All,
In the process of completing my list of parts, I am now trying to figure out what to do for the PROMs.
I found P-LAB's daughter board to use an EPROM instead. But these days it seems that even finding a UV EPROM is not that easy, and in any case they are a little inconvenient to work with.
However, finding an EEPROM is easy, and I am thinking of going in that direction. Has anyone has tried for example a 28C64 EPROM on a daughter board on their Apple 1 replicas? Is there any reason that this wouldn't be compatible with the Apple 1 circuitry?
Thanks,
-Erik
I built these adapters even though I have no problem to program bipolar PROMs. It works fine, Claudio (P-Lab) included in the circuit a jumper to switch pages on the EPROM. I used this feature to use the memory test of respected Uncle Bernie. Because bending the pins of his author's PROMs is not the best option.
03.jpg
04.jpg
Great. Did you use an UV EPROM, or an EEPROM?
-Erik
In post #1, ebruchetz wrote:
" But these days it seems that even finding a UV EPROM is not that easy, and in any case they are a little inconvenient to work with. "
" However, finding an EEPROM is easy, and I am thinking of going in that direction. "
Uncle Bernie comments:
Finding a UV EPROM is not difficult at all. For instance, Anchor Electronics in Santa Clara has them:
https://anchor-electronics.com/
If you click at "Inventory" on their website, you can download their current price list. You will find that they have almost all the ICs and IC sockets (and passive parts, too) needed to build an Apple-1, except for the DS0025 and the rare Signetics ICs (2504, 2513, 2519) and a few 74XX TTLs also are missing, but for the missing ones in their TTL list, you can use the equivalent types in their LSTTL list (i.e. 74LS00, 74LS123, 74LS154, 74LS161). I tried all of these LSTTL substitutes out, and they do work, but DO NOT substitute the 74157 with a 74LS157. They also have the elusive LM323K 5V/3A regulator for the low price of $8.95 --- these are the same Fairchild UA323K in TO-3 which came with my last kits, and they work fine. (The UA323K is Fairchild's drop-in replacement for the LM323K). No need to buy Chinese counterfeits which are re-stamped bipolar transistors in TO-3 and of course won't work. Nor do you need to buy the new LM323K still made by Texas Instruments from www. mouser.com --- they ring up a price of $65.70, each. The deal from Anchor is must better. But back to EPROMs / PROMs / EEPROMs.
Other vendors catering to hobbyists have EPROMs, too, like www.jameco.com, but I think Anchor has the better prices.
So getting EPROMs is easy. They even are reasonably cheap, if you buy from the right places. The problem with EPROMs is that once they are programmed (the super super cheap ones usually are "socket pulls", programmed) you need to erase them with UV light. A quick search on Ebay shows that you can get a new UV eraser from China for ~$30 including shipping.
Now, programming them needs another piece of equipment. But these EPROM programmers are cheap to buy, too. You just need to watch out that the programmer you buy can program the EPROMs you need to program.
EEPROMs of course are an alternative, but they may be too fast for the Apple-1, if they are of the newer types. There are some subtle timing issues with any 6502 system, when using fast memories, but the elephant in the room is the fierce kickback of the modern, fast EEPROMs into the power and ground rails. These issues can be mitigated with proper design of a small adapter PCB having local, high performance bypass capacitors, and it will work. But don't expect the use if modern, fast EEPROMs in the Apple-1 is child's play.
PROMs are very difficult. It is almost impossible to find PROM blanks at good prices anymore. What once did cost $3 each, went up t0 $10 each in the past four years, and when I looked for them a few weeks ago to get Michael Ng going with the Gen2 ACI kits, I found some IC brokers which want $70 (or more) each, for some lousy 512x4 PROM. Seems all the reasonable priced ones were sucked up, or the IC brokers just got too greedy. Same with the Signetics 2519, I was able to buy the first ones for $17 each (30% of which are bad, though) but today the sole source left on the planet wants $65, each, plus shipping from Israel. So alone the pricing thwarts any of my attempt to make more Apple-1 IC kits.
Next issue is how to program the PROMs. You need a good programmer which is able to do it properly. I use a Data I/O System 19 programmer which was painstakingly refurbished and calibrated to bring it up to snuff. Don't try to program any bipolar PROM with anything else than the old programmers of back in the day, the pin drivers of the more modern programmers typically are too weak to to a good job here, and you will get programming failures and fuse growback. But the old programmers with the beefy pin drivers are ... old ! And this means unless they are painstakingly rebuilt and recalibrated, they also fail to do a good PROM programming job in most cases. If they work at all.
I think this sums up the difficulties with sourcing and programming PROMs for the Apple-1 quite well. Stay away from Chinese sellers of PROM blanks - these typically are "refurbished" ones, meaning: pulls, and they are programmed, and hence, worthless. But many such fly-by-night sellers on Alibaba and elsewhere don't tell you that. They might even not know the fact that unlike other ICs, pulled PROMs are worthless.
- Uncle Bernie
Hi Uncle Bernie,
> Finding a UV EPROM is not difficult at all. For instance, Anchor Electronics in Santa Clara has them:
Oh, good to know. I will pay them a visit soon.
My worry was less about finding the UV EPROMS, even used, than having to get the UV eraser. But it's good to know that there are solution to both. If an EEPROM was known to work out of the box, I would go for that, for the added simplicity of not having to have an eraser (which I don't currently have).
> A quick search on Ebay shows that you can get a new UV eraser from China for ~$30 including shipping.
It turns out that they are even cheaper ($8 + $10 shipping!) if you order directly from AliExpress. They are probably of very poor quality, but if they work for the occasional erase (and don't burn down my house), that's fine with me.
> These issues can be mitigated with proper design of a small adapter PCB having local, high performance bypass capacitors, and it will work. But don't expect the use if modern, fast EEPROMs in the Apple-1 is child's play.
Good to know. I am going to try the UV EPROMs first.
You have convinced me not to even try going with PROMs. I am completely fine with this as I don't need my replica to fool anyone. It will be the kind of replica that you can build in 2024, even if that means a couple of compromises.
Regarding the other chips, I am getting the TTLs from Unicorn Electronics, and I already got from eBay:
- the DS0025 (untested, two of them in a lot)
- the 2504, 2513, and DRAM (all supposedly tested) from a single seller (not cheap!)
- the LM323K (for similar price to Anchor's) (I didn't know about the Fairchild UA323K)
For the 2519 I am going with P-LAB's daughter board for now. I already have the PCBs and the chips. I was turned off by the ~$80 price with shipping from Israel, and I had read and agreed with your comments about this in previous posts.
So I am getting there in terms of components. Most of the rest will come from Mouser/Digi-Key, and the 6502 (maybe from Anchor) and 6820 or compatible (probably from eBay as most sources except some Chinese ones appear to have dried up).
Thanks again for all the help!
I haven't had any trouble finding EPROMs either. I've bought them through Aliexpress and eBay numerous times. The only real hassle is that you need a UV eraser, but that isn't hard or expensive. I got mine off Aliexpress for around $25 including shipping. Also you need a programmer that will do them. This is a little more of an issue. A lot of the older types like 2716 and some 2732 types which are commonly used in Apple II items require 21V programming voltage. Unfortunately a lot of the inexpensive USB based programmers can't do this. I am lucky to have one of the older Minipro TL866CS programmers which can do that. The newer TL866-II+ (which is itself already discontinued) can't. Neither can the current T48 or T56 models to my knowledge. There are some other more expensive programmers or older ones that can do the older devices, but many of these require a hardware parallel port which is hard to find on modern PCs and often the software for older programmers requires older versions of Windows or even MS-DOS.
I got this very cheap Willem EPROM programmer from AliExpress along with my UV eraser and it's pretty awesome: https://www.aliexpress.com/item/1005004450760802.html
It can program 2716 chips that require 25V and I am yet to find an old EPROM that it cannot read or program. The only issue is that it needs a computer with a parallel port, but I was able to pick up an old laptop that has one locally for under $20. I installed the 32-bit version of Windows 10 and the English version of the Willem software runs fine on it.
Good to now!
I think I have an old PC with parallel port that might work. Locating the "latest" version of the software seems a little challenging but I am sure it's out there.
-Erik
Actually, there is a simple solution for erasers thanks to the pandemic. Some companies sold "UV Sterilizers" that really are actual UVC lights. As long as you use basic safety practices, these will erase EPROMs very well to the point I don't even use my actual eraser anymore. I find these at the dollar store for ,well, $1.
IMG_4571.jpg
As for the programmer voltages, the new T48 and T56 will both program 21 and 25 volt EPROMs. The older TL866CS will program to 21 volts but the TL866-II Plus will only go to 18 volts.
IMG_4570.jpg
As for the PROMs, I have been using my EPROM boards and by setting them to the F000 range I can use all 4K of space. I added a menu at F000 to select between Monitor2, Wozmon, mini assembler or Mike Willegal's memory test.
Boards.png
Very useful information. I like the idea of a more modern programmer that doesn't require an old PC to work. In addition, those are able to test 74xx family chips, and also program the ATMega644P which I might use for the keyboard.
I suspect that leaving an EPROM in the summer sun for long enough might erase it too. Good to know that cheap sources of UV work too. How long do the sterilizers take to erase an EPROM?
It all depends on the individual EPROM. I've had them erase in under a minute but some others took as much as 4-5 minutes. While I've never tried it, anecdotal evidence says the sun can erase an EPROM but it'll take a week or more. Time for a new experiment.
I actually have a T48 but can't use it yet because the Open Source minipro software is still working on support for it. I mostly only use Linux or MacOS and I don't have Windows. I might be able to use the T48 running the xgecu software under Wine but i hate doing that if I don't have to. I did test running the software that came with my original TL866CS back around 2016 but the Open Source software is so much more convenient.
I have tried only EPROMs with UV erase, 2716 for my EPROM card and replacements 2513 and 26c64 for this adapter. All kinds of different ones, branded, from Aliexpress, Soviet analogs and even a vintage Toshiba from the late 70's, all fit and work fine. So why pay more?
I now only buy them on Aliexpress, it's always a bit of a lottery but if there are problems the money is quickly refunded.
I also 2 years ago bought an AV washer on Aliexpress for 20$ with shipping, it copes with its tasks, although the quality of parts and assembly is not happy.
Screenshot_2023-12-17-12-19-55-854_ru.aliexpress.buyer_.jpg
I have the same one from AliExpress and it's great. It has a timer and I can set it in another ventilated room, so I don't have to breathe the ozone it produces. Also erase times depend on the EPROM type. About 5 minutes is enough for most western chips, however the old Soviet ones can take as long as 20 minutes, sometimes more.
Here is the software I use for the programmer in the link above, along with the jumper diagram: WillemProgrammer.zip
In post #10, ebruchetz wrote:
" I suspect that leaving an EPROM in the summer sun for long enough might erase it too. "
Uncle Bernie comments / warns:
It is true that the sun (and daylight in general) can erase EPROMs. This is how my Apple-1 color graphics card became "defective". I had put no protection label over the window of the video EPROM (these labels are scarce nowadays) and the color graphics card was laying around in my living room for a year or so, not being used anymore. I only have skylights. Normal glass should not let UV light through but it's bulletproof glass (we have >= 1 inch hail all too often) so it may be different than normal window glass. Still, when I then started to work with the card again, it did not function anymore. It turned out that the video EPROM only contained a sea of useless 'FF' and 'FE' bytes, and a few with a leftover '0' bit elsewhere. It's quite stunning that daylight, through windows, is able to erase EPROMs. Buy maybe the conspiracy theory is right which says that LED light bulbs were mandated to make us all blind. So the medical-industrial-complex could take away our life savings. Same clever strategem as with the cancer and diabetes causing industrial "food". But alas, I lack the lab equipment to measure the UV spectrum and intensity of these LED lights. For reading and drawing schematics etc. I still use incandescent light bulbs I have stockpiled.
The takeaway from this information is:
once the contents of EPROMs (or UV erasable ICs in general) has stabilized / is fully developed, put one of these UV protection stickers on it. The real deal was a sandwich of plastic foil to write on, aluminum foil to block light 100%, and self adhesive glue. Every IC programming center in the 1980s and 1990s used millions of these stickers per year. Now they are all but extinct, along with the programming centers. Not sure if paper stickers offer full protection - they are translucent. Oh, and last but not least, always protect your eyes from UV light, religiously wear high quality sunglasses when being outside in the sun, expecially when living or hiking / skiing at higher altitudes. We have lots of older people with cataracts here in Colorado Springs. The UV erasers / sterilizing lamps also are very, very dangerous, avoid exposure to their light (except exposure of the devices to be erased). These UV sources can turn you blind of you look into their light. (Hope the LED light bulb conspiracy theory is untrue, otherwise we would need to wear sunglasses even at home).
- Uncle Bernie
I make my own circular stickers using paper backed aluminum tape and a 10 or 12 mm leather hole punch like this one: https://www.aliexpress.com/item/1005001834155956.html
They look pretty good aesthetically and you can write on them with a pen:
Stickers.png
Great, thank you!
Here are pictures of my EPROM daughter board (thanks P-LAB!). I opted for a black PCB instead of green, for a change:
IMG_1375.jpg
IMG_1376.jpg
The board looks great, is it a matte mask or just a photo without glare?
Just in case I have to tell you that any dust or fingerprints can be seen on the black mask much better than on the classic green surface. A friend of mine built a replica of Apple II rev0 on a board with a black mask, so called Dart Vader edition, and faced this problem.
Glad you like it. I didn't specifically order a matte black, just "black" from JLBPCB. It is not super glossy but also not really matte. This is just an experiment to see how it would look. On that small a board I don't think fingerprints/dust will be a problem, but it's good to know it can be a consideration.
I finally decided to order the cheap UV EPROM eraser, and the XGecu T48 programmer.
I haven't had a chance to try the eraser, but I did use the programmer. I normally use MacOS, and the good news is that there is a command-line open source program called minipro that now works with the T48. This allowed me to program the 27C64 in no time without using clunky software. So I now have Wozmon on my EPROM, ready to be plugged into the PROM sockets.
I realize that it would be a good idea to have a ZIF socket on that board, and I might assemble a new EPROM board with that. But for now that will do. I have yet to assemble the main board of the Apple 1 but I now have all the components.
IMG_1592.jpg
IMG_1593.jpg
I hadn't heard the T48 support for the Open Source minipro was available now. It wasn't just a few weeks ago. This is good news.
Indeed. I had to build it from source, but that was quite easy!