WHR-HP-G54 Flag 0x3758 instead of 0x2758/0x1758 ???

Post new topic   Reply to topic    DD-WRT Forum Forum Index -> Broadcom SoC based Hardware
Goto page Previous  1, 2, 3, 4, 5, 6, 7  Next
Author Message
GeeTek
DD-WRT Guru


Joined: 06 Jun 2006
Posts: 3768
Location: I'm the one on the plate.

PostPosted: Wed Oct 10, 2007 23:17    Post subject: Reply with quote
Nice looking tests ! That is quite a machine you have there. I think I would trust your dedicated "G" mode test for determining the power from a legal standpoint (as in determining legal MAX ERP for FCC laws), but I would still consider the peak power to be the actual output. Either way, you have accomplished what a lot of people have wished to see, a precise and accurate power measurement of these radios. I'm not sure I understand your reference of 30 dBm and the attenuator value of 40 dB, but I would be real quick to beg ignorance and punt the ball ! It has been a long time since I have used a rectum spanalyser, and they were nothing near the quality and sophistication of your machine. Thanks for the screenies !

Edit - You may want to post your shipping address. There will probably be a few people who would want to donate various radios for you to test ! Laughing

_________________
http://69.175.13.131:8015 Streaming Week-End Disco. Station Ripper V 1.1 will do.
Sponsor
dicksons
DD-WRT Guru


Joined: 07 Jun 2006
Posts: 980
Location: Coal Creek Canyon, Colorado

PostPosted: Thu Oct 11, 2007 18:55    Post subject: Reply with quote
tparris, You're god-like !

You've provided more factual data on this subject in a few posts than the rest of the forum combined in the past year+


PS. Curious - what kinds of doings are you into up there with these 25 Km links ?

_________________
linksys GSv2, Gv4, Gv2, GLv1, G-TM, Buffalo wbr2, whr, whr-hp, whr-g125, wli-tx4-g54hp, Moto wr850gp, Alix.3C2
tparris
DD-WRT Novice


Joined: 08 Oct 2007
Posts: 12

PostPosted: Thu Oct 11, 2007 19:46    Post subject: Reply with quote
Thanks dicksons,

I am just fortunate to have some experience in RF design, and have access to some decent test equipment. I hope I can contribute a just little bit to the dd-wrt community. I really do like what is being done here. I would be more than happy to test other hardware if any one desires. You can e-mail me, and we can hash out details. I would also be happy to run some more thourough tests on the WHR-HP-G54s if there is a desire in the community.

We need the long links for communication with remote instrument sites. I work with/design radar systems (for upper atmosheric research... academic... not commercial), and we have radars (and other instruments) in some remote places (often close enough to some sort of infrastructure to use radio data links). To operate the instruments and retrieve data, we often use radio links. We have used Cisco links (specifically desinged for long distance applications) in the past, but they are flakey. When one of our cisco links on Kodiak Island (it's a 25 km link) died last year, I found info about some people in Peru that made a 210+km link using WRT54Gs and some high gain antennas. So, after a little more searching, I found this DD-WRT group, loaded DD-WRT onto a couple of linksys routers, and we've had a stable link ever since. The WHR-HP-G54 that I tested in the previous couple of days is being installed today at one of our sites in King Salmon, AK.
tkoyn
DD-WRT User


Joined: 11 Feb 2007
Posts: 426

PostPosted: Fri Oct 12, 2007 14:25    Post subject: Reply with quote
tparris wrote:
I measured the output of a WHR-HP-G54 with an Agilent E4404B spectrum analyzer. I tried this with DD-WRT V23, sp2 and V24, RC3. There was no difference. Here are the results;

0mW setting = 0dBm output (1.0mW)
1mW setting = 0.1dBm output (1.0mW)
2mW setting = 4.1dBm (2.6mW)
4mW setting = 7.5dBm (5.6mW)
8mW setting = 11dBm (12.6mW)
16mW setting = 13.7dBm (23.4mW)
32mWsetting = 17.0dBm (50.1mW)
64mW setting = 19.5dBm (89.1mW)
128mW setting = 20.1dBm (102mW)
251mW setting = 20.0dBm (100mW)

And for the power settings you asked me to measure;
28mW setting = 16.5dBm (44.7mW)
50mW setting = 19.5dBm (89.1mW)
80mW setting = 19.7dBm (93.3mW)
Thanks for your outstanding contribution to the DD-WRT community!

Can you please post a complete table like this for the test done in the manner to replicate the FCC (peak) figures? Also, can you do the test both this original way and again consistent with the peak method for the stock firmware g-only (25%, 50%, 75%, 100% power) and for V23 SP1 (settings you did for dd-wrt)? And do the tests again in b only mode to see the effect of b versus g on the measurements considering the different encoding methods used (both in dd-wrt, and stock firmware if it allows b only). Finally it would be good to get all the figures into the wiki in a nice table. Also can you include the setting 100mw in the dd-wrt tests in addition to the ones you have in your list?

If you manage to disable the transmit power limiter by altering the hardware, a complete set of tests for the altered version of a WHR-G54-HP would be useful to see the impact of the alteration in dd-wrt and stock firmware.

Please let us know if you find out anything more about the effect of the boardflags, as I am confused on those.

I would also be interested in seeing test results for WHR-G54S and WHR-G125 stock and DD-WRT.

Thank you so much!


Last edited by tkoyn on Sat Oct 13, 2007 2:50; edited 1 time in total
tparris
DD-WRT Novice


Joined: 08 Oct 2007
Posts: 12

PostPosted: Fri Oct 12, 2007 17:27    Post subject: Reply with quote
tkoyn,

I will certianly test the WHR-HP-G54 as you have requested. You'll have to have a little patience though. My WHR-HP-G54s got installed yesturday, so I cannot make the measurements until I get more. I have some on order, and they should arrive next week. I will make the measurments, take screen shots, and post the results as soon as I can after they arrive.

I do not have a WHR-G54S or a WHR-G125, nor do I have plans to get them (I just get wireless routers when I need to use them in a remote location, and I probably won't get more any time soon after these WHR-HP-G54s arrive next week).

I do have a couple of WRT54Gs left from another project, and I will test those more thoroughly.
tkoyn
DD-WRT User


Joined: 11 Feb 2007
Posts: 426

PostPosted: Mon Oct 22, 2007 13:17    Post subject: Reply with quote
tparris wrote:
I will certianly test the WHR-HP-G54 as you have requested. You'll have to have a little patience though. My WHR-HP-G54s got installed yesturday, so I cannot make the measurements until I get more. I have some on order, and they should arrive next week. I will make the measurments, take screen shots, and post the results as soon as I can after they arrive.

I do not have a WHR-G54S or a WHR-G125, nor do I have plans to get them (I just get wireless routers when I need to use them in a remote location, and I probably won't get more any time soon after these WHR-HP-G54s arrive next week).

I do have a couple of WRT54Gs left from another project, and I will test those more thoroughly.
Thank you in advance. Did you recieve your new WHR-HP-G54 routers yet?
tparris
DD-WRT Novice


Joined: 08 Oct 2007
Posts: 12

PostPosted: Thu Oct 25, 2007 2:45    Post subject: Reply with quote
Here are some of the requested measurements: http://www.dd-wrt.com/phpBB2/viewtopic.php?t=22201
I put them in a new thread because they are not really related to the orginial topic of this thread.

I measured 802.11b also, but there was almost no difference between G-only and B-only modes for the stock Buffalo firmware, DD-WRT, and Tomato. This concerns me a lot. Either changing settings in the firmwares does not change things as they should, or the units still transmit 802.11g beacon signals even if set to 802.11b only modes (the spectrum of sich a signal could 'hide' the spectrum of the 802.11b signals).

I can post photos of the 802.11b measurments if you desire, but they are the same as 802.11g.
sdoboze
DD-WRT User


Joined: 25 Jun 2006
Posts: 72

PostPosted: Mon Oct 29, 2007 20:21    Post subject: boardflags 0x3578 Reply with quote
Hi

I run my WHR-HP-G54 in client mode.
I was getting ~ (approx.) -80 noise levels.
Then I used the commands:
nvram set boardflags=0x3758
nvram commit
reboot

and my Noise level went to ~ -90, 10db better.
I think this should be changed in the dd-wrt firmware, when auto-detecting this model of router.
See the bugtracker entry here: http://www.dd-wrt.com/dd-wrtv2/bugtracker/view.php?id=2776

Regards

edit: put the corrected 'Before' picture up, as the first copy looked misleading (from before the connection stabilized).
edit: inverted the 7 and 5 in the original post: corrected.
edit: I've updated the wiki with the info



before.jpg
 Description:
Before
 Filesize:  43.73 KB
 Viewed:  15316 Time(s)

before.jpg



after.jpg
 Description:
After
 Filesize:  43.26 KB
 Viewed:  15327 Time(s)

after.jpg


dicksons
DD-WRT Guru


Joined: 07 Jun 2006
Posts: 980
Location: Coal Creek Canyon, Colorado

PostPosted: Tue Oct 30, 2007 23:03    Post subject: Reply with quote
back to the original topic and to confirm what sdoboze says,

0x3 makes my noise go down by 10 also on v24 rc4 as measured in the wireless status as well as site survey

would this flag also be valid in v23SP2, I wonder..... I'm all out of sp2 HPs at this point or I'd test it myself

if someone's still running SP2 could you try that 0x3 flag and do some wireless statuses and site surveys before and after ?

thanks,

_________________
linksys GSv2, Gv4, Gv2, GLv1, G-TM, Buffalo wbr2, whr, whr-hp, whr-g125, wli-tx4-g54hp, Moto wr850gp, Alix.3C2
Spudz76
DD-WRT Novice


Joined: 04 Oct 2007
Posts: 3

PostPosted: Thu Nov 01, 2007 19:30    Post subject: Reply with quote
Just tossing my hat in the ring, on a HP-to-HP link of only 180 meters with two identical 21 dBi grid dishes horizontally polarized, the 0x3758 boardflags vs stock 0x2758 also dropped my noise floor by exactly 10 dBm to -92 dBm. I am using txpower of 60 in Tomato 1.10 now, seems best and confirms results as shown by most testing over the lifetime of this thread.

What sort of RSSI and quality do you guys get? I'm looking for a sort of maximum I can expect, because my quality never goes much above 50. And is that enough for 54Mbits? I really only seem to get about 1MB/sec file transfer rates which is like 8Mbit link speed average. It seems with my overkill dish gain and supposed 3/4 watt output I should see better than a 50 quality, and get full speeds. Am I just overpowering everything?
dicksons
DD-WRT Guru


Joined: 07 Jun 2006
Posts: 980
Location: Coal Creek Canyon, Colorado

PostPosted: Thu Nov 01, 2007 19:43    Post subject: Reply with quote
I confirmed on a v23 sp2 buffalo running in the wild

this is a client running a couple miles from one of our WISP APs

with 0x2758

wireless status

rssi -53 *noise -63 snr 15* quality 50%

site survey

rssi -53 noise -71

AP sees buffalo's SNR as 26

------------------------------------------------------------------

with 0x3758

wireless status

rssi -54 *noise -81 snr 27* quality 50%

site survey

rssi -53 noise -81

AP still sees buffalo's SNR as 26

_________________
linksys GSv2, Gv4, Gv2, GLv1, G-TM, Buffalo wbr2, whr, whr-hp, whr-g125, wli-tx4-g54hp, Moto wr850gp, Alix.3C2
redhawk0
DD-WRT Guru


Joined: 04 Jan 2007
Posts: 11476
Location: Wherever the wind blows- North America

PostPosted: Thu Nov 01, 2007 23:41    Post subject: Reply with quote
I have a pre-RC5 release of Std-Nokaid installed on my HP-G54 due to some WAP testing I was developing with Eko.

The noise went from -78db to -89db.

It looks like setting the boardflags to 0x3758 is a winner in Repeater Mode.

Or....is it just an illusion???? Wink

redhawk

_________________
I do NOT provide personal assistance through chat or phone....so please don't ask.
MichelG
DD-WRT Guru


Joined: 12 Feb 2007
Posts: 903
Location: Galati - RO / Kishinev - MD / Geneva - CH / Tarragona - E

PostPosted: Thu Nov 01, 2007 23:49    Post subject: RC-4 Std Client Bridge Reply with quote
Also tested and the WHR-HP-G54 got 10dB better. Noise is now around -94 / -96dB. Before it was fulctuating between -82 / -87dB.

Seems to also working here Very Happy

Michel

_________________
2x EOC-2160 - AP 2x14dB Yagy to 450m link
2x EOC-2160 - RP 2x17dB Panels - 2x 7Mbps
1x WRT54G v2.2 - V24-SP2 13577VPN Eko
All Senao EOC-2610 V1.10 Std 15362 BS
1x UBNT Bullet 2 - RP 18dB Panel - 11Mbps - 13Km link
Kingdomcome
DD-WRT User


Joined: 24 Dec 2006
Posts: 134

PostPosted: Sun Nov 04, 2007 23:58    Post subject: Reply with quote
I tried changing to 3758 for a couple days, but did not see any increased performance. Noise did decrease by 10 db, but I also noticed a signal decrease of 10 db resulting in no change of the SNR, which I believe is the tell-tale number here. I will also say that observed stability of the wireless network was not as solid, though almost neglegible, as with 2758. I would hope with this new partnership and Buffalo's "commitment to the open source community" that the confusion with the boardflag will be settled. BS himself has questioned why either of these settings work....
_________________


Last edited by Kingdomcome on Mon Nov 05, 2007 18:42; edited 1 time in total
dicksons
DD-WRT Guru


Joined: 07 Jun 2006
Posts: 980
Location: Coal Creek Canyon, Colorado

PostPosted: Mon Nov 05, 2007 0:52    Post subject: Reply with quote
in my post above snr went up - no decrease in rssi

this is on a real link over a couple miles, not on a table next to the AP or in the same house.

I believe my eyes, what others believe it up to them.

_________________
linksys GSv2, Gv4, Gv2, GLv1, G-TM, Buffalo wbr2, whr, whr-hp, whr-g125, wli-tx4-g54hp, Moto wr850gp, Alix.3C2
Goto page Previous  1, 2, 3, 4, 5, 6, 7  Next Display posts from previous:    Page 6 of 7
Post new topic   Reply to topic    DD-WRT Forum Forum Index -> Broadcom SoC based Hardware All times are GMT

Navigation

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You can attach files in this forum
You can download files in this forum