New Programmer baby born at my house...

[Foster]

When I uised that software I left the file offset at 0000 and the changed device to C000
That way the file would load into the buffer you see on screen starting at 0000, but would be placed in the device at C000
Now you know why I like the SST27SF512's if you made a mistake with them you can re-use them immediately

[uniman]

My 2-cents.  I have GQ   rev 1.04

Unplug your USB.
Open GQ twice.
In one window set Device to 27c512.
Then fill 0000 to FFFF with 00

In the second window open your IGT eprom file.
Put your cursor at the top left and highlight the entire file, (left click, drag to end of file) and copy.

Paste at location C000 in the first window.
Save it.

Close the second window.

Plug in your USB.
Select 27C512 again.
Insert blank eprom, ID it, read it.
Reopen saved file and write.

[stayouttadabunker]

When I used that software I left the file offset at 0000 and the changed device to C000
That way the file would load into the buffer you see on screen starting at 0000, but would be placed in the device at C000
Now you know why I like the SST27SF512's if you made a mistake with them you can re-use them immediately

Yes,
I'm going shopping...I am running out of chips here....
I should sell them on ebay..."AS IS"...  

Really, I need to make an UV eraser...

I entered "C000" into the 'Device" offset box instead of "0xC000".
The code memory has data starting from "00000000" and all data ends at line "00001FF0".
All the lines after that show "FF".

I noticed on line "00001F50" the percentage of the file is shown as 85.02%.
That's a tough reel chip that I like to play with!
On line "00001F30" there are some letters and numbers that say "S+ 4667".

[stayouttadabunker]

My 2-cents.  I have GQ   rev 1.04

Unplug your USB.
Open GQ twice.
In one window set Device to 27c512.
Then fill 0000 to FFFF with 00

In the second window open your IGT eprom file.
Put your cursor at the top left and highlight the entire file, (left click, drag to end of file) and copy.

Paste at location C000 in the first window.
Save it.

I'd love to try that .
I have the first line "00000000"
But my last line is :"0000FFF0"
I don't have "FFFF"

ARGH!!!

[Foster]

the trick is FFF0 to FFFF is the last line
Programmer software and HxD when dealing with 8 bit (1 byte) width data displays 16 bytes per line
FFFF is the last byte in the row.

Address                                                                 
0000FFF0  FF FF FF FF FF FF FF FF--FF FF FF FF FF FF FF FF ................

Address      0  1  2   3  4  5  6  7    8  9 A  B   C   D  E  F
0000FFF?  FF FF FF FF FF FF FF FF--FF FF FF FF FF FF FF FF ................
replace ? with 0 to F

[StatFreak]

..
I believe that what you need to do is to put 0000 in the file offset box, and c000 in the device offset box.
...

When I uised that software I left the file offset at 0000 and the changed device to C000
That way the file would load into the buffer you see on screen starting at 0000, but would be placed in the device at C000
Now you know why I like the SST27SF512's if you made a mistake with them you can re-use them immediately

Foster is confirming my suggestion. If you load the file and burn the chip with "0000" in the file offset and "C000" in the device offset, it should work.

[uniman]

Like Foster said,
Yes you do.
You only see the address of the beginning of each line.


So the last line is really;
  FFF0  FFF1  FFF2  FFF3  FFF4  FFF5  FFF6  FFF7  FFF8  FFF9  FFFA  FFFB  FFFC  FFFD  FFFE  FFFF

[stayouttadabunker]

That's great Foster,
Man....... where were you when I burned all these blanks and none of them work? lol
I kept on getting "62-1"....argh....I was starting to wonder if it was the reel jumper on the 10MHz boards.
But I normally use my Double Cherry Bar woth a 16MHz and there's no reel or game jumpers.
Anyways, before I make another attempt ( I sound like that dumb magician now where nothing worked...lol)
Are there any other fine points I should be aware of?
JDKMunch is eagerly following this thread because he has a GQ-4X coming in the mail as well.
If it's better for you, please send me an email instead...
Again, thank to you and everyone else for all the help you've given me on this....I'm sure it'll benefit future NLG members.

[StatFreak]

Like Foster said,
Yes you do.
You only see the address of the beginning of each line.


So the last line is really;
  FFF0  FFF1  FFF2  FFF3  FFF4  FFF5  FFF6  FFF7  FFF8  FFF9  FFFA  FFFB  FFFC  FFFD  FFFE  FFFF

Yup.  

Remember, you're counting from 1 to 10 on each line, only in base 16! Just think of each line ID as increasing by 10 with the single digits in the row from left to right.
FF00     0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
FF10  ...

--------------------------

Now, if you really want to get confused, , each single hex digit represents a four-digit binary number:
0h  =  0000 b
1 h  =  0001 b
2 h  =  0010 b
3 h  =  0011 b
4 h  =  0100 b
5 h  =  0101 b
6 h  =  0110 b
7 h  =  0111 b
8 h  =  1000 b
9 h  =  1001 b
A h  =  1010 b
B h  =  1011 b
C h  =  1100 b
D h  =  1101 b
E h  =  1110 b
F h  =  1111 b

10 h =  0001  0000 b
1F h =  0001  1111 b  (notice how each hex digit represents it's own chunk of four binary digits.

So   C000 h  =   1100  0000  0000  0000 b
and  FFF0 h  =   1111  1111  1111  0000 b
and  FFFF h  =   1111  1111  1111  1111 b

Actually, hex makes it really easy to convert to binary, which is why they use it.

[stayouttadabunker]

Uniman,
I agree with you and thanks...
my last line looks like this though:

0000FFF0 FF FF FF FF FF FF FF FF - FF FF FF FF FF FF FF FF

8 columns of "FF's" separated by "-" then 8 more column's of "FF's".

Weird stuff, what friggin' lanquage is this stuff written in ...pig latin?
It's ridiculous looking , undecipherable garbage.
The only thing I see on the right hand side of the box is very little type written stuff.
I make out IGT near the top and I see the reel # near the bottom.
The rest is garbage....I need a pig latin" descrambler...again...lol

[stayouttadabunker]

[Remember, you're counting from 1 to 10 on each line, only in base 16! Just think of each line ID as increasing by 10 with the single digits in the row from left to right.
FF00     0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
FF10  ...]

Oh!!!! I GET IT NOW!!!!

jeesh, this isnt easy....
I'm learning a lot though, awesome stuff really...nothing I would try to woo a woman with though. Good thing I'm already married...lol...
What kind of books can I read to figure out what all that junk means though?
I getting kinda interested about it...I've just opened up a few more doors to my inquisitive mind (as usual).
Or is it all taboo stuff and just worry about making back ups of my collection like I wanted to do in the first place?

[uniman]

Here is link I found helpful when I got started;

http://www.arlabs.com/help.htm

[StatFreak]

Remember also that the value that you see in each position is just that: a VALUE. (Not the address.) So that row of FF FF FF... means that the value of FF is stored in each of those locations.

Think of the numbers on the side and across the top as the addresses of houses on a street, and the 2-digit numbers that you read across the line as what's IN the houses.

          0     1     2     3     4     5     6     7         8     9     A     B     C     D     E     F
FFE0   FF    FF    FF   FF    FF    FF    FF    FF  -   FF     FF    FF    FF    FF    FF    FF    FF     . . . . . . . .  . . . . . . . .
FFF0   FF    FF    FF   FF    FF    FF    FF    FF  -   FF     FF    DA   1D   FF    FF    44    61    . . . . . . . .  . . . . . . Da

The green numbers are the addresses of the "houses", the black numbers are what's stored in each "house".
In this example, the memory address of FFFAh has the data value of  DAh stored in it, and the memory address of FFFBh has the data value of 1Dh stored there. These values are often written as 0xDA and 0x1D

The dots and text that you see on the right represent the ASCII value of the hex numbers stored in each address. If there is no ASCII equivalent of the hex number then a dot is printed. Otherwise the ASCII text value is displayed.

In this example, the value stored in address FFFEh = 44h.  44h = 68 decimal = the ASCII character "D".
                      the value stored in address  FFFFh = 61h.  61h = 97 decimal = the ASCII character "a".   So you see Da on the right for those addreses.

If the readable data on the right looks like garbage, it's because the binary data is actually machine code and just happens to correspond to random ASCII values.
If the readable data on the right actually says something in English, it's because the binary code is being used to represent characters that will be displayed or printed out for the end user at some point when the program runs.


<EDIT> corrected and added more detail to the explanation.

[uniman]

And, when you erase an eprom it is actually filled with 1's. 1111 1111   1111 1111 etc
In Hexadecimal, these ones are represented by....              F F             F F

When you write to an eprom, you pushing out selected 1's.
So it ends up looking like...............................................    1f               a4             14         etc
In binary .................................................................0001 1111   1010 0100   0001 0100

[StatFreak]

I believe it was Jay who explained it to me. (CRS ) When you 'erase' a UV EEPROM, you are actually melting a substance that flows and covers all of the 'contacts' on the chip. The substance conducts electricity, so every connection is turned 'on' with a value of 1.

When you burn a chip, you are literally burning away this material at each node that you want to be 'off' (a value of zero). Once burned away, that memory node can't be reset to a 1 without melting the substance again under a UV light. Eventually, there comes a time when there isn't enough of the substance left (or it doesn't flow properly anymore ??) and the chip can't be reprogrammed again.

[stayouttadabunker]

It is somewhat similiar to the days when I programmed house alarms...I could give a ASCII text value to the side so I could use it as reference to what I programmed in. It helped me decipher what I did so could have something to fall back on when I came back to it at another time. Too many zero's and ones could drive your eyesight batty!

Uniman's link is something I seen before, not all of it is entirely new to me, but I just havent played with it enough.
It's so much more fun to understand everything that goes on with a machine than, I find, than just dropping in a few coins - I like to see what makes it tick....reminds me when I was a budding teenager, and of stunningly beautiful girl that asked me to fix her broken radio I later blew up...

[stayouttadabunker]

That's must be why the UV erasable-chips can only be erased about 25 times...I saw this somewhere a few day ago. ( Also a form of CRS? )

[Foster]

I think the CD and DVD RW discs have about the same number of erase rewrites as well, maybe more but I wouldn't chance it when they are so cheap.


 

[Op-Bell]

I think I can answer two questions here.

First, the main one. Think of an EPROM as a notebook. A 27C64 has one page, and the larger 27C256 has 8 pages. The address lines tell the EPROM which page you're looking at. Since the 27C64 only has one page, it doesn't have enough addresses for eight - those pins are still there, but used for other things and tied to fixed voltages. When you drop in a 27C256, those fixed voltages select one of its 8 pages, which is probably not the one you wrote on. So the trick is to use the burner software to copy the same data to all eight pages, at 0000, 2000, 4000, 6000, 8000, A000, C000 and E000. Then it doesn't matter which page gets selected, you still read the same information. Load the 27C64 one-page data - it goes in at 0000 - then copy it to all the others one at a time, then burn the 27C512.

Second, how do EPROMs work. First I have to tell you how MOSFETs work, because an EPROM is just thousands of MOSFETs in a matrix. A MOSFET has a layer of semiconductor (silicon) with two end terminals, and a metal gate electrode on top insulated from the rest by an infinitesimally thin layer of glass (silicon dioxide). Normally current will not flow between the ends because the silicon behaves like two diodes back to back. When you apply a voltage to the gate, the electric field attracts electrons into the area under the gate oxide. This makes a conducting path between the two ends, so current can pass and the MOSFET turns on. Hence Metal-Oxide-Silicon-Field-Effect-Transistor (MOSFET).

Now the trick with EPROMs is that under the gate metal, there's a "buried gate", another piece of metal completely enclosed in glass and not connected to anything. When a bit in the EPROM has its MOSFET turned off, it reads as a 1. When you program the EPROM you apply a high voltage to the gate that breaks down the insulation so that current flows through the gate oxide, and through the buried gate, into the silicon. When you remove the voltage the current stops, but some charge carriers (ions, not electrons, but let's not go there) were only half way across, and they get trapped in the buried gate and can't get away. There they sit, attracting electrons into the area under the gate oxide, so the MOSFET stays turned on for ever. Now this bit reads as a 0, not a 1. You can't "unprogram" this bit electrically with an EPROM (you can with an EEPROM, but that's another story). So you can program more 1 bits to 0, but you can't program a 0 back to a 1.

To erase an EPROM you shine energetic UV light on it. This causes a photoelectric effect, stimulating the charge carriers in the buried gates until they have enough energy to jump through the gate oxide and leak away. When they're all gone your EPROM is erased and you can program it again. There is a lifetime effect, because the programming process leaves a few free ions scattered through the molecular structure of the glass insulator, and after a while it doesn't insulate so well and the charge trapped on the buried gate leaks away, but it's at least 1000 program/erase cycles. A worn out chip can be recovered by annealing it in a hot oven to restore the crystal structure, but this is a laboratory experiment, not something to try at home (it destroys the package).

Hope this helps.

[StatFreak]

Op-Bell, I'm going to have to speak to someone in charge around here about letting me give you extra karmas.  

Thanks again.

[Foster]

I thought Global Mods could do anything they wanted!!!

OR they should be allowed to do so!!!

[StatFreak]

I thought Global Mods could do anything they wanted!!!

OR they should be allowed to do so!!!

Only one K in 12 hours - that goes for everyone. Only the admins can manually edit the numbers.

[jdkmunch]

That's great Foster,
Man....... where were you when I burned all these blanks and none of them work? lol
I kept on getting "62-1"....argh....I was starting to wonder if it was the reel jumper on the 10MHz boards.
But I normally use my Double Cherry Bar woth a 16MHz and there's no reel or game jumpers.
Anyways, before I make another attempt ( I sound like that dumb magician now where nothing worked...lol)
Are there any other fine points I should be aware of?
JDKMunch is eagerly following this thread because he has a GQ-4X coming in the mail as well.
If it's better for you, please send me an email instead...
Again, thank to you and everyone else for all the help you've given me on this....I'm sure it'll benefit future NLG members.

I am .... this is really eerie!  It's like looking at my future.   Were you able to burn a working chip?

[uniman]

That is the most complete explanation on eproms I've ever seen. Thanks Op-Bell.    K+

Copying the program to "every page" is what I do to my Universal eproms. Old Universals use a 24-pin 2732 game(reel) eprom in a 28-pin socket. Instead of looking for 2732's, I use 28-pin 27c128's, copy the file four times and then save it as "game#-27c128".   

[Ron (r273)]

Foster,
Can I send you a copy of my GQ-4X software?
I need to know what to type into the dialog box.
It's too hard to explain but easier for you to see if you peruse the software yourself.
You'll see what I mean.
I've sent this software to Statfreak and r273...no one's been able to figure out how to burn a copy and get it to work without getting that dreadful "62-1" error code on my S+.
What is this box for ?

I will throw this out to you.

Is the "Fill Buffer With Data" reached by clicking on the A-B icon on the top of your window? If so mine works similar to that. I only change the "to address" to 0xC000 and the "Data" to 0x2000 when using the 27C512.

Ron