I know a few of you have Needham's Electronics programmers, so maybe there's a slight chance someone has gone before me on this and has an answer.

I'm trying to put my large look-up tables for 16-bit math (log & trig functions, multiplication, inversion, square root, etc.) into 27C801 1Mx8 EPROMs and am having problems with the Needham's EMP software. The first one was that I couldn't get the S buffer to go beyond FFFF even when I had the K buffer, L device, and T file out to FFFFF (five F's, for 1MB, instead of 64K) and a 1Mx8 device selected. I can now get it to go to FFFFF only when I try to load file 1 declared in the .INI file; but the file doesn't actually load, so then I have to click on "Load file from disc" again and it loads but still the editor still won't show more than 64K.

The next problem is that it says the Intel Hex file has a checksum error in it; so using a binary search method as I added and removed lines to keep trying, I narrowed it down to the line that gives it the problem; but when I do the checksum by hand, it's fine, just like every other line I've tried, including the one before it which the EMP software accepted.

Has anyone here tried going beyond 64Kx8 on any of Needham's programmers that use their EMP software? Even if this is a known problem with no solution, at least knowing that will keep me from wasting more time trying to do something that will never work. Needham's is out of business, so I can't ask them.

My next thing to try is to just make my own programmer, controlled by my home-made workbench computer. (The Bytek programmer I bought recently on eBay, mentioned at viewtopic.php?t=2066 , only goes up to 32Kx8.) I could also supply the data in something like a serial flash memory for the user to load into large (and faster) RAM, but I know that would be more intimidating for some users. I want to put the files on my website anyway, but I wanted to have this cross-check first to be extra sure everything is correct with the Intel Hex files first.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

got a Needham, but I never went up that far. Sorry.

Am willing to bet the 8 meg was added as more of a kludge so some editor and RAM checks dont go up there or double, which is why you are getting the checksum errors.

If you want, the next time I get to my place, I will bring my Needham home and you can drop by since my parent's place is closer to you than my place and see if we can has something out about it.

_________________
"My biggest dream in life? Building black plywood Habitrails"

Thanks Tony. After I wrote the above, I was thinking you were one of the ones who had a Needham's. I'm pretty sure Wally Daniels is another, but he doesn't check in very often. When you get that far maybe I could email you a file or two or three (at least my .INI and an Intel Hex file) and you can see if you get the same errors I do. Maybe we can try the same stuff together on the phone. I'd really like it to work but I suspect you may be right about the kludge. I also just read that some inexpensive programmers didn't do these big EPROMs right, and even damaged them sometimes.

I'm exploring some other avenues to make this available, but I'm not sure which way I'll go yet. One is the 29F080, but it's not available in a DIP so next I need to check prices and availability on adapters. I also want to make a PIC-based RAM pre-loader, but the people who are interested in that won't necessarily be the ones who want the big math tables. The RAM pre-loader sets everything up in a RAM-only system before releasing the RST line, so you don't have any slow ROM slowing you down, and your programs can also change the interrupt vectors. The info comes from a tiny 4Mx8 serial flash in SO-8, one I've already used on a 65SIB module. The larger memory would normally be for the '816, but a 6502 could address it too with artificial banking.

Get well!

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Reason I said about being closer, convalescing at my parents house after kidney removal and the upcoming chemo. I brought my computer with me but left the programmer at my place. I plan to bring it with some other things next time at my place. My parent's place is a lot closer to you in Commerce, out by the Citadel. If I remember, I think mine is an emp10 which goes through the parallel port. I also have a card one in the workshop computer, but dont use that one anymore.

the Needhams I love, they are still pretty much the home Cadillac of eprom programmers. Had tried the Wilhelm which I didnt like and got rid of pronto. Wishb their program was open source or updateable or a version hackable. Sigh...

_________________
"My biggest dream in life? Building black plywood Habitrails"

New snag, and I'm hoping someone here knows a solution.

I used a GWBASIC program to get the hex files into the PC, but that was ten years and two PCs ago. As I was getting ready to do more tables that I think are important for the website and the ROM, I tried to run the GWBASIC program, and the line OPEN "COM2:9600,N,8,1" AS #1 gives the "Device Unavailable" error message. The manual says that means the hardware is not installed (not true-- I use it regularly and it works fine outside of GWBASIC) or that I invoked GWBASIC with the /C:0 option, shutting down the port (also not true). There is also no re-directing. With everything connected as before so the PC could receive the tables formed in my HP-71 hand-held computer, GWBASIC still says the COM2 port is unavailable, even though the LEDs on the RS-232 mini-tester were all lit with the right colors.

The GWBASIC program worked years ago on previous PCs, but now I need it again and I can't find any way to make this line work on this PC with an otherwise-working COM2 port. I tried PATCHCRT.EXE on it, to make some old DOS programs work on fast PCs (it has something to do with calculating parameters for a delay-loop return and causing divide-by-zero errors on PCs running above 200MHz), and I get this screen output:

The second choice seems most likely, ie, not having exclusive read/write access, since PATCHCRT did three passes so it probably doesn't have any trouble recognizing its existence or contents. There are no "archive" or other such attributes on it.

Any ideas?

There are other ways to form the files in a PC itself of course, and there are probably other ways to get them into a PC after forming them in my HP-71, but such alternatives will require a lot more learning and progam re-writing. I really don't want to take the time to re-invent something that was already working!

If I can't find a solution soon, I might move on to the assembler macros for program structures. I could post the math-table material as-is, but I don't like the idea of having it there, unfinished, and advertising it as, "New and exciting!" I've put a lot of time into it, making me want all the more to finish it.

TIA

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Was GWBASIC open and running in another window when you tried to run that program? Windows could have had the EXE locked.

Thanks for the response. I haven't used Windows in 7 years. DOS is the only OS I've ever had on that computer. (And this one I'm on right now that I use for all the newer stuff uses only Linux.)

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I suppose it would be ok to post a preview here, but I'd prefer that you not post links to it yet, since it's not finished. I have not linked to it on my own home page or links page yet either. It's at http://wilsonminesco.com/16bitMathTables/index.html . [Edit, three weeks later: It's no longer just a preview.] It starts with an introduction to scaled-integer math and how to use it, and why floating-point is not nearly as necessary as we initially tend to think. Then there are ideas for interfacing, how to look up a value, descriptions of the various tables, rational-number approximations (some accurate to 9 or 10 digits even though they're specified in numbers of 16 bits or less), how Intel Hex files work, and how the files were calculated.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

So it's a new PC but an old version of DOS that you're booting into and an old version of GWBASIC, and both those pieces of software are known good? Any BIOS settings which might come into play?
You can't just
MODE COM2: 9600,8,N
COPY COM2: FILE.TXT
Control-c
and then let GWBASIC act on the file?

Doggone!

I looked up the COPY command and I couldn't find anything in the manual that implied it could be used with a COM port, but I thought, "Oh well, I'll just try it!" I got a similar error message. Then I wondered if something else could have gone wrong, so I connected back up to the workbench computer and went to give it a command, and realized that it was COM1, not COM2! You know how it is when you get used to something being in a particular place and you click on it without paying attention to what it says in the box anymore? I had forgotten it was now on COM1 on this computer, although I was clicking on the right thing. Anyway, it's working! Hurray! Hopefully that's the end of the problems.

To answer your question though, the BIOS date is 2001, and I got the computer, used, free, just a few years ago. The OS is DOS 6.22, and the GWBASIC is from the 1980's. The processor is a Celeron running at 902MHz.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?