Thinkpads Forum

Ok,

I have been on a mission to figure out all the configuration options of the T60 with N cards, WWAN options, 3rd ant, Rev A etc etc...I was getting confused.

Conclusions:

It would seem that for those of us who want faster WWAN performance by using a WWAN card that supports Rev a or a faster bus speed with the Express card standard(vs PCMCIA)...you may want to opt for buying your Ext WWAN card from VZ direct vs lenovo's ext WWAN option the VZ V620 ext WWAN card (PCMCIA) at $249

Also, based on what I learned, I am not so sure of the advantages I get by getting the lenovo internally integrated WWAN with 3rd ant (Non Rev A btw) when I can get a Rev A Ext WWAN card today with dual Ant support.

So here's my reasoning from what I learned after talking to Lenovo support and VZ support today:

1) The t60 3rd antenna has to do with the buldge on the side of the LCD and is provided only for the "built in" VZ WWAN option i.e. (internally integrated into system) and not the external VZ WWAN card option at $249. ( I did not realize their were two options).

Re: Performance of Int vs Ext
Rep said, typically the ext WWAN card is actually faster anyway?.

Also, the 3rd ant. has nothing to do with the N card WIFI as I thought it did. That works off of the two Ant already configured for WIFI. ( I had thought the N card was somehow tied directly into the use of a 3rd ant - not according to this tech at lenovo?)

Update: More replies below suggest Lenovo is wrong - N is connected to a 3rd ant.

Also, it sounded as if adding the ncard was not a big deal. You add a chip like device on the motherboard. I assume this means you add the lenovo branded a/b/g/n chip card. (you might talk to tech more), but I heard no complicated wiring issues relative to N. Its the int embedded WWAN device that gets wired to 3rd ant.

Update: Again - if their is a 3rd wire than you would have to factor that in as well. Easier to start with N card.

2) Performance Question:

So, relative to selecting the best external WWAN option for a T60 - I see no real "performance" benefit to getting the internal lenovo WWAN because

a) ext cards are faster says lenovo (disagreement on this issue still- likely needs clarifcation express vs pcmcia etc)

b) the new ext cards have two antennas - one built in and one you can mount thus eliminating attraction of the 3rd ant built into t60 in you assume use of an ext rev a card.

***Please weigh in if you have different opinion

and with regard to external options, it would seem that the lenovo/VZ ext wwan option does not make sense when you can now buy the newer version of the V620 called v640 (with faster expresss interface) or opt for the new Rev a version although its the slower PCMCIA WWAN card.

Question -

If its true that the external WWAN express cards perform faster than ext WWAN PCMCIA cards at the bus layer, which offers better WWAN performance (the faster Express card "without" rev a WWAN support) or the slower PCMCIA card "with" Rev A? Assume the region your in supports rev a). Oh - and their is also an ext USB card that supports Rev A. Any thoughts on performance USB vs PCMCIA?

Rev A regions VZ support s today: http://b2b.vzw.com/broadband/RevA.html

VZ sells three ext WWAN cards(2 PCMCIA, 1 USB) which support Rev. A today...One is ready to go the PC5250(also supports MAC) and the Air 595 and a USB option. Each has built in ant and support for 2nd ant. mounted into the card with a wire you can mount wherever.

Not sure when express and rev a cards will ship at least from VZ

Rev A Info:

EV-DO Rev A will deliver even faster data speeds and greater efficiencies for web browsing and accessing your corporate intranet and email.

Faster downloads — typical speeds of 450-800 Kbps.*
Faster uploads — typical speeds of 300-400 Kbps.*
Lower latency — improved performance for many applications running over the wireless broadband network.
Backward-compatible — BroadbandAccess Rev A-enabled devices work seamlessly across the BroadbandAccess and NationalAccess service areas.

[/u]

There are too many errors in the information people have given you. Rather than address them one at a time, let me explain the setup as follows:

Re: WLAN antennas
-- Internal WLAN (WiFi) uses 2 built-in antenna wires for non-"N" cards, and 3 built-in antenna wires for "N" cards -- however, the 3rd wire for "N" cards is mutually exclusive with the presence of the WWAN internal antennas.

-- If there are WWAN internal antennas, then even the "N" cards use only the 2 non-"N" wires, leaving open the middle connector on the "N" card.

Re: WWAN antennas
-- There can be 2 built-in WWAN antenas. One is the "bulge" at the right of the LCD case, the other is under to back left side of the keyboard bezel. The internal WWAN cards connect to these 2 antenna wires.

-- For several good reasons I doubt very much that external EV-DO Rev.0 cards perform better than the internal EV-DO Rev.0 cards with the internal antennas. No point in debating this; there is no engineering reason why the external cards should have an advantage and I don't believe they do.

Summary:
There are at most 4 antenna wires built in: 2 for WLAN and 2 for WWAN. If there is not WWAN, then the position of one of the 2 WWAN wires can be used for a 3rd antenna for the WLAN "N" cards. (Note: The 3rd "N" antenna is not the same antenna as the WWAN antenna it replaces, although it occupies the same physical location in the LCD panel.)

interesting.

Thus, we are back to yes 3 built in ant for the N card.

Re: Int/Ext performance

I think the techs observation was limited to hw connection speeds and not access stabiliyt issues ie more antennas etc. But, that may be wrong as well.

This info must be so new that these reps have not had the frequency of relatd questions that we get here on the MB.

Relative to the other part of my post, the issue of Rev O and Rev A remains open as a perf issue IMO.

I think the issue of antennas becomes mute. I would think an ext rev a PCMCIA card(with dual ant) overides internal WWAN with 2 ants, but I'd love to learn otherwise.

And if the n card does get 3 ant its nice to know its benefiting beciase thats a direct link that I cannot replace that I am aware of.

Alltweed wrote:I would think an ext rev a PCMCIA card(with dual ant) overides internal WWAN with 2 ants ... And if the n card does get 3 ant its nice to know its benefiting

Of course an external Rev. A card will outperform an internal Rev. 0 card. This is the result of Rev. A vs. Rev. 0 -- it has nothing to do with external vs. internal antennas.

However, if/when internal Rev. A cards become available, I believe they will be preferable to external cards -- for both convenience reasons and performance reasons -- for people who do not need to remove the cards in order to use them on multiple machines.

Separately, the 3rd "N" antenna provides the "N" card the full functionality of its design. However, for me, were it a choice between A=(3 internal "N" antennas with external EV-DO card) and B=(2 internal "N" antennas with internal EV-DO card), I would pick B. Of course, this is a matter of personal preference based on the perceived cost/benefit.

Absolutely! Agreed.

I thought you might catch that

as I did not preference the correct context meaning my remark had nothing to do with your reply, but I wanted to try put the antenna issue into some context relative to rev a and perhaps ext express.

Wonder if the ext express cards(non rev a) have anything over internal Rev O WWAN?.

Thanks for feedback.

Alltweed wrote:Wonder if the ext express cards(non rev a) have anything over internal Rev O WWAN?

For stock Rev.0 external cards using their built-in antennas, the performance differences with internal Rev.0 cards are likely to be inconsistent and vary with the peculiarities of specific chipsets and designs. For cards which use the same chipset, there is not likely to be any difference in performance.

The advantage of the better (i.e., larger) internal antennas is likely offset by the disadvantage of longer leads which connect to the internal cards. Further, there is the possibility that the very small UHF connectors on the cards are improperly connected to the very small UHF connectors on the antenna leads. The external cards don't have this potential for bad connections since their antennas are built-in. Still, assuming that connections are good, performance differences between similar external and internal WWAN cards are likely to be statistically insignificant.

One last point: Some external WWAN cards have the connector(s) to use discrete (i.e., non-built-in) antennas. This opens the door to major performance leaps which are not possible with internal cards which do not have that option.

EOMtp wrote:One last point: Some external WWAN cards have the connector(s) to use discrete (i.e., non-built-in) antennas. This opens the door to major performance leaps which are not possible with internal cards which do not have that option.

Excellent point. Perhaps you will want to use your laptop in a room that has poor reception. By placing an external antenna by the window, or even on the roof, you can dramatically increase signal strength. I use an external antenna with my Sierra Wireless 860 AirCard (Cingular). I have read on the Cingular forums that some have been able to get a 3G signal (HSPDA) as far as 20 miles from a cell tower using an external antenna mounted on the roof of their home. Note that external antennas come in different sizes and mounting configurations. The most common is a 4" magnetic mount antenna. This is what I use.

Regarding internal vs external WWAN cards. On the plus side, external cards are easily replaced to keep up with rapidly advancing technology. Additionally, you can change carriers without the headache of opening up your laptop. On the minus side, many if not most external cards stick out of the side of the PC Card or ExpressCard slot. If you carry your laptop around, you will likely want to remove the card when you carry it to avoid potential damage. I like the 860 Aircard in that it does not stick out at all.

I see that Sierra Wireless will have USB WWAN modems available this quarter; both EVDO rev. A and HSDPA 3.6 Mbps. This looks like a very attractive option to me. The only thing is, from looking at the site, I do not see that it has an external antenna jack - I could be wrong though.

Sierra Wireless wrote:The AirCard 595U for EV-DO Revision A networks and the AirCard 875U for HSDPA networks are scheduled to be available Q1, 2007.

http://www.sierrawireless.com/product/usbmodem.aspx

Alltweed wrote:(...)and with regard to external options, it would seem that the lenovo/VZ ext wwan option does not make sense when you can now buy the newer version of the V620 called v640 (with faster expresss interface) or opt for the new Rev a version although its the slower PCMCIA WWAN card.

Question -

If its true that the external WWAN express cards perform faster than ext WWAN PCMCIA cards at the bus layer, which offers better WWAN performance (the faster Express card "without" rev a WWAN support) or the slower PCMCIA card "with" Rev A? Assume the region your in supports rev a). Oh - and their is also an ext USB card that supports Rev A. Any thoughts on performance USB vs PCMCIA?(...)

OK, so let's look at some performance numbers:

PCMCIA (byte mode): 80 Mb/s
PCMCIA (word mode): 160 Mb/s
CardBus (byte mode): 264 Mb/s
CardBus (word mode): 528 Mb/s
CardBus (burst mode): 1.07 Gb/s
ExpressCard (USB): 480 Mb/s
ExpressCard (PCIe) 2.5 Gb/s
USB 1.1: 12 Mb/s
USB 2.0: 480 Mb/s

EV-DO Rev. A: 3.1 Mb/s

Now, comparing those rates, wouldn't you agree that it really doesn't matter how you connect your EV-DO modem?

I find it odd that Lenovo would run a 3 antenna N card using only two antennas, as this implies no load on the third transceiver (should the laptop being using a WWAN card occupying the third antenna).

Either the card is smart enough to shut down that transceiver (entirely possible, if it detects no load on the port), or they must feel the output power of the third transceiver is too low to possibly damage the transceiver (and at 250 mW, it could easily do damage).

Running a transceiver without a load is generally a no-no on most RF systems, even the relatively low power WiFi gear. Might not cause immediate damage, but you're also getting zero performance out of the system as well, as with no load / impedance on the transceiver, it can neither Tx or Rx worth a [censored].

This would basically eliminate the MIMO advantages of 802.11n, AND possibly damage the card.

For some reason, the information from the Lenovo tech isn't adding up here.


Patrick

For 802.11n MIMO, do you not require an antenna pair for each independent stream? That would imply a minimum of four antennas, not three, if correct -- and mean that the third antenna is really only for WWAN, and not for 802.11n.

But I'm on rather thin ice here -- I certainly may have misunderstood MIMO in my readings.

This is certainly something that I have not studied. A quick Google search brought up the following article which seems to indicate that only (2) antennas are required.

DEVICEFORGE.COM - by James M. Wilson (Aug. 9, 2004) wrote:Supporting at least two spatial data streams will require a minimum of two transmit antennas on all 802.11n implementations. Support for more than two transmit antennas, or two spatial streams, should be optional, with the maximum number limited to four for practical reasons.

http://www.deviceforge.com/articles/AT5096801417.html

My understanding is that 802.11n support anywhere from 1-4 antennas, with more antennas yielding higher bandwidth. This link seems to support that...

http://www.radio-electronics.com/info/w ... 02-11n.php

GomJabbar wrote:This is certainly something that I have not studied. A quick Google search brought up the following article which seems to indicate that only (2) antennas are required.

http://www.deviceforge.com/articles/AT5096801417.html

The same article also says ...

Each spatial stream requires its own TX/RX antenna pair at each end of the transmission (...)

Basic 802.11 pre-n requires an antenna pair as well. This seems to imply that you would need to go to 4 or 6 antennas. On the other hand, jeffm's link makes it much more cloudy again.

Wikipedia's articles on 802.11 and MIMO also imply multiple transmit and receive antennas. But maybe it's possible to have an asymmetric configuration, with # receive antennas <> # transmit antennas, a la ADSL?

Perhaps 802.11n does not require more than one antenna, but 802.11n MIMO does require at least two antennas. From the same link that I posted above:

DEVICEFORGE.COM - by James M. Wilson (Aug. 9, 2004) wrote:Each spatial stream requires its own TX/RX antenna pair at each end of the transmission (Figure 1). It is important to understand that MIMO technology requires a separate radio frequency (RF) chain and analog-to-digital converter (ADC) for each MIMO antenna. This increasing complexity ultimately translates to higher implementation costs as higher-performance systems are required.
------------------------
Intel believes both MIMO technology and wider bandwidth channels will be required to reliably satisfy the higher throughput demands expected from 802.11n. Choosing conservative increases in channel bandwidth, combined with conservative approaches in MIMO technology will enable cost-effective solutions that meet such requirements. A combined approach, employing both MIMO and 40-MHz channels, will enable the IEEE 802.11n technology to reach even higher performance as Moore's Law and CMOS process technology improvements advance DSP capabilities.

Article below offers some additional information.

BROADCOM CORPORATION White Paper - 802.11n: Next-Generation Wireless LAN Technology (April 2006) wrote:The emerging 802.11n specification differs from its predecessors in that it provides for a variety of optional modes and configurations that dictate different maximum raw data rates. This enables the standard to provide baseline performance parameters for all 802.11n devices, while allowing manufacturers to enhance or tune capabilities to accommodate different applications and price points. With every possible option enabled, 802.11n could offer raw data rates up to 600 Mbps. But WLAN hardware does not need to support every option to be compliant with the standard. In 2006, for example, most draft-n WLAN hardware available is expected to support raw data rates up to 300 Mbps.
-------------------------
With all the optional modes and back-off alternatives, the array of possible combinations of features and corresponding data rates can be overwhelming. To be precise, the current 802.11n draft provides for 576 possible data rate configurations. In comparison, 802.11g provides for 12 possible data rates, while 802.11a and 802.11b specify eight and four, respectively.

http://www.broadcom.com/docs/WLAN/802_11n-WP100-R.pdf

The wikipedia link that tomh009 mentioned linked to chapter 11 of 802.11 Wireless Networks, The Definitive Guide. This chapter is entitled A Peek Ahead at 802.11n: MIMO-OFDM.

Here's the link: http://www.oreilly.com/catalog/802dot11 ... r/ch15.pdf

Anyway, there's a lot of information there, but a quick perusal seems to support the idea of 1-4 (or maybe 2-4) antennas - unless I'm mis-understanding something (which is very possible since I'm not an engineer)...

The number of spatial streams can be equal to 1, 2, 3, or 4. It must be less than or equal to the number of transmission antennas. Support for at least two spatial streams is mandatory.

and...

The number of transmit antennas, noted by xTX, where x is the number of transmit antennas. It ranges from 1 to 4, although a single antenna is only supported for 40 MHz channels. All 20 MHz channels must use at least two transmit antennas, though they may have only one spatial stream.

Here is some more information from the Broadcom White Paper quoted above:

BROADCOM CORPORATION White Paper - 802.11n: Next-Generation Wireless LAN Technology (April 2006) wrote:Diversity exploits multiple antennas by combining the outputs of or selecting the best subset of a larger number of antennas than required to receive a number of spatial streams. This is important because the draft-n specification supports up to four antennas, so devices will probably encounter others built with a different number of antennas. A notebook computer with two antennas, for example, might connect to an access point with three antennas. In this case, only two spatial streams can be used even though the access point itself may be capable of three spatial streams.

This seems to indicate that each antenna (for 3-antenna systems) works off of the other two - somewhat analogous to 3-phase power systems.

So in relation to the origianl post, it sounds like you should be able to install the draft-n card in a thinkpad that does not have the 3rd antenna, but with reduced performance.

jeffm wrote:Anyway, there's a lot of information there, but a quick perusal seems to support the idea of 1-4 (or maybe 2-4) antennas - unless I'm mis-understanding something (which is very possible since I'm not an engineer)...

The number of spatial streams can be equal to 1, 2, 3, or 4. It must be less than or equal to the number of transmission antennas. Support for at least two spatial streams is mandatory.

Right ... now 802.11a/b/g has a single spatial stream, so that's the "1" option. But note that this quote says the number of spatial streams must be less than or equal to the number of transmission antennas. Several of the other articles imply that you, in fact, require an antenna pair for each spatial stream -- a transmit and a receive.

Now, if that three-phase comparison holds, maybe a three-antenna scheme would get you two spatial streams -- don't know. I may have been an engineer for many years, but software engineering doesn't help much with wireless technology!

The quote I took previously stated: "A notebook computer with two antennas, for example, might connect to an access point with three antennas. In this case, only two spatial streams can be used even though the access point itself may be capable of three spatial streams." I take this to mean that three antennas equals three spatial streams capability. And if a notebook with two antennas has two spatial streams capability, then that would seem to indicate that a single antenna would have one spatial stream.

I believe I will have to retract my analogy of 3-phase power systems. It appears to me that each antenna is capable of a single spatial stream; since four antennas is the maximum specified and four spatial streams is the maximum specified.

I believe the confusion here is with MIMO. MIMO requires a minimum of two antennas. Not all 802.11n devices are MIMO capable. There is a diagram in the first link I posted above from DEVICEFORGE.COM that shows a simple MIMO setup with two antennas and two data (spatial) streams. AFAIK, the transmit and receive antennas are one in the same. The same antenna is used for both purposes.

Furthermore, that same White Paper states the following regarding 802.11n:

BROADCOM CORPORATION White Paper - 802.11n: Next-Generation Wireless LAN Technology (April 2006) wrote:802.11n * Standard Approved - Not yet ratified * Maximum Data Rate - 600 Mbps * Modulation - DSSS or CCK or OFDM * RF Band - 2.4 GHz or 5 GHz * Number of Spatial Streams - 1, 2, 3, or 4 * Channel Width - 20 MHz or 40 MHz

I believe jeffm is correct in his deduction: "So in relation to the origianl post, it sounds like you should be able to install the draft-n card in a thinkpad that does not have the 3rd antenna, but with reduced performance."

"I believe jeffm is correct in his deduction: "So in relation to the origianl post, it sounds like you should be able to install the draft-n card in a thinkpad that does not have the 3rd antenna, but with reduced performance.""

The issue I was trying to resolve with regard to ordering the N card was learning whether it comes with a 3rd Ant or how you would order it with 3rd ant to get best perf. If it does, is it seperate from the integrated WWAN 3rd Ant or does it share it?

The lenovo tech told me it has no 3rd ant. The only 3rd Ant on the T60 is the buldging lcd for the integrated WWAN and it is not connected in anyway to the N card. Thus, there appears to be no way to order a T60 that optimizes N use - you just order it so it gets plugged into MB at time of config connecting to the existing two ant.

Eitherway, people were reporting better stability and performance for b/g with the N card and thats why I wanted the N and if I can get super G perf - all the better..

Alltweed wrote:... Thus, there appears to be no way to order a T60 that optimizes N use - you just order it so it gets plugged into MB at time of config connecting to the existing two ant. ...

Incorrect - please read my post above (the second one in this thread) for clarification. The T60/p can be configured with a 3rd antenna for "N"-card use, and this antenna would be in place of the "bulge" on the LCD cover. Here are the part numbers:
-- Wireless LAN 3rd antenna (14.1 in) = FRU #42R9909.
-- Wireless LAN 3rd antenna (15.0 in) = FRU #42R9910.

Wow - Your right.

I meant to go back to re-read your post earlier and forgot. Thanks. Now that adds some clarity.

Thus, w/o the int WWAN card, a third ant can be attached to N card and it fills that space where the buldge exist on lcd panel.

FYI: The first thread I saw on the N card/3rd ant issue is at:

http://forum.thinkpads.com/viewtopic.php?t=33434

The important thing to note about antennas in a WiFi environment is that they are not dedicated to either Tx or Rx; they are in fact, transceiver antennas, not Transmitter OR receiver.

So if you have 2 antennas, it's possible to do 2x2 MIMO (two Tx, two Rx), as the laptop would only be transmitting, OR receiving at any single point in time (It's not a Full-Duplex, Frequency Division system).

With a third antenna, you can do 1x2 SIMO, or 2x1 MISO (single input, multiple output, or multiple input, single output). This would have the effect of being able to RECEIVE multiple spatial streams, yet only transmitting a single stream. In a WiFi environment, this optimizes the system for downstream communications, which the vast majority of WiFi access tends to be.

All of this is dependent upon the chipset in the device, and the drivers running that chipset.

My primary concern remains that of having an unterminated transmitter powering into no-load. This tends to cause feedback and either an over-voltage or over-current situation, which will often smoke the amplifier.

Without knowing exactly what's on those Mini-PCI cards, I can't be sure that the third antenna may be possibly a Rx only system.


Patrick

There are some more good articles on the Atheros XSPAN site:
http://www.atheros-xspan.com/

If you browse their chipset solutions, you will see that all their 802.11n chipsets and reference designs, including the AR5008-3NX that I believe Lenovo uses, have three transmit chains. Now, Lenovo appears to be only using two of those three, at least ub some configurations. Given that IBM/Lenovo has a longstanding relationship with Atheros, I have to believe that not using all the antennas is OK ... either (1) the transmitters won't burn out, (2) the chipset is clever enough to detect the lack of an antenna and turn off the transmitters, or (3) the ThinkPad BIOS is capable of doing the same.

P.S. Patrick, thanks for clearing up the transmit/receive/transceive antenna confusion!

Guys,

I just looked inside my T60 8744-5BU in a quest to diagnose an erratic wireless signal. (If you have any ideas I'd welcome them. That post: http://forum.thinkpads.com/viewtopic.ph ... ht=antenna )

The news for this thread is that my a/b/g/n card has leads connected to all three antenna terminals exactly as drawn in the HMM. I was not inclined to crack open the lid to see if there were actually antennas at the other end, but one assumes so. The wires went off in the right directions.

(The 8744-5BU comes configured with a/b/g/n and no WWAN.)

RA