Windows New ALFA 036H Modded Drivers + Homebrewed & Improved by Me

Yes they left quite a lot out...such as forcing wireless modes and fixed rates.

All the functions were always there in the drivers all along though.

They just forgot to add them. I just turned them all on and cleaned up the whole inconsistent install mess across all the OS versions.

Couldn't stand the shoddy half wit implementation as it stood, the missing stuff, and the stuff that didn't work across OS

Many weeks of trial and error on a dual boot, and poking around many an old .inf.

I'm doing the same atm for AR9271 and Ralink 3070 drivers to give us more options, but they'll be ready when they're ready.

Infact the b, b/g, b/g/n wireless mode selection ability on Atheros and Ralink is already figured out and done (and works)

But there's like 8 different OS in x86 and x64 to think about for Atheros and Ralink totalling 16 different alterations to think about testing them, and just one man working.

Don't just wanna throw stuff out there if it isn't tested. RTL8188RU can wait...drivers are a complete headache for that one.

But I'm quite sure i'll be able to implement some kind of control for the 8188RU to kill the bright blue led from flashing and creating police sirens on the ceiling at night for when your running a black wallpaper and want to go into full stealth mode without having to use tape or bluetack.

--- Double Post Merged, 7 Feb 2014 ---

I today found an anomaly with the ALFA 036H Windows 7 1.316 version of the driver.

I have long suspected it to be flaky and the root cause for the random disconnections, at least especially so on Windows 7 x64.

If you look at the statistics column in the GUI, for the 1.1316 Windows 7 version of the driver the RXRetry count is unusually abnormally low.

So low in fact to be downright wrong and the driver isn't responding the same compared to all the others.

The driver is bugged because no other driver gives these unusual results.

When you load the Vista64 1.1313 or 1.1314 tx control version of the driver up on Windows 7 you will get a completely different answer
and in fact a normal amount of RX retries inline with what a 8188RU NHR gives with any flavour of OS driver, or using the 036H driver on Vista
or indeed what Windows XP gives using either 1.313 or 1.314 or any below that for that matter.

Could this be the cause of all the disconnection troubles under Win 7/8 ? Is the Windows 7 version of the driver actually completely flaky and broken
and mishandling RX retries eventually causing the adapter to disconnect under Windows 7 ?

I would tend to think so based on this discovery, so I would therefore recommend switching to the Vista version of the driver 1.1314 if you want p/c
and 1.313 if you don't, and going no higher than that and staying away from the Win7 specific version of the driver altogether if you suffer from
unexplained disconnects.

I can't prove this is the cause of any disconnection under Win7 because other factors may be at work, but looking over the anomaly of how this driver responds
compared to all the others seems it might be highly likely to be possibly bugged in it's handling of RX retries with packets going into a black hole.

 

err, no. 036H only as per the title says.

The 036NH uses a RaLink chip instead of Realtek so they aren't even the same chip or s/w, although I'm still working on unlocking the RaLink driver for the 036NH's people...and the Atheros ppl, and the (god help them) 8188RU NHR v2 ppl.

(yes the blue light can now be shut off on the NHR without having to use sticky tape & bluetack ! )

I posted the most up to date drivers in all the Windows chipset threads...take a look for the 3070L. As for custom ones for the NH, just wait some more time and check back. Wireless mode selection is already done on Atheros and Ralink to kick them both down into 11b mode to extend the range and increase reception sensitivity viability, but I'm still busy researching more options that are useful to unlock on 4 different adapters across 6 different OS versions, so is going to take quite a while...

 

I haven't come across any tx power adjusting drivers in my travels except old ones where the GUI allows you to select from 100% tx downwards on profiles.

TXpower command on Linux also has no effect for Ralink 3070L.

It may show set at 30dbm, and respond to the commands correctly, and even display the new value of 30dbm, but the power is never actually changed for real unlike RTL8187L.

Unless you hack the CRDA...

 

It's annoying because the 036H and 036NHA are the only 2 adapters to suffer from random dropouts from the drivers having a hissy fit. 036H more annoyingly so than 036NHA which is around once a day.

I saw that the Windows 7 version of the 036H driver is the only driver for the 036H to get the RXretry count badly wrong where every other version of
the 036H drivers gets it right, and <= 1.1314 don't seem to disconnect either. I've also ruled out power and heat.

Interestingly my Kasens G9000 RTL8187L doesn't drop like the ALFA on those 1.316 drivers even though the rxretries are still wrong, but then measuring the intensity level of the amount of inductance fed into the PC speakers using the 5dbi antenna on both ALFA and Kasens while running a speed test with the adapters in the same position, tells me that the Kasens G9000 has no where near or close the same level of output power as the ALFA 036H does despite behaving the same for most purposes. What I can't explain is the Kasens G9000 seems to be around 10% more sensitive than the ALFA when it comes to RSSI results. The Kasens can't always use the same amount of drivers the ALFA can either. If you go back too far in history, the Kasens isn't recognised by the drivers below a certain version.

The amount of inductance in the speakers (and therefore output power) goes in this order for me :

036NHR
036NH
036H
Kasens G9000 RTL8187L
036NHA (almost quiet)

Which seems inline with the power output of those respective adapters.

But the Kasens is as light as a feather, and it's power amp sucks compared, same as all the other chicken brand 036H ALFA knockoffs floating around.
But on the other hand they don't disconnect. Maybe they turned up the power a little too much to remain stable on the 8187L ALFA, I don't know.

--- Double Post Merged, 8 Feb 2014 ---

It appears so. Yes.

 

Probably. There is nothing much more to do anyway...

But in hindsight I would recommend using the Vista x86/x64 driver files on Win7 instead,
as the only one version of native Win7 ones are broken, report wrong RX retries data
counts in the realtek GUI, and are well known to frequently "drop" for no apparent reason.

The vista ones of x86 or x64 flavours loaded on Win7 do not suffer the same flaws but you can
still have power control enabled as long as you take x.1314 version and not x.1313 or below.

Better to forget about power control drivers altogether and flash in permanent new power values to the 8187L
EEPROM using Mass Production Kit unlocking then the p/c CCX slider out to 17dbm, then you can use the normal
non p/c 1.1313's without worrying about having to fiddle with the power drivers anymore which would never achieve
the full 17dbm anyway but 0-9dbm for OFDM and 0-11dbm for CCK using the p/c power slider unless you were to force 17dbm
directly using the CCX tab setting, while doing it this way "works" this is then lost on reboot.

So...instead I have 15-17dbm power settings burned permanently on my Kasens G9000's eeprom for OFDM and CCK for channels 1-14, so I no longer need to be using the p/c drivers anymore at all.

The normal non p/c ones do just as well with a flashed EEPROM with jacked up power values.

(17dbm is TX output max limit under ETSI for point to point links anyway)

 

I'm still working on putting something out for the 036NH ralinks. I haven't forgotten.

I think the 036NH has the most solid hassle free drivers of all the ALFA adapters although still quite bare in any sort of control.

It's finding the time to research and test changes across 4 adapters and several OS versions.

Not every "unlocked" feature is available in every driver depending on OS.

Vista/7 drivers are very different structured in most cases to XP ones.

 

The features are all ready there and available waiting to be unlocked in the realtek .sys files scanning
with a hex editor/binary.

Then search the t'internet to find out what that feature does and how it's implemented correctly in the .inf
(very time consuming to research and test everything)

Windows 8 is problematic for 8187L, since

a) I prefer to avoid it as an OS altogether for various reasons. So I can't or won't test any 8187 Win8 solution.
b) you can't modify any .inf files on x64 Win8 due to new patchguard implementation unlike vista or 7
which requires the .inf is also "signed".
c) there are no existing native RTL8187L Windows 8 drivers. Driver support ended in 2011.

 

My v2/nhr is something of a complete joke slowly driving me insane since the time that I've had it.

I think I'm beginning to become aware that I have been conned by the hype.

It's a Turkey, A bad apple. It's garbage compared to the other 3. I'm beginning to seriously hate it with a passion...Why ?

The signal strength with the latest drivers is half of what it should be, and I still can't hit anything at range using *any* of the new 8188RU drivers unlike the other 3 adapters can.

And if I use the "old" 1021.1 NHR drivers, then I can hit things at range normally (like the other 3 Alfa's can), and the signal strength is shown acceptably correct (although still magnitudes down compared to 036H) 74% vs 92% with 036H but then I get "random" disconnections and unexplained random TX errors. And trying the power control drivers blue screen the damn machine useless as they are.

The receiver is messed and It's crap at detecting networks compared to the consistency of the others which is another random feature depending on which way the wind is blowing.

And you can no longer buy a new NHR v1 anywhere that I found. When I hear about people getting 100 AP's listed with an NHR or it being a "signal sucker" I'm completely puzzled and baffled as to "why" or "how" this is possible, because this is not my experiences. Currently my V2 NHR does not detect any more AP's than the others do.

Since the v2 is supposed to be an improvement/tweak over the already existing v1, I don't understand what makes the v1 so "good" compared to it. The situation is a farce...

Can you upload some pics of what the NHR v1 and v2 is doing with it's signal strength levels in Realtek UI using old and new drivers ? And how it behaves thanks. I need some visual "proof".

I need to know if it's the same for everybody's NHR with the bonkers drivers, or I have been sold a lemon because at the moment, I don't believe it's the strongest adapter on the market by a long stick, because it can't even connect to what the others can. And infact with the default 5dbi it's the weakest of the bunch. Sometimes I have to use a stronger antenna to even get a connection, not so on the others which will merrily work all day off 5dbi around the same approx. range.

And it's been like this crap "playing up" since day 1 with my NHR from an "official" Alfa distributor in an un-tampered box.

I don't like how the drivers behave with the clutzed signal strength and TX RX rates of 24/24 rising on activity. And if it's been this way for 3 years as standard as normal, I don't have much hope of this ever being fixed, and nor will I be buying yet another problematic "broken" NHR.

I want to like the NHR, and for it to work, but I'm not buying into the hype. Even my crappy 8191SU rock bottom of the barrel cheap dongle at least shows the signal strength and bars properly.

Soon as I have a handle on what is really going on, I'll be emailing Realtek and/or Alfa to fix this complete mess because either the hardware is fundamentally broke, or the drivers are.

This is supposed to be a premium high power chip/adapter for pity sakes. For me at least atm (don't know about anybody else) it completely sucks "in how it behaves", rather than "sucking in" signals.