DIY Computer Science

Filling up the Screen and erasing it

2012-08-20T07:54:00.003-07:00

Next project is to get the screen to fill black with a key press. This should be a little easier. This project wasn’t easier but it also wasn’t harder. After getting the screen to fill with key press and erase with the key up I had a problem checking for memory boundaries. As I was trying to figure it out I decided to post on the forums and ask for help. This is what I typed out

I made the Fill.asm program and it will fill the screen to the bottom with black and if the key is let go it will erase all the black. The Issue is that it will keep going passed the screen memory and on erase it will go all the way passed the memory in the other direction.

[begin]

How can I check for memory boundary?

I tried

(FILL)

@Store

D=M

A=D

//here I need to check if the address is too big if so then just jump back up

@24575

D=D-A

@KEYDWN

D;JEQ

But it just goes way out of bounds.

[End]

After typing this and before submitting I reread 4.2 example and realized I was changing D and that I could read memory as much as I want as long as I don’t wright to it. So I did this.

@Store

D=M

@24575

D=D-A

@KEYDWN

D;JEQ

That way I was reading the memory and checking if it was out of bounds at the bottom

@Store

D=M

@SCREEN

D=D-A

@KEYDWN

D;JEQ

And out of bounds at the top.

So after a very long delay I finished chapter 4 still needing to do more examples because I still don’t fully understand ASM because using only 3 calls one has to make things so much simpler and the order of every last command is important.

It is now my belief that all computer science students should take a very basic ASM class after taking BASIC or PASCAL but before taking C/C++ class’s. It helps with so many different things.

Multiplying by adding

2012-08-20T07:52:00.002-07:00

When running the test script I was clearing all the locations but the test script would put in the data first before I cleared then all the results would be wrong. I just had to make the result 0 every go around but since the other two memory locations were going to be over written any way I just left them alone.

After doing this project I realized that if one plans out there project and Knows that the variable is going to be 0ver written then one does not need to clear it each time but always zero out variables that are not going to be over written but just read then written to.

The key to the multiply project is to know that when your multiplying you are adding itself to itself as many times as the multiplier says to. Also using a loops and jump statements are good.

Here was my laid out plan

//Start at fisrt memory location

//then stick that in to be computed

//then add it to itself as many times as it is supose to be multiplyed by

//use the multiplyer to controll the loop or counter

Part of the code that add'ed the data

@R1

D=M

@R2

M=D+M

@LOOP

0;JMP

After a very long break. Getting under way....

2012-08-20T07:50:00.000-07:00

On and off for five months of trying to figure out the Hack assembly language and re reading chapter 4 five or six time on top of taking pages and pages of notes I learned how to use it. Assembly language is different than any other language I have learned. Calling the memory location then putting data into that location.
Here is what I played with to figure out the ASM

@1 //this is the address to the memory where the program starts at
//And since the system just started that memory location should be blank or 0
D=M // since the above address memory location is empty then D is the value of the memory
@5 //this is a new value
D=D-A //A is = to 5 and D is equal to 0 so the new Value of D is -5
@100 //the vaule in the A register is now 100
D;JEQ // D = -5 JEQ is Jump to 100 in the program if D is equal to 0
@7 //A is now 200 if 7 then program doesnt move at all
M;JMP // Just jump to 200 in the program if 0 but if M Jump to 200 in memory also

Something like this should have put in the book with that much info on what the coded did.

Progress on creating all the hardware

2012-03-29T13:26:00.002-07:00

I am making all the schematics and taking pics of all the hardware. I am also videoing all the progress.
Next thing I will be doing is simulating all the schematics before I build the more complex parts.

Just trying to get my blog out there

2012-03-29T11:17:00.002-07:00

{EAV:a0afc2bea14f8c11}

So I want to draw as many people to my blog as I can. Why you may ask? I want to be able to get comments that will help me better understand the people who would want to buy a DIY computer and help educate people the easy and fun way

The memory map

2012-03-22T08:44:00.001-07:00

Activity 34

Creating the memory map was a little confusing because the output switching but after a little trial and error and studying the memory map in the book and arranging the chips in the order of the book I got them all to work. Now I need to figure out how to make my own keyboard and video on my own and implement them. I do have a book on creating my own video game console so I am going to combined that with what I have and it should lead me in the right direction.

Update on the 4bit ripple comparator

2012-03-20T13:49:00.002-07:00

After looking at more info on the comparator I decided to create a 1bit comparator in Logisim. (http://ozark.hendrix.edu/~burch/logisim/)
It came out great because then I found the equation for the comparator and used it to make the 1bit into a 4bit ripple comparator. Here is how it is.

The inputs are always true or on tell it is attached to another one. I am not 100% that this is the best way to do it. Let me know if there is a batter way

Logic design software

2012-03-19T18:53:00.001-07:00

A few logic design software's.

WinLogiLab

http://www.griffith.edu.au/professional-page/charles-hacker/resources/winlogilab

The logic lab
http://www.neuroproductions.be/logic-lab/

Simple solver
http://home.roadrunner.com/~ssolver/syn.html

Logisim
http://ozark.hendrix.edu/~burch/logisim/

I am going to try a few out and see which one works best for me and see if it cant speed up creating some of these more elaborate logic deign

Here is a list of more but not all of them are free.

http://www.electronicsoft.net/en-us/dept_3.html

Set less than or SLT

2012-03-19T17:04:00.001-07:00

To be more MIPS compatible I need the function STL. After a lot of research and some help from cadet1620 I found from a book that I had called "Fundamentals of Logic Design 4th edition by Charles H. Roth, Jr." that STL is part of a set called a Comparator. Also I found it on this web site http://www.electronics-tutorials.ws/combination/comb_8.html. The logic set does not match exactly between the two so I am going to have to figure out exactly how it works so I can add it to my ALU. Also I might have to add my functions to the CPU emulator to incorporate my additions. More to come as I figure it out more clearly.

After looking closer I found that it is called Digital magnitude comparators. The IC that goes with it is a TTL 7 485. The logic diagram helped me understand what was need to cascade 4bit comparators. I found the info here http://www.dauniv.ac.in/downloads/Digitalsystems_PPTs/DigDesignCh12L1.pdf

Using the software

2012-03-18T22:42:00.000-07:00

i went though the Assembler simulator and the CPU simulator and they worked like expect. I tried out all the samples and they worked great. Note I really like how the assembler simulator translate the assembly to binary now there should be one for windows it would be really cool to see the binary code that runs a basic exe file.

Update to my 16bit Adder

2012-03-18T01:55:00.001-07:00

So with a few additions to my 4bit adder with look ahead carry I believe I can now detect over flow situations. I have not tested it 100% but I did do a few test and it worked every time

Check it..
OUT out[4],carry,OvF; //added OvF as an out pin

then added
Or(a=cout12,b=cout13,out=c1,out=carry);
Xor(a=out42,b=c1,out=OvF); for the over flow detection

Now just to add it to my ALU

8bit adder

Add4(a=a[0..3],b=b[0..3],c=c,out=out[0..3],carry=carry1,OvF=OvF1);//ignore overflow in first set
Add4(a=a[4..7],b=b[4..7],c=carry1,out=out[4..7],carry=carry,OvF=OvF);

16bit adder

Add8(a=a[0..7],b=b[0..7],c=false,out=out[0..7],carry=carry1,OvF=OvF1);//ignore overflow here
Add8(a=a[8..15],b=b[8..15],c=carry1,out=out[8..15],carry=carry,OvF=OvF);//out put here

First step to optimize ALU

2012-03-18T00:56:00.001-07:00

I found a good schematic for look ahead carry adder 4 bit and added it to my 4bit adder. But before I did that I was thinking why would adding more gates make it go any faster then I thought the adding part the less gates it acutely goes through the faster and since the carry is just essentially copying the input and working on it in parallel of the actually addition I can see how it could be a lot faster. With no further ado here is the code I got from the schematic.

[code]

CHIP Add4 {
IN a[4], b[4],c;
OUT out[4],carry;

PARTS:
This is my first 4bit adder
/* HalfAdder(a=a[0],b=b[0], sum=out[0], carry=c1);
FullAdder(a=c1,b=a[1],c=b[1],sum=out[1],carry=outa3);
FullAdder(a=outa3,b=a[2],c=b[2],sum=out[2],carry=outa4);
FullAdder(a=outa4,b=a[3],c=b[3],sum=out[3],carry=outa5);*/

//A0 B0
And(a=a[0],b=b[0],out=out1);
Or(a=a[0],b=b[0],out=out2);

And(a=out2,b=c,out=out3);//needs an input
Or(a=out1,b=out3,out=out4);

Not(in=out2,out=out5);

Or(a=out5,b=out1,out=out6);

Xor(a=out6,b=c,out=out7);//needs an input
Not(in=out7,out=out[0]);

//A1 B1
And(a=a[1],b=b[1],out=out11);
Or(a=a[1],b=b[1],out=out21);

And(a=out21,b=out4,out=out31);//needs an input
Or(a=out31,b=out11,out=out41);

Not(in=out21,out=out51);

Or(a=out51,b=out11,out=out61);

Xor(a=out61,b=out4,out=out71);//needs an input
Not(in=out71,out=out[1]);

//A2 B2
And(a=a[2],b=b[2],out=out12);
Or(a=a[2],b=b[2],out=out22);

And(a=out22,b=out41,out=out32);//needs an input
Or(a=out32,b=out12,out=out42);

Not(in=out22,out=out52);

Or(a=out52,b=out12,out=out62);

Xor(a=out62,b=out41,out=out72);//needs an input
Not(in=out72,out=out[2]);

//A3 B3
And(a=a[3],b=b[3],out=out13);
Or(a=a[3],b=b[3],out=out23);

//And(a=out2,b=,out=out3);//needs an input
//Or(a=out3,b=b[0],out=out4);

Not(in=out23,out=out53);

Or(a=out53,b=out13,out=out63);

Xor(a=out63,b=out42,out=out73);//needs an input
Not(in=out73,out=out[3]);

//C1
//4 way and
And(a=out1,b=out21,out=cout1);
And(a=cout1,b=out22,out=cout2);
And(a=cout2,b=out23,out=cout3);//out
//3 way and
And(a=out11,b=out22,out=cout4);
And(a=cout4,b=out23,out=cout5);//out
//
And(a=out12,b=out23,out=cout6);//out
//4 way and
And(a=out2,b=out21,out=cout7);
And(a=cout7,b=out22,out=cout8);
And(a=cout8,b=out23,out=cout9);//out

//4 way or
Or(a=cout3,b=cout5,out=cout10);
Or(a=cout10,b=cout6,out=cout11);
Or(a=cout11,b=cout13,out=cout12);//out

And(a=cout9,b=c,out=cout13);

Or(a=cout12,b=cout13,out=carry);

}
[/code]
If any one knows how I can break this up into basic adder and faster carry sections let me know. I haven't annualized the code that close to see what would need to be done.
The more help I get the faster I can get all of this done and have a open source system that any one could build.

Update

2012-03-17T12:21:00.001-07:00

Update 4

Creation of all Logic gates with NAND IC 4011 CMOS

Gate types	Done
AND	X
OR	X
NOR	X
NAnd	X
NOT	X
XOR
XNOR

They are all done now moving on to composite gates

Update on the Switch to CPU Kits

2012-03-17T02:01:00.002-07:00

Update 3

Creation of all Logic gates with transistors

Gate types	Done
AND	X
OR	X
NOR	X
NAnd	X
NOT	X
XOR
XNOR

Again the exclusive gates seem to have more components then any others. I will try them after I complete the IC gates.

A further update on Switches to cpu's

2012-03-16T19:25:00.001-07:00

Update on step one create all basic logic gates with switches

So far I have made

Gate types	Done
AND	X
OR	X
NOR	X
NAnd	X
NOT	X
XOR
XNOR

Unable to make the Exclusive gates right now. It would take a lot of pushbutton switches and lots of fingers. I believe I could make the XOR gate with a dtsp slide switch which I do not have at this moment

I am now going to recreate these gates with transistors.

From Switches to CPU (playlist)

2012-03-16T17:15:00.001-07:00

('http://www.youtube.com/p/28A81D76E3FD63FB?version=3&hl=en_US',)

this is the video series that is going to go with it.

Update one on the up coming kits

2012-03-16T13:02:00.002-07:00

Step one crate all basic logic gates with switches.

And [X]
OR[X]
NOR[X]
NAND[ ]

I have some short videos and pics showing that the gates marked are working will upload soon.
I have all the parts to make a few kits. Hit me up if you want to donate to the project and get one.

65% done with the class next stop ALU optimization

2012-03-15T13:48:00.000-07:00

All done with chapter’s one through three. Most of the hardware is all done. The next chapter is on creating an assembler language. After completing this one I am going to go back and fix up the ALU to make sure it will be compatible with the MIPS architecture which then will be able to run some real world software and it will also be a RISC CPU which intel killed a long time ago but looks like it’s making a comeback.

After fixing up my ALU I am going to start on my kits that will revolutionize the way people understand how computers work and not be fooled by the propaganda on tv and what they say is new but really is crap. Also having a 100% open source system means you know whats in it and know that they cant be tracking you in any way because there will be no identifying items since you made them all your self.

RAM16K 16bit

2012-03-15T13:38:00.002-07:00

Activity 29:

Ram16k 16bit. All is well

An issue I found with the simulator is that the error are at the bottom and cant always see it because it’s a java window that is not that user friendly. The errors should pop up in a separate dialog box. Easy to read and refer back to.

Just had to redo some math and got this one to work.

RAM4k

2012-03-15T12:53:00.001-07:00

Activity 28:

RAM4k 16bit my prediction is that it will be just the same as the last one. Just a little copy and paste with a few changes here and there and tada the RAM4k.

RAM512 16bit

2012-03-15T12:43:00.001-07:00

Activity 27:

RAM512 16bit. Just as easy as the last ram was. Its just a combination of the last set of ram chips.

PC the rest of the story

2012-03-15T11:38:00.001-07:00

Activity 26:

Continued.

The order of things is when Mux are being used for if then else statements the last one goes first and the first one goes last.

So goes increment then load data the reset even though that seems backwards in coding since. It makes since in hardware because the last mux will change the output of all the rest.

That was the piece I was missing. It now works perfectly.

With that information I get

Increment then

Load then

reset

with the register data having a feedback loop in it.

The Beginning of the PC(Program Counter)

2012-03-15T03:21:00.001-07:00

Activity 26:

The book has the PC(program counter) as being the next activity so that is what I am going to do. With this one I have a data in and then load, inc, and reset. The inc is to change the output and the load and reset changes the input.

The thought proses.

1)if load is true then load data

2)if reset is true the rest the PC

It wont matter if load is true if rest is true because only one can excute and once because if you sent both their would have to be a rule on which one got excuted first and that may complicate things.

3)if inc is true then output gets +1;

I am using a register to hold the data the question is to tell it to rest or load new data.

I am going to work with just rest for now then move on to the next issue.

I have change it to work with load first because I think reset just works on the output.

The more I think about it the load just gets inputted into the register.

Now I am going to work with just the output.

I have it set up the way I think it should be but it’s throwing an error saying that if all is 0 but for increment I should still get a 1 the second time around. I am not fallowing this logic.

So after reading the forums and reading my old HDL code I am still confused I think I will need to draw out the PC in actual schematic or pseudo code even though the book does say this

/**

* A 16-bit counter with load and reset control bits.

* if (reset[t]==1) out[t+1] = 0

* else if (load[t]==1) out[t+1] = in[t]

* else if (inc[t]==1) out[t+1] = out[t] + 1 (integer addition)

* else out[t+1] = out[t]

I know what that means in c but I am not sure what that means in logic gates and hardware.

I belie MUX are just like if then statements but what direction should they flow because I am going to need more than one.

TO BE CONTINUED…..

RAM64 is just one step in many.

2012-03-15T01:53:00.001-07:00

Activity 25:

Create a Ram64 bit. After reading the section on it and double checking the inputs of the Ram8 with a little guessing I got it all to work out just fine. Trial and error sometimes works but is only worth trying if one is willing to go back and figure out why it work then it is all worth it.It builds on the last activity.

a RAM8 16bit

2012-03-15T00:59:00.001-07:00

Activity 24:

A16bit/8 register(memory). Not sure how this one is going to be made. After reading the chapter and rereading the chapter then I came upon the solution just by looking at what the input I have to work with and the output and figured out how to get it accomplished but using the anything’s that have multiple selectors. Another words I made it work and now onward to the next set of problem’s still remembering I need to fix my ALU so its better faster and usable in real world with real ISA’s