Hello, I have this Apple IIe that won't boot and all I'm getting is horizontal bars. I've done some troubleshooting and so far found that the clock is good and the CPU is getting a reset signal on pin 40. I scoped pins 19 and 20 on the MMU and found that only 19 (ROMEN2) is ouputting a pulse. Pin 20 (ROMEN1) is always in high (off) state. Does this mean that the problem is in the EF ROM? I haven't been able to find any information on whether both of these signals should be on at the same time or if ROMEN1 is off because it has to read the other ROM first.
I also checked the address lines on the CPU (pins 9-25) and I'm only seeing a pulse signal on the first 3 (A0-A2). The rest are all at 1.5v which I'm assuming is low state but I'm not actually sure. Shouldn't low be something close to 0v?
Is there anything else I can check to narrow down what the issue is? I have a scope but no logic analyzer.
It appears to be stuck in the lo-res screen.
Does pressing crtl-reset do anything?
Have you done the power-on-self-test?
I'm not able to get it to do anything else. Can't get it into self test mode so I assume it's not loading the ROM for it. I did scope the reset pin on the CPU and it is getting a reset signal with ctrl-apple-reset so I know the keyboard works at least. What I'm trying to figure out is why only the first 3 address lines on the CPU have any signal on them and if that has something to do with why it isn't reading the ROM or if the ROM is just bad. I ordered replacement ROMs but they're going to take a while to get here and I want to see if I can figure it out in the meantime.
I'm probably not the best one here to help you, but I think logic LOW shouldn't be 1.5v. I'd think finding out why would be your best lead.
If that can help you, here are two logic analyzer captures from the MMU at power on. The vertical line is on the rising edge of /RESET.
Notice how the Apple IIe does 3x Page-1 accesses followed by a WORD read at address $FFFC (the reset vector) and the value at that address is $FA62.
On the second image, you can see that /ROMEN1 is falling a bit later.
If you like, I can give you the data file, readeable with KingstVIS, to compare with what you see with your scope.
Why did you think your ROMs were bad?
MMU_Power_on_1.png
MMU_Power_on_2.png
Have you metered all the power supply voltages? 1.5V isn't something you should normally see on a //e motherboard. In a perfect world the power supply should put out +5, -5, +12, -12. Realistically its usually only close, but usually +/- a few tenths. More often a little low. But if logic low something is really wrong. 1.5V for high isn't much better. Start with the power supply because if it is way out of whack all else will be.
The voltages from the power supply are all good, so it isn't that. I built a basic NOP generator with an arduino to test the CPU and I'm seeing activity only on the first 2 address lines. The third one is always high and the others are all low at 0v. So I think the CPU is dead and I'm going to have to try replacing it with a working one.
I also tried building a ROM reader described in this blog post and I'm getting bad inconsistent data from both the CD and EF roms every time I try to read them, so I'm not sure if the roms are ALSO bad or if I'm doing something wrong. I'm kind of a novice when it comes to digital circuits so this troubleshooting has been a good learning experience for me.
What logic analyzer did you use to produce those graphs? Maybe I should buy one to have on hand if it isn't too expensive.
One of the most common failures on a //e that can cause all kind of wacky results is the 74LS245. If that is out then addressing will not be right. Hopefully your MMU and IOU are good as those big custom ASICs are hard to come by although people are working on replacements made with modern CPLD or FPGA devices.
You might want to get an inexpensive chip programmer like a Xgecu Pro T48 (successor to the popular Minipro TL866 family). Those are well under $100 and they can read most EPROMs and program most that require 21V or lower programming voltages, and EEPROMs, GALs, etc. I've made replacement EPROMs for //e using one of the earlier TL866. Supposedly they can also do some 74xxx testing also, but I've never used that myself.
What is generating your clock signal? The 6502 can't retain data when the clock is suspended for more than 40 microsecond; all internal registers values 'evaporates' to 0. If you use the arduino to generate the clock, it's possible that your clock frequency is too low.
I don't know exactly how you NOP generator works, but you may need to change for a faster clock (like the Apple IIe's clock directly).
Also, I think I recall reading somewhere that halting the clock for too long on the NMOS 6502 may deteriorate it's stability. I may be just misremembering, but be careful in any case.
As for the ROMs, if you followed the directives on the blog post and you get bad data from them, then I suppose that your ROMs are bad. But I'm bit surprised that both your ROMS are completely bad.
I use the Kingst LA5032. Before that, I used a Kingst LA1010 which is relatively acceptable for very basic and simple tasks. If you intent to continue playing around with electronics once you have repaired your Apple IIe, then a good logic analyzer will save you hours of frustrations. Otherwise it may not be worth it.
I was using the arduino to generate a 1mhz clock signal and verified it was correct with my scope. With the ROMs I was getting partial data from them where some pieces were being read as zeros, but in inconsistent places each time, so I'm not sure what to make of that. One ROM one was just giving me nothing every time. What are the chances that all these ICs could be dead at the same time? The dates on them are all 82 so I guess they are pretty old at this point.
Also, what kind of ROMs are these original chips? This page lists them as being EPROMs if I'm understanding it correctly, and don't EPROMs have a limited life compared to mask ROMs?
Original chips that came in //e units were mask programmed ROMs, although they (unlike ][+) have pinouts that are generally directly compatible with EPROMs so you can use EPROMs or even EEPROMs of the right size.
It is unlikely that all your ROMs are bad unless something blew them out. While they do fail occasionally their failure rate is not all that high in general. EPROMs do have more likelihood of losing programming, especially if the UV erase window is not covered and they are stored somewhere they might get UV light exposure. However even then, I have seen lots of EPROMs that are 40+ years old that still have their data intact despite the stickers covering the window having dried up and fallen off years ago.
Are you sure your reader is working right? As I previously mentioned you might consider getting a commercial chip programmer.
I guess I will have to wait and see until I get the replacement ROMs. Maybe I'll try building a 74LS245 tester in the meantime to check that chip as well.
I second this sentiment. Your ROM reader is suspect.
I doubt your ROMs are bad. It's possible the CPU is bad, but I've only ever seen that happen two times since 1980. Unless something happened like a power surge or you pulled a card out of its slot while powered up.
Pulling a card with the power on in a //e normally blows either the 74LS245 or the MMU or IOU instead of the 65(c)02. It could certainly happen but as you note, they don't seem to go bad very often. I've only seen one or two bad ones in the same time frame as well. They're pretty durable and reliable parts.
Turns out it was the CPU after all! I swapped it and everything is working fine now. Didn't need the replacement ROMs but I guess it's good to have a set on hand for future repairs.