My first 10BASE5 network segment

[Warner Losh]

On Thu, Aug 25, 2016 at 11:16 PM, tony duell <ard at p850ug1.demon.co.uk> wrote:
>>
>> You will need to terminate the coax.
>>
>> (terminator)---------------<tap>------. . .
>> ----<tap>----------------(terminator).
>
> He said he has connected a 47 Ohm resistor at each end of the
> coax. That's close enough to the correct 50Ohm terminal to
> work.

47 and 56 were resistors of choice back in the day for make-shift
termination on thin-net networks in a pinch.

>> Anyway, you need to terminate the line or your going to have so many issues
>> you may not even get a packet to make it from one end of the line to the
>> other.
>
> Correct. You won't. No matter how short the coax is. You will get collisions.
>
> The reason is that the transmitter in a coax MAU is a current source which
> effectively develops a voltage across the terminators. The receiver is a
> voltage detector. A collision is sensed by the MAU if it sees more voltage
> across the coax than it should. This (on a correctly terminated cable) means
> that 2 transmitters are putting current into the cable/terminators at the
> same time.
>
> But if the termination resistors are too high or missing (even if only one
> is missing) then a single transmitter's current will develop enough voltage
> across the coax to be detected as a collision.

10base5 also had rules for minimum bend radius as well as tap locations
to be at the maxima of the reflection point. For early gear, failure to
put it at a vibration node would often result in unreliable behavior, though
I can't recall if that included collisions or not.

Warner