Hi you guys,

Thanks for all the suggestions etc..

aka j:

I totally agree with you about the use of libaries. I hate using other peoples code, and consider it a sign of defeat! As of yet, I don't think I've ever written an assembly program using a subroutine that I did not write, and whilst I'm working alone, I'd like to keep it that way. Incidently, I don't suppose you could suggest a good place on the Internet where I could learn exactly how floating-point maths works. I have a reasonable understanding of fixed-point, although a bit more info on that wouldn't go amiss.

George:

I have to admit, your comparison of the 8051 code against the MSP equivilent was a bit biased to say the least. The MSP family sounds interesting and I will certainly look into it, thanks for introducing it to me!

Erik:

I wouldn't say that hobbiests stick to just the one family of microcontrollers. I have to admit the cost of buying books and development kit for endless different types of architecture can be a little high, and for most hobbiests (including my self), it can prove impractical. From what I understand, it is not unusal for designers in the comercial world to stick to one type of microcontroller. Changing even from a Dallas 8051 derivitive to an Atmel one could incure all sorts of costs.. different production and programming equipment, training costs etc...

I'm not sure I understand what you mean't in your second posting.

Peter:

There is certainly lots of life left in the 51.. I mended an RTX industrial robot arm a few weeks ago that used two Intel 8052s on its control boards which was built in 1986, when I was just a nipper! It shows what a brilliant architecture the 8052 must be for it to still be found in todays advanced designs!

Hans:

The 8051 does seem most suited to control, and it does a brilliant job. It's great when you need to do a lot of single bit manipulation, as is often required in control applications. My problem is that my current application really requires a lot of maths, preferably floating point, but at the moment it'll have to be fixed-point for various reasons. I don't think the 8052 is a good choice, firstly because it is 8-bit, and I can imagine all sorts of carry and borrow nightmares occuring. Secondly, there is no limit to how fast I want the calculations to be carried out. I want things to happen as fast as I can afford to and practically make them!

Stephane:

Hopefully you've received my e-mail. Thanks for the offer!


=======================================

Finally, for a hobbiest, and perhaps unusually? I tend to only work on projects that are of some commerical, or scientific nature. I know lots of hobbiests are interested in using microcontrollers for projects like digital dices etc... Of course everybody has to start somewhere, and no doubt people enjoy making circuits like that. I on the other hand jumped in at the deep end when I was learning about microcontrollers and have always tended to use them in fairly complex designs. I would love to be involved with microcontrollers as a job, but even though I'm an electronic, electrical and computer engineer, don't have the relevant qualifications, or commercial experience to get into that field.. not yet anyway!

The unfortunete position this puts me in is that I am very interested in progressing to the higher performancc microcontrollers and DSPs availible, but because I'm a hobbiest can't affort it, or have the time to be able to do so.. Even things such as the package type of a microcontroller determine what I can and cannot use.. anything more complicated than an PLCC package usually puts a chip out of my reach!

I think I have pushed 8-bit devices as far as I can, there's not much else I can do with them, so now it's time to progress!

Again, cheers for taking the time to read this even more lengthy posting!!!

Matt.