[Gllug] NT Embedded?

[Nick Hill]

And, quoted from Mocro$oft site:

>The team that developed a critical battery-monitoring system for the U.S.
>Navy's unmanned autonomous research submarine, chose the Microsoft® Windows NT®
>Embedded operating system along with VenturCom's RTX technology to achieve the >deterministic real-time performance required for the project. The system not >only collects data on more than 1,800 battery cells every 3.7 seconds, but also >provides easy portside connectivity to shore-based Windows NT?based systems to >offload data and upload application changes. With Windows NT Embedded Target >Designer, the project team was able to fit all the operating system >capabilities required into 22 MB of flash memory. With no moving parts, this >monitoring solution delivers quiet, reliable, 
>unattended operation for up to 12 hours underwater.
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Does NT stay up that long?

Perhaps they have installed a watchdog timer and use flash to store the values between reboots?

I could probably have programmed a similar system with PIC microcontroller assembly, in 8Mbit of ROM in 2 months, from scratch, without operating system which can fit on a single £1 chip.

I have programmed a universal autodetect UART (Serial controller) in software, capable of 230,400Kbit along with a zero-hardware (bar a capacitor and resistor) A to D converter in less than 512 bytes and no operating system. I was going to use it as a basis of a hall effect sensor for Earth's magnetic field, to be used in a car. The problem was the signal to noise ratio in reading the magnetic field value, made the system unreliable. (heat and electrical noise vs Magnetic signal). I created a sophisticated learning algorithm for the circuit to learn it's own parameters and change them over time to compensate for temperature, ageing, electrical noise etc. All in machine code and assembly.

The noise was too random in the inexpensive hall effect devices available, for me to get a good magnetic reading. Hall effect devices are really designed to measure much stronger fields from proximate magnets.

I later decided to use magnetic saturation as a way of detecting magnetic field. I soon got fed up with the project and lack of knowledge on magnetic saturation theory so I started manufacturing memory modules.


-- 
Gllug mailing list  -  Gllug at linux.co.uk
http://list.ftech.net/mailman/listinfo/gllug