maximize distance, we want both a powerful and a very sensitive transceiver.


Typical internal transmitters found in laptop PCs or external USB cards have a transmit power in the 100-150mW (milliwatt) range. In comparison, the most powerful external transmitters are 500mW and above. Internal laptop PCs are quite sensitive. Their sensitivity can be in the 80-90dB range, whereas external transceivers have sensitivities as high as 96dB. To interpret these sensitivity numbers, it is important to know that a 3dB increase implies a doubling of distance that can be covered.

 Posted by at 7:00 pm WiFi card specs Tagged with:  Comments Off on USB WiFi adapters explained in detail
Jul 222009

A receive-amplifier can have some advantage if used with a highly directional antenna.  This is because the directional antenna will eliminate  (or at least reduce) other strong signals that might be nearby, which might overload the receiver.  With a non-directional antenna, the amplifier will be amplifying unwanted signals to detrimental levels — levels that desensitize the receiver or step on  (override) useful packets.

  • Analog amplification introduces a lot of noise and thus deteriorates the overall connection
  • Analog amplifier used with an omni-directional antenna:  One savvy customer verified that it merely deteriorates their ability to establish stable connections. Other people report similar results:
  • “To be frank, I have almost never seen an amp help when added to a client device. The only time I feel one is appropriate is when it is installed at the top of a long coax run in order to overcome cable loss.”    Source:

Counterpoints to the above arguments:

Distortion and noise can cancel out the advantage of amplification.  However, if the amplifier is a quality amplifier, the affects of distortion and noise will be minimized.  Problem is, the specs on low-cost amplifiers are often missing or lies.  A person just has to try it.

That said, the Alfa WiFi USB adapters already have a pretty good receiver.  Adding 15db more receive-gain might not do as well as hoped because it might be mostly amplifying noise, although I think a little more gain could help.

In our experience, an amplifier can be a noticeable benefit on the the average wireless router (if used with a directional antenna).  Years ago, we used 800mw amps to cover a small town with great success.

We believe it is best to start with a good antenna, before you try using an amplifier.

We’ve had many experiences in which a customer tries to use a WiFi booster on a USB WiFi adapter and damaged or “fried” the booster (and also situations in which the USB adapter was also fried).

What you would have to do to make the two devices work together without damaging one or the other, is to turn down the output power from the USB WiFi adapter to a level that is in the middle of whatever the power amplifier can accept.

Our customer Reed White, and RF engineer, provided all of the following commentary and info (March 2013):
If the power is turned down to the proper level (a number that is equivalent to about 15dbm output), can you say why anything would be damaged? I am well experienced with RF, and would like to understand why damage would be done if proper adjustments are made. Amplifiers are pretty common these days, and of course they would not feed their output power to the AWUS036H (instead to the antenna), although some receive gain is usually added. I guess you must have some experience with this.

Using signal-booster with an omni-directional antenna:  Most experts agree that this is a lousy option. In technical terms, it results in poor signal to noise ratio. In laymen’s terms it sounds as if someone stands next to you with a loudhailer and is tries to talk to you. You hear a lot of noise, but it becomes difficult to discern a message (due to the pain of the loud noise and spurious extra sounds that have been amplified).

With a power amplifier and with an omnidirectional (non-directional) antenna, a powerful transmitter can disturb a large area of other people’s systems.  By disturb, I mean slow down or render erratic.  This may not be an inconvenience to the guy with the high power transmitter, but it is inconsiderate of everyone else.  Therefore, it is generally considered best practice to use a very directional antenna for point-to-point connections — ideally directional antennas at both ends, with no more power than really needed.

I would say that the 1000mw of the AWUS036H is powerful enough that it should have a directional antenna in an urban environment where there may be tens to hundreds of other people sharing the same frequency.

One catch is that the 2.4GHz signals do get reflected, especially in urban areas where there are many buildings.  A person can easily get fooled into pointing a directional antenna at a reflection.  The powerful signal is then blasted where it can disturb others.  It is usually best to point the antenna directly at the other station, when possible — unless a dense building is in the way.

Amateur radio operators and RF engineers know that a good, well-positioned antenna is usually the most cost-effective way to improve range and reliability.

[end of Reed White’s commentary]

General description of use of signal amplifier or booster:  Inserting an analog booster in between the transceiver and the antenna, the RF signal can be amplified.

 Posted by at 6:48 pm Signal booster/amp Tagged with:  Comments Off on WiFi Signal Amplifier / Booster: When it helps & when not to use one
Copyright © 2004 - 2022 Data Alliance lnc. All Rights Reserved