Slashdot Mirror

The mappers in an NES cartridge are really simple. You can emulate most of them all in the same PIC. The PIC is fast enough to read from the address buses and update the data busses for both 'ROMs' in one NES clock cycle.

I have plans drawn up to make an NES game cartridge with a Compact Flash card, a PIC, 2MB of SRAM and some latches. The parts are cost prohibative right now though, so I haven't assembled or tried it yet.

Microchip has a new part, the MCP2150, which is designed to fit between a microcontroller's serial port and the infra-red transducers. Here are a bunch of application notes that apply. Earlier this week my local Microchip representative got me a couple samples of the MCP2150. I'm looking to make an interface between my Palm Pilot and a datalogging microcontroller. When will I get to it? Who knows- it might just add to my stack of sample parts...

One thing I really like about Microchip is that they are pretty committed to distributing free samples and free/low cost tools for development. If you've ever priced the development hardware from Intel or Motorola, you can see why this is such a big attraction to the small time/home developer.

Microchip has a new part, the MCP2150, which is designed to fit between a microcontroller's serial port and the infra-red transducers. Here are a bunch of application notes that apply. Earlier this week my local Microchip representative got me a couple samples of the MCP2150. I'm looking to make an interface between my Palm Pilot and a datalogging microcontroller. When will I get to it? Who knows- it might just add to my stack of sample parts...

One thing I really like about Microchip is that they are pretty committed to distributing free samples and free/low cost tools for development. If you've ever priced the development hardware from Intel or Motorola, you can see why this is such a big attraction to the small time/home developer.

Microchip has a new part, the MCP2150, which is designed to fit between a microcontroller's serial port and the infra-red transducers. Here are a bunch of application notes that apply. Earlier this week my local Microchip representative got me a couple samples of the MCP2150. I'm looking to make an interface between my Palm Pilot and a datalogging microcontroller. When will I get to it? Who knows- it might just add to my stack of sample parts...

One thing I really like about Microchip is that they are pretty committed to distributing free samples and free/low cost tools for development. If you've ever priced the development hardware from Intel or Motorola, you can see why this is such a big attraction to the small time/home developer.

If your using something large enough to require a OS, you either ...

1. Don't have an embedded application

2. Have over designed your product

Let's take a look at how you would code up a PIC 16C74 from Microchip.

org 0x00
goto START

org 0x04
goto INT_Vector

org 0x20
START
; Code to setup all your I/O states
MAIN_LOOP
; code you execute in the main loop
goto MAIN_LOOP

INT_Vector
; code to save register states
; code to handle interrupt(s)
; usually time critical I/O and timing
; code to restore register states
retfie ; return from interrupt

I give you a complete embedded O/S. If you need more than this, please refer to rules 1 and 2.

This chip with a FTDI FT8U245AM latched onto Port D gives you a complete embedded 20Mhz USB device (with linux drivers).

You need an OS? That is the rumor that keeps Microsoft and Slashdot in the green.

TastesLikeHerringFlavoredChicken

If your using something large enough to require a OS, you either ...

1. Don't have an embedded application

2. Have over designed your product

Let's take a look at how you would code up a PIC 16C74 from Microchip.

org 0x00
goto START

org 0x04
goto INT_Vector

org 0x20
START
; Code to setup all your I/O states
MAIN_LOOP
; code you execute in the main loop
goto MAIN_LOOP

INT_Vector
; code to save register states
; code to handle interrupt(s)
; usually time critical I/O and timing
; code to restore register states
retfie ; return from interrupt

I give you a complete embedded O/S. If you need more than this, please refer to rules 1 and 2.

This chip with a FTDI FT8U245AM latched onto Port D gives you a complete embedded 20Mhz USB device (with linux drivers).

You need an OS? That is the rumor that keeps Microsoft and Slashdot in the green.

TastesLikeHerringFlavoredChicken

Take a look at www.microchip.com. On every page they serve, they have a unobtrusive link called "Page Options" at the top where you can choose what page you want to get: text only, graphics or Java frame. As it turns out, I use all three versions from my university ethernet connection, depending on if I want the heavy-duty search in Java (like a MSFT help search, index, etc box), I just want to browse (I'll use graphics) or I really need something fast (text-only). It's not polite to NOT give these choices to the user!

It works great! I don't know how much more it costs them to do this, but it definitely makes for happy customers. Each version is based off a different root directory on the server and all three are probably generated automatically without the web designer having to think twice.

As far as having something else to do, generally it's looking at one or two other active Netscape windows.

I've just recently started helping with this research project at the U of MN, and I must say that the little scout robots are pretty damn sweet.

Some other cool things (that you may have read on the various sites, but I thought I'd say them anyway) and notes about them:

• They can jump with the help of a spring 'foot.'
• They each carry a black and white camera to send video to the controlling ranger.
• There's also a scout that has a camera that pops out of the tube and can pan and tilt. Check the Demos Page for videos showing this and other features.
• For testing and demonstration, the scouts can be driven around with a Palm Pilot hooked up to a transmitter.
• A PIC chip is used as the processor for the scout.
• Need to get a scout through a window? No problem. The ranger can shoot scouts through glass (see the end of the first video from the Demos Page.

With all sorts of cheap electronic gadgets available, there isn't the same incentive to play around with electronics as there has been in the past. Who needs to build a shortwave radio to find people to talk to, when the Internet has made long distance chatting trivialy easy?

Work the other interests into it. If the child reads about Tesla... build a Tesla Coil. (I guess that would need some pretty good adult supervision, too).

What you can do is find things that interest the child and see how you can work electronics into it. If they are already into computers and basic programming, microcontrollers are a good start (and a personal favorite of mine) They can teach about the basics of computers without becoming overwhelming- the Microchip PIC series, especially the 16F84 is a good choice, relatively cheap, can be programmed with *really* cheap hardware, and the development tools from Microchip are free. You as the instructor must become proficient in them first, to not add to the frustration of your student. Blinking LEDs can be an introduction to the wide world of electronics- because once they know they can do it, they will be primed to learn more. There are a number of other microcontrollers that have even simpler programming interfaces than Microchip's- such as a Basic Stamp from Parallax.

To really keep their interest, you need to have payoff early- they need to see that they can do neat stuff before the theory gets pounded into them. Thankfully, with digital electronics, you don't have to worry as much about currents and Ohm's law to get results. The math will come, but it doesn't have to be the first part. They will be much more interested in learning why something works, after seeing it work, but you have to do both- a lot of the kits that you can find just show how to build it- without giving any insight into why it works. The why may be up to you. It will require work and commitment on both the teacher's part and the student's part.

This is all well and good for a hobbyist. In fact, there's quite a bit of Linux software for doing instrumentation this way.

If your application is professional or you want dead accuracy, please do not do this. How calibrated is that input amp? How about noise? What if you have to manually calibrate?

If you must use a laptop, NI sells some cool PCMCIA cards for this sort of use. Expect to pay for it, though.

That said, I'll have to mention that any x86 is probably overkill. I'd likely use a PIC and some choice Analog Devices chips. They are dirt cheap, and would be much easier to make immune to noise. You can control the whole design, which is a good thing. The less bloat, the less chance of interference.

For extremely critical apps, you'd probably end up doing some expensive testing, but you'd end up doing that for anything off-the-shelf anyway.

As for the data transport, if the speed isn't critical, you could do this. Just pick a frequency that you can use freely and make sure that nothing else around will interfere with the signal.

There are also a lot of other ways to do it. Radio is cheap to do, and more fault-tolerant than IR.

Look around, you might find something else.

1. Probably something to deal with the fact that the board containing the LEDs for the clock is rotating, so transferring signals to them becomes kind of complicated.
2. A kind thin board featuring strips of copper and lots of holes drilled on a standardized grid. Dead handy for building circuits without using real printed circuit boards
3. No idea, sorry.
4. Likewise.
5. A PIC 16C84 is a wonderful microcontroller made by Microchip. It's a RISC design, and uses a Harvard architecture. Commonly clocked at 4 MHz, which lets it execute 1 million instructions per second. Very popular among us hobbyists because it stores its program in 1024 words of EEPROM, thus making it easily erasable and reprogrammable using just electricity--no UV light or anything like that.
6. A DIP resistor, I would guess, is another name for a "resistor network", which is just a bunch (seven or eight seems common) of resistors mounted in a standard DIP (that's dual inline package) capsule. Very handy when you need lots of same-value resistors for e.g. LED current limiting.
7. "16C84" is just a shorter form of "PIC16C84", of course. The word programmer in this context referes to a piece of specialized hardware which is used to transfer instructions into the EEPROM of the chip. The net is full of build instructions for those, and they're all pretty nice and simple.

This is stupid. Talk about overkill.

How much processing power do you need to process a few sensors and drive a couple wheels? $200 bucks worth of palm? Uhh... how about $4 bucks worth of microcontroller like a nice 16C64 from PIC Microship microcontroller? Enough to make a hardware guy cry! Onboard RAM, ROM, EEPROM.... mmmmm..... timers and interrupt vectors.... 20mA drive current.... whoo-hooo! Just need a few H-Bridges and a couple infrared leds and detectors. You wanted to know what to do with those old 5 1/4" drives... its time to get out those irons and solder suckers!

I suppose this kit is nice for those software types that don't like to touch hardware, but what fun is that?

If you REAALY dont want to challenge yourself, and don't want to touch hardware, I would think a nice Lego mindstorms kit would be better suited and more flexable. Reconfigurable, and you can always buy more motors and sensors.

A freaking GUI for a robot? And thats a selling feature? Who needs a GUI anyways ;-)

The other chip is a 93C46 serial EEPROM.
I'd imagine this is where the serial number is stored -- anyone have the equipment to read these things? I'm kinda curious what else may be on it...

There's a data sheet here.

--K
(And of course, my AC misfire gets modded up... ;)
---

Put it on a PIC! Put it on a PIC! Put it on a PIC! Put it on a PIC! Put it on a PIC! *sigh*

I design products with 8 bit devices, and I've used a couple larger chips here and there. There are many important features that designers need in microcontrollers:

• LOW COST !! 50 cents less times 10k units/month is a big deal. You could task an engineer for a couple man-months to rewrite code for a 50 cent lower CPU, after an initial release. Usually using an 8 bit chip instead of a 32 bit solution saves at least $10, sometimes much more.
• Low power consumption, under 4 mA is what I generally consider low power. A few years ago I designed a product that uses 9 A, 32 kHz CPU, wakes from full shutdown at 4 Hz. Try that with a linux-capable 32 bit chip! Low power also means a low cost power supply... at 1-2 mA, a resistor and zener diode can sometimes be used!
• Multiple vendors, or at least some assurance that the components will be available in the required quantities.
• Programmable program memory on-board... with in-circuit firmware download is a plus.
• Small physical size

Most projects in the embedded market just don't need a lot of CPU power. At high volumes, it's easy to pay even the most expensive engineers and programmers to re-write code to run on a cheaper chip.

I should probably disclose that I have a small website with 8051 related resources (open source), so take my words with a grain of salt, but until 32 bit microcontrollers are less expensive and use less power than their 8 bit competitors, I'd expect the bulk of the market will probably stay with the 8 bit chips.

First of all, I have been working on MicroMouse for the last two years, and one of the things I have been really been hunting around for is a microcontroller that has enough memory (both flash and RAM) as well as enough processing horsepower to do motor control, navigation, etc. We needed something with built-in serial interfaces, timers, interrupts, I/O ports, etc. because we wanted to make this thing as small as possible. I have looked at many different parts including PICs, the SX, various 80186 processors, the Hitachi H8 and SuperH, ARM, and many others that I'm not thinking of right now. The first thing I've learned about these devices is that the newest ones (and the ones we wanted to use) use either insanely small pin spacings (which requires a custom PCB) or they are of the BGA type (which requires a minimum 4-layer PCB with soldermask as well as special mounting equipment). We finally settled on the Hitachi H8S/2357 even though it was a 128 pin QFP with 0.5 mm pin spacing.

The things I don't like like about these newer, highly integrated processors are that they are more expensive, they tend to be a pain to mount, and chances are, you probably don't need that much processing capacity anyway. While our current versoin of the MicroMouse uses an H8 as the main processor, it also uses a couple of SXs to operate the sensor array. While these SOC's will certainly have a market, it will certainly not eliminate devices like PICs and other smaller microcontrollers from the industry.

FreeBSD is dead ...
long live the PIC assembler

It sounds like you are going way overboard with your system requirements. For 2-3 bucks you can get a RISC PROM MicroController (10 bucks for the erasable ones) These can be programmed to do a lot (it involves assembler, but thats half the fun)

I have used the PIC16C7XX series from Microchip to do a number of useful things (it has 2 I/0 ports and a A/D converter) it really wouldn't be too hard to get one of these things to control some sort of robotics.

What you do is you write your program with the free software and simulate it. (All software is free along with the specs. and manual for the PIC) After you get it right, you burn it on a chip and boom you have a chip that will do what you programmed it to!

something I've been working on... use hardware-asissted OTP. a pair of Keeloq devices, one on your serial port, one on the server's. The server's password doesn't increment until a valid password has been used.

when you SSH in the password is passed and the indexed incremented. Use the same thing for su. PAM would be wonderful here.

Any thoughts?

This is a simple EEPROM based microcontroller that runs at about 4 MHz, based on an internal oscillator. It costs less than $5 and you can build a simple programmer for it for less than $10. Yes, the TCP/IP stack IS real and has been around for quiet some time. Just do a search for "PIC microcontroller" on the web. They are made by Microchip.

You can find the code for the TCP/IP stack rather easily (sorry, no links off the top of my head), and the rest should be pretty simple. None of this is really that new.

-Andy