When you open the MAX 4000 as a technician,
you will see a huge mainboard with a lot of discrete components. From our view of 2003, it is an old design..
OK, you may say, I would never design such a big board in case of service an modularity. But is was 1992 or a little bit later when Ascend did start as a company designing the first Multiband MAX. The MAX 4000 was born later in 1994.
The MAX was the first dial in router, other say "access concentrator" or "access server" or whatever they gave it a name, which could easily compare against the best of the world.
From 1996 on the MAX 4000 was definitely the best in its class and they got it paid.
What was the reason for that success ? The first thing was the quality of the components. Most of them are extremely good. Secondly they did not overstress the specifications by thermal power and by speed like overclocking. There have been no internal connectors beside the power connector to run the MAX.
Taking a look onto the big mainboard of a MAX 4000 you can see three c-mos crystal oscillators and six crystals at all, each for its own group of components. That means, each functional group has its own master clock, the CPU, the Serial, the Ethernet, the four T1/E1 Wan channels, the high speed serial, the busclock for the digital modems slots. And the c-mos oscillators are specified for widest temperature and smallest frequency tolerance and drifting.
That was the reason, why the MAX did run all over the world in all environments even the Brasilian djungle with all different carriers and telcos. So the TAOS was enhanced and enhanced up to 7.0.28. Then the 2 Mbyte flash could no mor take any more bit and the flash is fixed onboard, no chance to expand this.
And now in 2003, when peaople have probelms with E1 lines, I am comming with my MAX, plug it in and it works fine, suprising a lot of experienced old computer people.