Hi,
I started an initiative few years ago to teach ourselves about computers, it was called msxmakers because was initially focused on msx.
our project try to teach ourselves how to build new hardware.
So, my idea is not focus on repairing old hardware but build it, some times using as older chips as Woz, sometimes adding new chips as long as we can study them and understand it, but new pcb designs and with gpl schematics when possible.
I was not looking to focus only on msx but I don't like names using "retro" and I though some people won't mind that.
I was confused, but I will create a new logo to fix this, called DIY computers as brand.
We use/develop new cases, new power supply and / or new card devices, buying old hardware do not give us any knowledge.
This is my new Apple II wooden case, please, don't look at my MSX logo and think I cannot pay an old US case, or I don't like to break a computer to build another.
Now is painted (an glued).
I used a rev 0 mainboard, charROM pcb replacement, Eprom pcb replacement, raspberry pico vga, Brainboard, etc.
DAM ][ disk board is not full working perhaps because it needs prodos and I still have not Languaje card.
I preffeer to build an SRAM, through hole, gpl design, but I might need to design it by myself.
Even I know electronics, I have not too munch knowledge on 6502 or Apple II general use. I started to read several good books fount on the net.
Thank you all.
In post #1, diycomputers wrote:
" I prefer to build an SRAM . . . design ..."
Uncle Bernie comments:
Be advised that the Apple II system architecture is not very well suited to use SRAMs, especially fast modern SRAMs. It just won't work !
The reason is that there is a flaw in the 6502 bus cycle timing approach, which did not matter in Y1975 when the 6502 came out, because the memory IC at the time were too slow to notice the problem. Only ten years later, in Y1985, both RAM and ROM memories had become fast enough to notice the problem: for instance, you can't gate write cycles off fast enough to prevent the modern fast SRAM to "see" the crumbling addresses made by the 6502. Which means that in some SRAM architectures, those with an "address change detect" circuit, the crumbling addresses will initiate another internal write operation, this time to the crumbled address, which trashes the byte there, eventually leading to a program crash.
I think that Woz knew about this trap and this is why both the Apple-1 and the Apple II ignore the "official" PHI1 and PHI2 signals offered by the 6502, and make "fake" PHI1 and PHI2 signals. As the story goes, when Chuck Peddle visited the now famous Garage, he (Peddle) got furious over this Woz trick and accused him (Woz) of having no clue about proper 6502 system design. The truth is that Woz' solution makes the Apple-1 and Apple II bus timing slightly more robust against delays in the TTL decoding and memory gating logic. Typically, the chip selects are gated off by the "fake" PHI2 a bit earlier than the "real" PHI2 from the 6502 could do, before the addresses crumble and may wreak havoc.
Now, with DRAM which use /RAS and /CAS signals to store the addresses internally, this issue does not manifest itself.
But when using the 6502 with modern, fast SRAMs, then you need some sophisticated timing state machine to generate "safe" SRAM control signals.
I have been working on a fix of this issue which affects any 6502 system for a while now, and the added circuitry needed as a solution comprises the "CPU island" in my "Replica 2e" project, see post #23 in this thread:
https://www.applefritter.com/content/uncle-bernies-replica-2e-ww-prototype-apple-iie-replica
At the moment I run the system without the added circuitry being activated. It's too soon in the development to add another layer of complications. But once these added circuits have been brought on line, it should be possible to add a fast, modern SRAM without much added circuit effort.
Note that said timing issue also affects the "Soft Switches" in the Apple II. If the soft switches are implemented with fast CPLDs or FPGAs, then the gating of the soft switch latch enables requires some extra effort to make them work. They must be gated off before the addresses start to crumble. And depending on which signals are available, this may not be trivial.
I have not yet investigated the various CMOS versions of the 6502 if their designers did something to mitigate the problem with the address hold time at the PHI2 cycle end. If they didn't add circuitry to increase the address hold time, then with the faster CMOS processes these 65C02 will be even more tricky to make them work together with modern, fast SRAMs.
It will be fun to try that out ! (But I'm not yet there ... development plan must be followed).
The bottom line is that I don't recommend SRAM based implementations of the 6502 at the moment. The problem / timing trap mentioned above is real. Once I have solved it I will publish the solution and then anyone could replace the DRAM in the Apple 1 and Apple II with modern, fast SRAMs.
Other than that, you have built a very beautiful wooden Apple II case. And then you make a photo with a slanted keyboard in it. Oh the horror ! This was a mistake which was made all too often back in the day when hobbyists had to make their own enclosures using primitive hand tools from the household's tool box, and ill-fitting keyboard cutouts were the result. Your keyboard cutout looks perfect (CNC machined ?) so why do you show us a photo with an imperfectly mounted keyboard ? This flaw diminishes the otherwise great impression your work could make !
Do you intend to do a production run of these enclosures ?
- Uncle Bernie
I'm glad to see you in this forum . ( Me alegro verte en este foro ) .
This man is one of the people who helped me launch my Apple II clone ... https://www.applefritter.com/content/restoration-apple-ii-clone
And today it is helping me to build a Apple II Replica 0 ( www.willegal.net )
Replica_180622.jpg
Replica_180757 copia.jpg
Replica_183048 copia.jpg
Replica_183051 copia.jpg
Replica_194410 copia.jpg
Replica_194533.jpg
Replica_190648.jpg
Replica_182328.jpg
I took the photos when we met at another friend's place to tinker with the computers.
He was the first of my friends who has set up an Apple II Rev0 Replica, it has been easier for me to build mine, I have been following his steps and asking my doubts.
https://www.applefritter.com/content/construction-my-apple-ii-rev-0-replica
Thanks Uncle Bernie, all your suggestions are welcome.I have seen several language card designs for Apple II and I thought they were well proven, but if not, I will wait for you to complete your development.I was currently working on a card version based on the design here:https://github.com/garrettsworkshop/RAM128Although redesigning the card for through hole components. The main IC would be replaced with an AS6C4008-55 series IC which is probably older than the one used by Garretts but perhaps still too modern/fast.My purpose when looking for an SRAM card is twofold, first to avoid the ribbon cable connection to the motherboard, and second to experiment if possible with step-by-step circuits.Although I have built other computers before, I am still a novice with Apple II or even any 6502 based computer.In the work team we are starting to build our own circuits, it would be good to know the details of this operation in case it affects circuits other than memory expansions.Regarding my case and its keyboard:The photo showing the misplaced keyboard is because I cheated. I decided to take a first photograph when I had not yet fixed the keyboard to the case.My case is currently painted but not finished, I'll better post pictures when the painting is finished.In the workgroup we are in, we are also already building some 3D printing casings.My only intention with respect to the design was to have a light, comfortable and economical case to contain my Apple II.The design was created to be shared, I can help whoever needs it to have their wooden casing in Spain, but I don't think it helps anyone to send cut wood abroad with all the extra costs and risks that this entails. Here you are on it's current status. Feel free to erase the MSXmakers logo.
-DIYcomputers
There are language card designs that don't use the ribbon cable. They generate their own CAS / RAS signals to refresh the onboard memory.
Plus they have the added advantage of being able to be used in any slot and in addition to a "strapped" 16K card in slot zero.
For instance:
16kcard.jpeg
I used to use such an arrangement, with a regular 16K card in slot 0 and a card such as this in slot 4, for a total of 80K, so that I could copy disks using Locksmith 6.0 in two passes instead of three, which saved quite some time when I was doing multiple copies.
Good to know it!
But, Do you have the schematic for this design? Even I could I don't like to buy cards but to build them.
Hi, not sure if this is what you are looking for but there is a project. I wanted to build it after I build my Apple II Rev. 0, but since I never got it to work I put it on hold.
https://github.com/btb/LanguageCard
Good for you, but this project is also based on SRAM chips.
I have never seen a schematic for this type of language card. I do have one or two of them somewhere, but haven't really dug into them.
Presumably it's quite similar to a normal 16K card, but with a couple of extra chips to handle RAS/CAS
This card was probably designed by Unitron in Taiwan, and information on Asian clone makers' cards are almost impossible to find.
Here's the Unitron card. (They also made a standard version of the card with a strap to the last RAM chip on the motherboard. Why did they have both? No idea...)
UNITRON_16KRAMCARD.jpg
U2076_Z80_BOTTOM.jpg
Hi Sir,
Look at your pictures, I guess they doesn't fit.
Could you check if front or back are mistaken?
Thanks for sharing.
I don't see anything wrong with the pictures.
If you're not seeing the entire photo, right click on it and view it in a new tab in your browser, or right click on it, download it and view it in a jpg viewing utility.
They are high resolution images.
The Garrett's Workshop RAM128 Saturn is a solid card. I have GW memory cards in 2 of my machines so far. More to come.
Hi,
Your back board picture is called "U2076_Z80_BOTTOM_0.jpg"
What I saw was the 40DIP600 footprint. Sorry but even downloading your pictures I didn't see they fit.
Is applefritter cheating me?
regards.
I don't know...try using a different browser. It looks fine with Safari, Chrome and Firefox on my Mac, and looks fine with Firefox and Microsoft Edge on my Windows 10 PC