How to set tx power UP (increase tx power) in Windows 7

Don't you mean RT8187L? Just do the same as you did in XP. I have the TX slider using Realtek in 7 so I don't know why didn't you try it?

 

An ALFA 036H has a 8187B chipset that's news to me!

 

But does it have the imperfection (looks like soldiering iron mark) in the top shell on the left side right next to the sma connector?

 

No need for any replies with the casing smudge info because there is not now, nor has there ever been an ALFA 8187B adapter.

And neither has anyone made one on their behalf. It simply doesn't exist.

8187B is a weaker variant of the 8187L chip and wouldn't even belong in an ALFA adapter commanding the same price tag, simple as.

Patently obvious your operating nothing more than a fake. Holograms can be "faked", that's why I derived the unmistakeable ALFA "smudge test".

Because clones like to be perfect replicas in every way including hologram, so perfect in fact they forgot to slip with the soldering iron.

Hell, even the MAC address of an ALFA can be cloned away with my uploaded Mass Production Kit.

 

^^^ Maybe that's the reason he didn't reply back. He realized he has a fake ALFA 036H

 

Stronger than a real ALFA ? no. Not even close. The chipset isn't the only factor here at play. I've tested all this kind of stuff, and I'd be the first on here to extol the virtues of an 8187b if it were better...that hasn't happened.

Nevertheless.... (including the ramblings of articles trying to justify the misguided purchase of cheaper options which only concentrate on chipset matters, whilst completely forgetting about the piss poor PA attached and included inside which is just as relevant as the chipset)

You simply cannot have the same range, sensitivity, compatibility, or even the same output in OFDM power using an 8187b equipped device over an 8187L device pound for pound despite what you may have read.

eg. In an environment where the 8187b would detect around 9 AP's, the 8187L will detect comparably 16 instead. (dbi antenna's and all else being equal)

It's simply just not possible to have an 8187b behave at the same level of quality/performance as an 8187L chipped adapter does,
due to the 8187b being an a) less expensive realtek alternative (for a reason), and b) (less sensitive) variant of the chipset for the reasons of a) which explains the differences. In addition to c) high confidence of an utter crap power amplifer implemented inside a clone.

If it works flawlessly for you for 3 years for what you want it to do then good. Whatever rocks your boat, and keeps you happy. You just might be missing out is all I'm saying. And here's why I offer the truth and nothing more...

The lesser chipset is only half the problem. The first thing I would be looking at is the quality of the PA inside because this is what determines
the final output power. But being a fake I think you'll discover that it's pretty generic chicken brand stuff what all the others floating around use, my Kasens G9000 included.

As you see in the picture I provided (above) the real ALFA 036H v5 has actually not 1 but 3 amps. The Skyworks 2576L alone being rated for 26dbm operation.

Just one of the reasons it costs more money over the fake/knockoff clones jumping on the bandwagon amongst others that you might get 15dbm max out of (if your lucky) without popping the PA. The Chinese PA's are like little bombs waiting to go off..

So the facts on chipset difference and PA differences in quality are facts, so it's entirely incorrect to assert painting a picture of (or assume) that an 8187b chipped device equals, or alternatively even surpasses as a ridiculous assumption an 8187L (not least an ALFA variant of 8187L) in performance, it doesn't, because the truth is it's just not possible nor realistic given the comparable chipset specifications, and to add on top of that, the differences in the quality of power amps inside, since the chipset alone whilst (less sensitive) only tells one half of the story as to what kind of overall quality your going to be getting in the finished article of the ALFA clone pidgeons the world and his brother are making, and chucking out at all the unsuspecting folks that simply don't know any better.

*If* your AP's are close enough, you might not be able to entirely fathom the difference between a real ALFA 036H 8187L, and generic chicken brand adapter x or y, so yeah potentially you could save some money the closer your AP is.

That is until you start trying to hit things at range and broaden your horizons pushing the edge of what's possible, where the 8187b with it's likely inferior razor thin profit margin PA will just quickly fade away, and won't be able to keep up with the same sensitivity or capacity, compared to an 8187L adjoined with a quality PA for the longer haul.

With those kind of environments, the connection speed on the clones will sink like a stone, before fading into to nothing after some minutes at any significant measurable distance.

There is a simple reason why fakes will always continue to cost less money, and ALFA continues to charge a premium. Because the chipset itself is only 1 little part of the jigsaw.

 

Firstly the person who wrote the article is statistically unlikely to be a homosexual.

Secondly, Well......errr infact I do have measurements, It's not just any old drunken monkey opinion, it's called specifications, and facts and backed up by evidence (see below)

The specification of 8187B is significantly less than an 8187L. A proven fact most will agree with beyond doubt.

Anyone who believes it to be anything else on an equal footing or in any way superior to an 8187L are just deluding themselves when you peel back the layers.

I tested all 4 Alfa adapters and wrote one huge article about them on here since I own them all with no particular favoritism for any. An unbiased opinion.

Maybe you might respect Roklands opinion more than that single article or my take on it perhaps ?

(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.)

Ok...here is not an "opinion" but example specifications and testing of 8187B using Roklands testing of an 8187b adapter vs 036H 8187L
since you evidently pushed me into it, I'll give you an "independant opinion" to peruse over for future reference, that isn't my own "opinions" about the 8187b devices. There...

PDF:

http://www.rokland.com/support/awus036h_gsky.pdf

Further analysis:

RokBlog: Alfa AWUS036H vs. Gsky 500mw USB adapter

The GSky is an 8187b chipped adapter btw.

Quoted from article :

Alfa 036H 8187L

11 mbps -91 dBm at 8% packet size
54 mbps -76 dBm at 10%

Here are the figures for the 8187b adapter :

11 mbps -82 dBm at 8%
54 mbps -65 dBm at 10%

"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."

Which correlates with my own testing of borrowed 8187b adapters that all have junk PA's. Case dismissed.

In any event, It is a completely different chipset with a different driver package.

Since you asked about an NHA, an 036NHA for the future is a smart all round choice, and the fastest of all the Alfa adapters, since more processing is done in hardware instead of software based along with sensitivity close to -96dbm. Even the 036H itself is now getting far too old, no new drivers, and not really any good for decent multimedia streaming or extended duty cycles that the 11n adapters offer. Stay away from the NHR. It's crap, although I have discovered a way to con it into giving me a real unfaked 24dbm in Windows whatever the domain. Hehe.

The connections on NHA are very stable, even on only 1 bar of signal. Ever so slightly less AP's discovered than 036H or 036NH. You won't be disappointed with the speed however at least.

Hint for 036NHA: use the Windows XP driver for maximum range of which I posted latest drivers for in another thread...

The Win7 ones while works didn't have as much development, and don't get as much range as the XP ones. The XP x86 ones work forced on Win7 x86.

Bye Bye....

 

The maximum is (or should be) 17dbm for the 8187b/l. It's as far as the slider will allow you to go (unfortunately) you get higher limits to play with on the 8188RU chipped NHR...

You might be able to manually set a power over 17dbm ignoring the slider, however I've not noticed much if any difference past this figure, in any case, those are the EEPROM limits when edited with MPK at least for the 8187L.

Whereas it's 24dbm for the 8188RU NHR if you know which drivers to use
and what your doing. Well it actually goes up to 100dbm, but jumps back to 24dbm (around 251mw). Fed by x2 USB cables.

I'll see if I can rig something specific up for the 8187b in next days or so and post a link to some custom drivers made for exactly this purpose, whether installed on XP, Vista or 7.

By the time you get home the custom 8187b drivers will be ready.

EDIT

Ok, here is my first custom driver attempt made up for the 8187b *only*, which should allow CCX control in XP or 7 amongst other unlocks.

You'll have to have a play around with it and see if it all behaves the same as RTL8187L.

RTL8187B_LatestDRV_Customised_CCX_UI_1.00.0145.L

Don't worry about the lower UI version (it works better than 0.179 anyway and allows RTSET without breaking) all the drivers inside are still up to date and current taken from .0179 package.

Report bk if they work well or not...Should install CCX tab or CCX_IHV depending on OS installed to. If the contents of the CCX_IHV folder doesn't get installed automatically to \System32 just copy them there manually and just create shortcut on desktop to RTLUI2.exe. RTLUI2 will actually run on XP but it doesn't work, that's why you have the CCX tab on XP.

They should allow you to do what you are trying to do (ccx tab in XP or RTLUI2 in system32 for Win7/Vista)

(I don't have access to 8187b h/w anymore to test, but I know enough about 8187L and how the driver components all interoperate and how to jury rig one up.)

PS.

As you can probably imagine, 17dbm of forced measurable output in the realtek control panel, coupled with a 16dbi YAGI antenna is around 33 of EIRP. RTL8188RU 24dbm+16dbi more so.

A pretty powerful cocktail of hacked up drivers and sniper rifle antenna's to get noticed, not least by your neighbours wifi being persistently jammed and disrupted if your antenna is pointing at them, so be careful therefore what you do with the new "unlocks" to set the power as "high" as it will go or you may be unwittingly annoying someone, especially in an urban or city environment. The power is artificially low for a reason of being a "good" neighbour hehe.

Have fun...

Example of 036NHR v2 (RTL8188RU) unlocked power settings using my own homebrew custom drivers instead of the crap ones the adapter comes with.

24dbm is maximum output from a UMTS/3G mobile phone (Power class 3 mobiles) . SiGe Skyworks 2576L on V2 NHR is rated for 26dbm+

There is little point raising the power to these kind of levels unless you do it on *both* ends. No point having a big mouth and small ears. Only the Realtek 8188RU adapters offer any kind of hidden power controls to smash past the 17dbm ETSI limits as long as you know what your doing.

 

The 036NH is the least problematic in any Windows OS and has lots of signal "bars". But the NHA is one of, if not the fastest ALFA adapter as most of the host processing is in hardware instead of software (driver). It just disconnects itself once every few days requiring a reset which is probably a driver issue that will never be fixed. And is the only adapter that supports all 6 wireless modes under Linux.

036NHA can work even with 1 bar signals fully. E.g. signal strength 10% signal quality 90%.

The NHR is a horrible adapter that is only really useful in Windows, and even then is seriously bad compared to the others at even detecting networks. Maybe it's just mine, but I don't plan on buying another just to "find out". My next adapter will be 051NH (when the ridiculous price comes down) so I can have 5ghz covered.

Which adapter to choose depends on what your needs are. Or you could just eventually end up buying them all so you have multiple choices like me because no 2 act the same.

You won't be able to hack and crank out 24dbm TX with 8187b, unless your running like me an 036NHR with twin USB leads and my own "doctored" drivers that I haven't shared anywhere yet.

I suspect you won't (like 8187L) get past 17dbm on 8187b (eeprom hard limits), but I guess you will just have to see if it "unlocks" since I can't test an 8187b.

Disclaimer: If you pop the (don't know what) PA inside "playing" with the power settings and/or due to a bad antenna don't blame me.

Since.....the higher the power, the better the antenna tolerance must be in tune and I don't know your setup.

Because SWR rises with power output. And high power really only makes sense if you do the same at the other end since wifi is a 2 way street, needing x2 sets of RX and TX.

No point "shouting" a strong signal if they can't shout back because while increasing the TX range, you won't be able to hear "them" because they can't shout back the same on bog standard 3dbi AP antennas. Just because you run high power doesn't mean that everyone else is/will which it needs to work both ways, and therein lies the problem of wifi range limitations.

Everybody has to own high power/doctored adapters, or all bets are off.

A discovered signal is one thing, connecting to it is another.

Power is not the be and end all of assessing adapter quality. It makes sense to communicate on as little power as you can get away with for interference reasons.
And health reasons, since were dealing with MICROWAVE frequencies, you know the same ones that can cook bacon.

Far too many watts along with a 24db parabolic grid is seriously not a good idea, unless you plan on building nothing more than a testicle zapper ray gun.

Useful though for x2 ALFA's "doctored" and distanced apart on an adhoc network where then both of them can "shout" all they like at each other.

Mass Production Kit Editing of Kasens G9000 RTL8187L (pam unknown)



XP install MPK, Load EEPROM, save EEPROM to MAP file for backup as 8187.MAP in Realtek program files folder and copy elswhere in case you mess up. First F (15 hex) is for OFDM power, second F (15 hex) is for CCK power. Apply FF power to all channels. Channel Plan changed to MKK1 (1-14) for Linuxy purposes. Increase MAC address unticked. TXPowerbase 0:0 for Skyworks PAM's or will disrupt network scan better 0:0 to be safe. Write Changes when ready. Done. Uninstall MPK and re-install driver.

Oh and don't fiddle with changing RF Chip vendor. While it's still possible to retrieve a bricked adapter, it's a pain in the ass to do so.



Now both 15dbm on boot in both 11g and 11b modes. Then take power slider raised to 100% in driver using edited EEPROM for 17dbm in both OFDM and CCK modes (50.1mw). All Done.

Test differences before and after editing using inductance feedback test on nearby sound speakers next to omni antenna under heavy network load like a speedtest. If the level of inductance increases, so has your (edited) power output which has now new defined higher limits accordingly. If you don't like the new "high" limits of power output use normal power slider to tone back down to normal levels previously without having to re-flash old eeprom .map file.

*Your final result will be limited to the quality of PA inside adapter. There may or may not be an improvement. For WifiSky ones, don't even bother.

*Actually on 8187L at least it's possible to manually set 30dbm and forget all about the slider.

Just that I haven't noticed any difference in ROM edits over 17dbm that's all or no benefits like the NHR's 24dbm limit, and the 8187 eeprom limits in MPK tell me 30dbm is just not possible (even manually set).

But remember that 3x the power over 5dbm doesn't mean 3x the distance !
And don't make the mistake of equalling signal strength with signal quality they're not the same !

It's possible to have higher penetration with more input power, but sometimes you have a more stable connection with lower power, so it's a double sided coin.

Before tinkering with any of the power tweaks above. Make sure your plugging it all through a powered USB hub. It probably doesn't need one or doesn't matter, but it's better than risking blowing motherboard USB polyfuses if I'm wrong.

*Infact it's essential you do actually need a powered hub contrary to others tests especially if you plan on moving your newly "juiced" adapter beyond the cable it comes with.

I tested these high power levels on my USB active repeater cable far away from the desktop machine 10m away in another room, and the power level at least in the 8187L's case scales back to 5dbm protecting itself against undervoltage and choking the adapter's scan of available AP's. But on a powered USB hub is stable and not a problem and stays at 15dbm all the time in CCK and OFDM modes.

I suspect the voltage drop on the 10m active USB extension is too much for the 8187L's newly flashed power requirements...This can be gotten around by plugging a powered usb hub fed by a wall outlet into the end of the active USB repeater cable (and not the other way round) it seems my cable doesn't like a hub first scenario but works the other way with hub last on the chain (active repeater cable was cheap and nasty, but voltage drop can be expected with all extensions and some devices don't like it, especially high power wifi adapters or HD's).

Let me know how your experiments with 8187b go...

 

I just de-cloaked everything while I was in there making them for you, same as 8187L that's all, nothing to worry over.

Thought you might find the extras useful for later...

Nothing happens because it's TX (output) only power. It won't change signal link quality or receive parameters of the adapter, they are fixed in the sensitivity of the chip and not tied to dbm/mw output.

The only way to test for this difference in values of output power without having to buy too expensive detection equipment, is a difference in better lower RSSI values testing @ AP end (to confirm stronger signal output power), or by stronger audible inductance into a sound speaker placed next to the Omni as you input different power levels higher the stronger the inductance fed into the speakers, like a mobile phone when it's about to go off next to speakers.

My 8191SU dongle only goes up to 16 (dbm). All realtek chips are different in what they show in RTLUI2.exe, but sometimes the value can be bypassed (manually) with a measurable difference, some maybe not (A lot depends on strength and quality of PAM chips (amplifer), and what output parameters this chip can tolerate and is specced for, before going non-linear, and generating noise in the output which is counter productive.)

So what your saying is, if you input more than 10 manually as a value the display jumps back to 10 and shows no more than 10dbm ( 10mw ) of output or shows more ?

If you can't push anything more than 10dbm, then I would have to assume the PA inside is just not capable of giving any more in it's output than this, unlike the 8187L's 8191's or 8188RU's I have tested an increase in output power with.

I'm afraid if this is the case, then ultimately this must rest at the cost of unsuspectingly being sold fakes/clones.

There is no extra room to experiment or play with in amplification stage, because with clones, corners are unfortunately cut.

And as long as it connects a reasonable distance away for 90% of buyers, most people wouldn't notice the difference in different output levels.