It looks pretty solid from the picture.

https://hackaday.com/2020/11/17/another-kind-of-bare-metal-6502-computer-powers-rpn-calculator/

Nice build.
Great display, looks like a HP/Agilent/Avago/Broadcom LED. Same as I ueses on my own build, but red.
I miss a picture of the big box inside and a little more about used componenrs. 1K memory, 8x2102?

I love this build: robust, useful, and with a toggle-switch front panel!

That thing looks amazing. The main unit is wire wrapped.

You can see it here.

https://www.youtube.com/watch?v=Hxb7a6DZVzI

_________________
Cat; the other white meat.

If anybody's interested, there's a lot more information (including schematics and program listings) on his various units on his Hackaday pages.

I particularly like his toggle-switch PROM programmer, something I've been tempted to build myself from time to time (until I think again about how much work it is actually to program things with it :-)).

_________________
Curt J. Sampson - github.com/0cjs

_________________
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?

Oh wow. Acutually, another thing putting me off a bit was the work and expense to build the switches into a decent case. (Yamada-san's aluminum cases are very nice, but even much cheaper ones start around $30—click the "データシートPDF" button for more details.) So I had thought of using DIP switches, too, but after a bit of playing with them, the significant additional pain of toggling those by hand (especially now that my eyes are getting old) made that a non-starter. So fortunately I never got to discover the cycle limits issue.

I had also considered building some diode ROM boards with DIP switches, but there again expense put me off; even 32 bytes starts to get pretty awkward.

On the other hand, with a bit of extra design work it seems to me that one could "multiplex" PROM programming on to a front panel device, so you'd need to build only one big box o' switches, which might be worthwhile. I never got around to doing a front panel because it seemed too complex (I was looking at the Altair "bus-jamming" design to load the PC directly), but actually now that I've seen those videos the much simpler "just reset to load the PC" seems to be quite practical to use, so maybe I'll go back to this.


Yes, he mentioned this in one of his comments on the project page on Hackaday, I believe. If the error was a bit set that shouldn't be it wasn't an issue, you just reprogrammed it, but if the error was a bit clear that shouldn't be, well, get ready for another long programming session. Or, if you're so lucky as to be programming the 6502, some really clever patching with BRK vector routines. :-)


I thought I'd seen a lot of EEPROMs by the early '80s, but these were 27xxx series EEPROMs that still erase the whole device in one go. My impression is that it wasn't until the 28xxx series parts (with a slightly different pintout, and not available until the '90s or '00s?) that byte erase became available. And those it seems you just write like regular RAM chips (with no special voltages, either), except that you have to wait a while for the device to respond again after you've triggered a write.

Those should be programmable easily enough from your front panel, so long as you run the write signal to your "ROM" socket (and probably disable that during normal operation to prevent rogue code from corrupting your ROM and crashing your machine).

_________________
Curt J. Sampson - github.com/0cjs

_________________
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?

Hm. I've always been a bit uncertain about this. I had thought that PROM was one-time field programmable, EPROM was UV-erasable, and EEPROM was electrically erasable. What do you call the UV-erasable PROMs, and how do you distinguish them from the electrically erasable?


Oh, thanks for that link! There's a lot of good information in there. It's also reminded me about all the issues I'd forgotten.

I now realize that the panels for Yamada-san's systems above don't do auto-increment, which is something I'd be loath to live without. This can be done by giving the panel its own address register, but this introduces some UI complexity in now that you need to be able to switch the address display between the CPU and the panel.

Jamming instructions on the bus (JMP to set the address and NOP for increment/load), Altair 8800-style not only fixes that, but additionally simplifies the UI in that now you need not set the reset vector and reset before running but simply load the address at which you wish to run. Of course, since I use multiple CPUs (6800 and 6502 at the moment, and Z80 at some point) it would be nice to have a set of DIP switches to set the opcodes for these instead of having them hardwired. (I think we're starting to see why I never get anything built. :-))

I notice that the Altair 680 (6800 CPU), unlike the 8800, didn't use instruction jamming but instead went with a Yamada-style interface. I wonder if this was just because that was cheaper, or if there's some issue with instruction-jamming on 6800 and/or 6502 systems. (There's the obvious support you'd need to add to take the 6502 off the address bus on many models, but that's not an issue for the 6800, W65C02 or 6510.)

_________________
Curt J. Sampson - github.com/0cjs

_________________
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?

Ok, through more careful reading of the Winbond W27C512 datasheet, which is careful never to use "EEPROM" but instead calls it an "Electrically Erasable EPROM", and Wikipedia I now see my confusion. The "EE" versus "E" prefix refers not to whether the the device is electrically erasable or not, but whether it's erasable per byte or only as a whole. That's not exactly intuitive, but there it is!

I guess I'll use "UV EPROM" and "EE EPROM" to distinguish the UV and electrically erasable EPROMs.

I see also that some EEPROMs can be erased as a whole, as well, such as the AT28C64 (/CE low, /OE 12 V, 10 ms low pulse on /WE) and the AT28C256 (via a 6-byte command sequence).

_________________
Curt J. Sampson - github.com/0cjs

_________________
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?

Well, that was what I was originally saying, but you said that 27xxx is not EEPROM (by this I assumed you were talking about all types of them, obviously there are many 27xxx devices that are not EEPROM) and that EEPROMs are byte-erasable, which the electrically erasable 27xxx are not. So now I'm confused. Is it the byte erasability or the electrical erasability that distinguishes an EEPROM from an EPROM?

Wikipedia (which I'm not claiming is utterly trustworthy on this) seems to state that byte-erasability, or at least block erasability, is an important part of the definition; the lead paragraph says it "is a type of non-volatile memory...allowing individual bytes to be erased and reprogrammed." And later it even comes right out with, "In the industry, there is a convention to reserve the term EEPROM to byte-wise erasable memories compared to block-wise erasable flash memories."

Sorry if this is a comes across as a bit fussy and pedantic, but I do want to be communicate clearly and if my previous convention of refering to, e.g., the Winbond W27C512 as an "EEPROM" is confusing, because it can be erased only wholly, I want to avoid that.

_________________
Curt J. Sampson - github.com/0cjs

_________________
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 think I prefer to stay close to current usage, where possible. Among other things, it helps the newcomer. To my understanding, those early products, which might well have been called EEPROM when current, are not what is presently meant when someone speaks of EEPROM. Of course most people have no reason to discuss retro products, so they don't need a term for something which can be reprogrammed without UV but is not bytewise reprogrammable. I think it falls to us, the minority, to make ourselves clear. We can bemoan the loss of a useful term, if we like.)

I think the evolution might have been
ROM - read only memory, mask-programmed
PROM - programmable read only memory, field-programmed electrically
EPROM - PROM which can be erased. Initially UV-EPROM.
EEPROM(1) - EPROM which can be erased electrically
EEPROM(2) - EPROM which can be erased electrically and also rewritten byte-wise.

I found this statement: "Flash uses NAND-type memory, while EEPROM uses NOR type. Flash is block-wise erasable, while EEPROM is byte-wise erasable."

And I found this Q&A: What is the difference between Flash memory and EEPROM?