OSDev.org

I've decided to build my own computer either from a x86, blackfin, or arm based architecture even though I haven't decided on my processor is there any software where I can make a virtual schematic of the computer and then test it because I don't want to have to spend $60 (processor) + the cost of making a pcb or paying a company to make one and then to find out it doesn't work. Even if I do revise my test it will probably still have errors and it would be nice to be able to test it out before the building. By the way I've contacted intel and they said that they've never heard of such a thing so I've decided to ask you and if there isn't then is there anyway to be able to test the schematic out without buying a pcb and soldering the processor on.

Thanks,
Jackson

sds2017 wrote:soldering the processor on.

You really don't want to do that for your choice of CPUs. That's machine work.

If you really want to homebrew such a CPU from scratch, have it made for you, or take a breadboard and a DIP-packaged 68000/8086/6502/Z80 for starters

I wouldn't solder it on myself I would send it to a pcb and soldering company to do it for me.

sds2017 wrote:I wouldn't solder it on myself I would send it to a pcb and soldering company to do it for me.

Your choice of CPUs are quite powerful, high clock rate and not really suited to homebrewing a system. Unless you mean an M-class (Cortex-M series) ARM, in which case maybe.

But the other problem with ARM is that they don't design processors, they design systems-on-chips. Which means there's pretty much nothing to do except hook the pins up to the uart, dvi and usb connectors

When I first made my homebrew system, I started with a Z80 (specifically a H180, hitachi's sooped-up version that has a MUL instruction and runs at 6MHz instead of 4), an EEPROM and two D-A converters hooked up to an oscilloscope.

Why not try that? you can do the whole thing with breadboard yourself and probably have a much more fun experience out of it.

sds2017 wrote:is there any software where I can make a virtual schematic of the computer and then test it

I think that the software simulator for that level of simulation would be too expensive. I never heard about such software.

sds2017 wrote:because I don't want to have to spend $60 (processor) + the cost of making a pcb or paying a company to make one and then to find out it doesn't work.

That's almost inevitable. Be ready to make at least 2-3 iterations of development.
BTW, for the ARM devices there are many ready development kits, on which you may develop you ARM-based software.

JamesM wrote:When I first made my homebrew system, I started with a Z80

Yes, that's a good choice. For the DIP components you even may test schematics without soldering. Search "Wishboard" on Google. This is the sample: http://www.dirkdirk.com/projects/raiden/?p=71
I often use such board for modeling not very complicated schemes (even prototypes of commercial devices!)
For non-DIP components I solder a simple adapter which put into wishboard.

Thank you for your responses. When you said them series was a good idea for development I thought it was more for cars computers but do you think it would be reasonable to use the Avr m-0. Also I doubt Intel would do this but is it poss ible for me to get the code or soft processor of an arm or blackfin to create my own program to do this.

In terms of EDA tools, I used gEDA http://www.geda-project.org/ which is decent (but nowhere near their commercial competitor). Also it might be worth noting that at ARM we put each chip through millions of cpu-hours of testing on every iteration. Literally.

sds2017 wrote:Thank you for your responses. When you said them series was a good idea for development I thought it was more for cars computers but do you think it would be reasonable to use the Avr m-0. Also I doubt Intel would do this but is it poss ible for me to get the code or soft processor of an arm or blackfin to create my own program to do this.

Hahahahahaa. Ha. ha.

No. You'd need to licence the IP, which you can't do with Intel but ARM will let you do, at a price of only around a million pounds, +/- 10 million.

I found out that some of sparcs processors are opdn source is it possible to use that souce as a add on to a library for testing circuitry

sds2017 wrote:I found out that some of sparcs processors are opdn source is it possible to use that souce as a add on to a library for testing circuitry

The SPARC *specification* is open source. The processor isn't. And you won't have access to a process small enough to fabricate an entire SPARC onto.

So any recommendations for building a MODERN homebrew computer(no z80, 6800, ect). About the x86 computers would it be good for testing if I used a ardiuno or atmel c programmable microcontroller to send signals to the x86 to at least do simple functions to make a hello world : programming :: simple light blinking : x86. And is the atmel arm based processors good and is the blackfin processor good. If all of these are no then my last question is how do big companies start of their design do they just hire a bunch of people to make a schematic and check everything thoroughly and then by everything put it together and if it works then mass produce it and if not then keep on checking.

well, if you decide to make your own based on x86, let me know if you can produce something like the bifferboard (preferably with same functionality and price-range) but with VGA & sd mmc included - I'd buy one, or ten

sds2017 wrote:So any recommendations for building a MODERN homebrew computer(no z80, 6800, ect). About the x86 computers would it be good for testing if I used a ardiuno or atmel c programmable microcontroller to send signals to the x86 to at least do simple functions to make a hello world : programming :: simple light blinking : x86. And is the atmel arm based processors good and is the blackfin processor good. If all of these are no then my last question is how do big companies start of their design do they just hire a bunch of people to make a schematic and check everything thoroughly and then by everything put it together and if it works then mass produce it and if not then keep on checking.

Short answer? No. I have no suggestions. If you want to produce a system-on-chip (i.e. something that gets fabbed onto a die) you're looking at 1million dollars for tooling setup at least.

Why not try and walk before you can run?

Closest alternative:

Get yourself an FPGA prototype board and implement your own SoC on top of that.

Agreed with berkus, but personally I might go relays instead (or if I can get hold of them, the original electronic valves. THAT would be an adventure).