MC75 WLAN performance vs MC9090

All,

Just looking for some validation of an issue I think we are seeing at a customer.

Up till a couple of months ago they were a big user of MC9090 terminals (approx 2,500 units across 330 sites) and whilst we would get the occasional connectivity problem they have been pretty stable.

Over the last 2 months they have replaced 90% of the MC9090s with MC75As and sinc then their helpdesk has received nearly 400 calls for devices losing network connectivity and in many instances it requires a warmboot to solve it.

They are running the latest BSP with the latest Fusion driver on top.

Is there really a huge difference in the performance of the MC9090 and MC75A WLAN, or do I have a fusion bug?

I will be gathering more data on this shortly but if anyone with any thoughts they would be appreciated.

Cheers,

IJ
Mario Guillen
Ian, I also have a U.S.

Ian,

I also have a U.S. Federal Government customer with a large deployment of MC 9090s who have been upgrading them to MC 75As.  They have reported that the MC 75As do not have the same WLAN range which the MC 9090s have.  They have not had to reboot, but do have to walk back towards the AP to reestablish an RF connection which makes the intended task more difficult.

Mario
Vote: 
Vote up!
Vote down!

Points: 0

You voted ‘up’


Mark Mann
Hi Ian, Has the customer

Hi Ian,



Has the customer reported that they are seeing Adapter Unavailable and must do a warmboot to get the radio card to re-connect to the WiFi network?



I've seen this happen on some of our newer terminals (MC55N)



Mark

Vote: 
Vote up!
Vote down!

Points: 0

You voted ‘up’


Anonymous (not verified)
Mario, I have an airline

Mario,

I have an airline customer that has found the same difference in WLAN range with our newer devices as your government customer.  We found that WLAN range decreased when the customer migrated from the MC9060 to the MC9090.  They are now seeing an additional decrease in range when migrating to the MC9590.
Vote: 
Vote up!
Vote down!

Points: 0

You voted ‘up’


Marc Fluhrer
If I remember correctly the

If I remember correctly the MC75A, MC95XX, MC9190 and the newer devices have the Jedi Radio and the old devices like the MC9090 have the Photon radio. The Jedi radio has a little less trasmit power then the Photon did, but with the latest BSP on the Jedi terminals you actaully don't see much of a difference in coverage.

The one place you may see a difference is if the WLAN at your customer site is locked at a Minimum rate of 11mbps and the cells are very spread out, meaning there is not much overlap of the 11mb coverage. You may lose connectivity at the edge on a MC9190 when you did not with a MC9090.

One easy solution is to see if the customer can go to a 5.5mb minimum rate to increase their cell size. 
Vote: 
Vote up!
Vote down!

Points: 1

You voted ‘up’


Mario Guillen
Marc, Have you been able to

Marc,

Have you been able to quantify the RF range difference between the Photon radio and the Jedi radio, or between the previous BSP for Jedi as compared to the most recent BSP?  What is the latest BSP number on which you observed the improved RF performance, and on which MC?

Thanks.

Mario
Vote: 
Vote up!
Vote down!

Points: 0

You voted ‘up’


Marc Fluhrer
I have a customer site I was

I have a customer site I was at on Sunday doing a WiNG update to 5.2.12. I brought a MC9190(BSP41) and MC9090(BSP43). Locking into a single AP with both terminals using the diagnostics on the terminals doing a ICMP ping test, They got pretty much comparable coverage walking down the aisles with both terminals. I was probably about 90 - 100 yards away from the AP. I did not keep going to see when the 9190 would not ping anymore since I was a significant distance in the warehouse away from the AP I was locked into. This warehouse is default data rates in WiNG, which dictates a 1mb basic rate, so I am sure this helps.

When we first installed it in November I had BSP 34 on my 9190 and saw the 9190 stop pinging about 50 feet before the MC9090 would stop pinging. Something changed, not sure what, not the power, but maybe something in the radio drivers. The test I did Sunday with the new BSP, both terminals reached the same spot the MC9090 did in November.

In November, both terminals will see the AP having the same db power, but what I saw with the 9190 on BSP 34 was that it would stop transmitting even though it would still see the AP at the same power level as the MC9090.

If you look at the Wireless Status in the Fusion driver, you can see the power levels it is transmitting at on both the MC9090 and MC9190.
Vote: 
Vote up!
Vote down!

Points: 0

You voted ‘up’


Mario Guillen
Marc, Thanks.  That's very

Marc,

Thanks.  That's very useful information.

Mario
Vote: 
Vote up!
Vote down!

Points: 0

You voted ‘up’


Ian Jobson
All, The customer who is

All,

The customer who is experiencing problems has a Spectralink voice deployment so the vast majority of their stores have very good (-65 to -67db) coverage throughout, and the back bone is Moto WS5100s with AP300.

The MC9090s were running DCP1.8 (latest CE5.0) build.

The MC75s are running BSP41 with Fusion 3.00.2.0-022R

So I know everything has the latest build and the quality of the RF in store is generally very good and has not changed in the process of swapping the terminals.

IJ
Vote: 
Vote up!
Vote down!

Points: 0

You voted ‘up’


Juan-Antonio Ma...
Silly question, yet relevant:

Silly question, yet relevant: what type of security are they using? Have you noticed/tested if using no security at all yields better results?
Vote: 
Vote up!
Vote down!

Points: 0

You voted ‘up’


Ian Jobson
Security is WPA-TKIP exactly

Security is WPA-TKIP exactly the same as the MC9090s.

Nothing has changed to the config or the infrastructure, the only difference is the handheld device.

IJ
Vote: 
Vote up!
Vote down!

Points: 0

You voted ‘up’


Juan-Antonio Ma...
Have you tried to create a

Have you tried to create a non-secure profile on the ws5100 (with anoter ESSID) and test with this?
Fusion versions are not only different due to chips (Jedi/Photon), but they also use a different authenticator software (actually APIs), as far as I know (pls someone correct me if I am wrong).
"Culprit" could well be this.

On the other hand, since you have been there, is there any chance that multipath effect could be happening (basically, lots of metal stuff, panels, racks, etc, you may have seen)?
MC75 are very much sensitive to this, we are actually replacing the whole infrastructure at a customer which is having huge issues with MC75 and multipath.

Finally, are rates 6 and 9 disabled? It seems as if they ought to with our infrastructure. Yet, I really ignore if this is an "urban legend" since I heard it several times but I could not find ever a document to state it! This was to be done with our old DRDOS terminals for sure, but now... I am not sure.
What I am positively certain of is that 802.11g forces 6 to be enabled -and actually as a basic rate.
Vote: 
Vote up!
Vote down!

Points: 0

You voted ‘up’


Log in to post comments