Propeller rulez

Andrey Demenev

Just a random thought (sorry for this "pathetic" title)

I am now doing a project using ATMega32, primarily for my personal use, with a possible perspective to sell the completed units. I was considering Propeller for it, but finally decided to use Atmel chip, to save few bucks. While writing firmware, I constantly ask myself : how would I do this or that if it was the Prop? And the more I think about that, the more often I answer to myself : "easier, faster, more beautiful". Just an example : I have two sensor inputs, each can accept either analog or digital sensor, each analog sensor can be in 1 of 4 voltage ranges, and I have to check for 1 of 4 conditions defined through config (first sensor fired, second fired, any of the 2 fired, both fired). For digital sensors, 1 or 0 can be active state (from config), for analog - above or below threshold. I have to check for all possible combinations, and that would eat tens (if not hundreds) of microseconds, a valueable amount of time for the application.With Propeller, I could just generate optimized code for each combination, and take decisions in microseconds. And this is just one routine - and there are others. And not only hardware costs money, development time does as well. .

Hell, I may be changing my mind and using a Prop for this. Anyway, I have made a mistake in schematic that cost me a burnt LCD, so I have to make another board anyway.

Cluso99

Too true Andrey. And don't forget lost time to market, and the additional grey hairs (or less hair) Haha.

This has been my point in using multiple props. Everything is so much easier.

MagIO2

Your random thought is nothing new for the audience in this forum ;o) You should better spread such "news" in other forums.

And I think the "saving time" starts much earlier. Buy a prop and a prop plug and you're in business. You can do I2C, LCD, TV, VGA, Keyboard, Mouse, RS232 and much more in as little as one day. And not only as a demo .. you can already use all that in a complex project. Try this with another controller ... maybe except with the other multi-core controller that's available but which is a big nono in this forum ;o)

Don't want to count that poor guys that set the wrong fuse bits when starting with AVRs. If you want to do some serious development with other microcontrollers you have to learn a lot of lessons, where you simply have to pick an object from the ObEx in propeller world.

Andrey Demenev

MagIO2 wrote: »

If you want to do some serious development with other microcontrollers you have to learn a lot of lessons

Well, I have over 5 years experience with "other", so the lessons are learned already

davidsaunders

Yes the Propeller is amazing. About 5 weeks ago, I was working on a VGA bitbanger using an ATMega and a friend showed me his Propeller Demo Board clone, and gave me a Prop Plug and a few Propellers, I was hooked instantly. I can do most things on the Propeller in about one third the time it takes on the ATMega, and to think I was happy to get 512x480x8BPP out of an ATMEGA with 8 resisters, 2 diodes, and a 32k SRAM, while the propeller gives so much more possibility.

I now only use one ATMega on my MultiDB68K, and that is only to push a video splash screen until the Propeller used for video finishes loading (all other peripherals are Propeller based), thus making the AVRs only advantage its internal FLASH (which is also a disadvantage). And I only need 4 Propellers where the initial design called for 16 ATMegas. Once the Propeller II comes out I will be able to reduce this to 2 Propellers, and have much better pATA, a much simpler Video adapter, much better sound, and way better 3D acceleration, not to mention the possibility of using the PropII directly for USB rather than relying on support ICs.