?
That's 20:1.

In reply to a post by RobertoS:

?
That's 20:1.

It may work out that way, but it is actually different!

Take a simple view of using voice circuits. Five customers connected to a small exchange, and only one out going line from the exchange, you have 5:1. Customer A can make a call, when customer B tries he cannot. Add another 5 customers and if you give them another single outbound line contention remains at 5:1. However, if you make both outbound line available to all users, the contention becomes 10:2 and rte chance of customer B being able tpo make a call is significantly improves.

Thanks for all your interesting replies but I have now cured the speed problem and getting 36-40Mb (my limit).

I tried my mobile android phone nearer to the modem and got 37Mb so knew the connection was OK. The main computer is in another room about 8 metres away through a double thickness wall and a large tall cupboard. It receives the BB via a USB RX on a 2 metre quality extension cable.
The modem is in exactly the same position as when installed and the RX hadn't moved (clipped to a wall shelf). However moving it just a couple of inches restored the speed. A bit of a mystery as neither units had moved position so I had wrongly discounted this as the cause.

It's amazing what a small movement positioning it can make, but I've noticed that with video senders, etc before at those SHF frequencies.

Thanks again for all your comments; I guess I was barking up the wrong tree regarding contention rates.

Mike.

So, it was a WiFi problem and not the connection!

In reply to a post by RepairExpert:

Thanks again for all your comments; I guess I was barking up the wrong tree regarding contention rates.

So after all that it was down to use of wifi for the final link? Did I miss that being mentioned somewhere? Standard advice for all issues is to start off the investigation with a wired connection.

Wifi should only ever be viewed as a convenience for mobile devices(*). It compromises reliability and throughput in exchange for not being tied to one place.

(*)Or an unfortunate situation if you really can't run a proper cable

Edited by Andrue (Fri 18-Jul-14 10:47:31)

LOL
Taking your two cases:-

1) One line, 5 customers. There is a 4:1 chance that B will not be able to make a call.

2) Two lines, 10 customers. When B tries to make his call there is a 9:1 chance that a line is in use, and an 8:1 chance that the other is in use.

He has more chance in (1).

This is nothing whatsoever to do with broadband contention anyway. That allows both A and B, or indeed all ten, to connect at the same time. The issue is the throughput speed, not the availability of connection.

I have never seen contention ratios quoted in that way.

50:1 can mean 50 users on one resource, but it can also mean 500 users on 10 resources. Both are 50:1.

I have, many times and still do see it in regular usage.

10 groups of 50 users each after 1 resource is truly 50:1 but if there are 10 resources available for 500 users it would be quoted as 500:10

In reply to a post by MHC:

I have, many times and still do see it in regular usage.

10 groups of 50 users each after 1 resource is truly 50:1 but if there are 10 resources available for 500 users it would be quoted as 500:10

That does make sense since, as your first post implied, there is a difference in performance between the two. 50:1 implies that you only get the full speed while no-one else is using it. 500:10 implies that several people can get the full speed at the same time (the equivalent of 9 other people at full speed).

Edited by Andrue (Fri 18-Jul-14 16:22:08)

Exactly.