Ah, you beat me to it. I had actually simulated a while back but hadn't gotten around to publishing the results. I ran the simulation a little differently, with the clock continuously rising and falling, the simulation acting as the real circuit would behave. It looked fine to me, so I think the GAL can be ruled out as a suspect.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

The schematic appears okay to me based upon the connections that are shown. What you don't show and may be a factor in whether or not the circuit will work is your Ø2 clock generator.

Also, are you sure your battery can produce enough current flow to support all loads plus the losses in the 7805 regulator? I assume you know the 7805 is not a low-dropout type, so the battery voltage should be at least 6 volts under load to assure regulation.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

noted and stored


I did not searched for that nor did I stumbled across one within the forum. It's just a little elaborated auxiliary schematic. If I were in need of s.th. similar I guess I would just use a free gate, two resistors and two LEDs maximum.


edit(1):

not intentionally !

Just a few days ago I won a fight with this damned software (after several hours) just because a non sequential pin definition !!
So I am currently well exercised with the odds using this program

BDD can correct me if I'm remembering incorrectly, but I seem to remember him saying earlier that it's not just a matter of not being able to tell what colors things are, but that the light green lines nearly disappear into the white background.

_________________
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?

Correct.

The particular color-blindness I have is tritanomaly. The principle effects are an inability to positively distinguish blue from green, and great difficulty in seeing either of those colors on a white or near-white background. I also cannot see pure blue on a black background. Perversely, I can readily see those colors on a yellow background, although blue and green still look the same.

Due to this little problem, I may misread the color code on resistors (I keep the DVM handy to verify what I think I'm seeing), and sometimes mix up pairs in network cable, which some years ago led to me purchasing a rather costly piece of test gear to verify connections and save myself a lot of grief. Needless to say, that piece of equipment has gotten a lot of use.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

$1.37? Down from $28.58? I would buy at least four. When you can't find it when you need it, find one of the other ones.. keep them here and there. And a spare. Or at home or elsewhere.
(BTW they (the EISTAR LP-1) start at a bit above US $8 on ebay)

I do that with tape measures. I'm always misplacing them.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

I'm using a 3-cell LiFe as my battery, so it's nominally 9.9V. It's a 2100mAh, and supposedly 30C discharge(2.1mAh x 30C = 63A). I don't believe that this battery would survive such a discharge(but it is a LiFe, so it's not inconceivable), but I'm not drawing anywhere near that for this project. It draws ~265mA with the keyboard and display attached, without the regulator(180mA without them).
The keyboard is an RC2014 serial keyboard, and uses an ATMega running at 16MHz. The display is a SparkFun serial graphic LCD, and it has a PWM backlight, which was on at about 50%, I think.

I think I'd like to swap out the linear regulator for a switching type, so it's more efficient, as I plan to run the "finished" version on a battery. One of my goals with this is portability, although it's looking like I won't be able to put it in my pocket, the way I'm going.

The Ø2 generator is on the schematic. It's in between the 65SPI and the RAM. I didn't label it as such, so that's probably what tripped you up. Both of the oscillators(baudrate and Ø2) are MXO45 can oscillators.

I suspect that I won't lose the logic probe. I tend to keep my tools in boxes so I don't lose them. My DMM and its leads are in one such box, and most of my hand tools are in another one.

I've attached a picture of the box I'm using for my tools, the mounting for the display and keyboard, and the battery I'm using.

This is only vaguely related to the original topic, but anyway.
As I alluded to in another thread, I'm suspicious of the wiring under and around the 65SPI(It's a right mess in there!). Because of this, I'm starting to think seriously about getting a PCB made, as a second prototype. The end result of this is supposed to be a handheld, so I'll need to squeeze it down a bit for that, but I don't want to be poking around a cramped board for now. I'm putting together a schematic in KiCAD, and I'm almost done, but I have a couple of questions.

I'm not using all the input and output pins on the 28L92. I know I can safely leave the outputs unconnected, but what should I do with the unused input pins, and should I break out the unused outputs?
I'm also not currently using the 65SPIs EXTC(External Clock) input. Should I wire that to Phase 2 or the DUART's clock, or tie it to GND or VCC?

I've attached the schematic as it currently sits, in black and white for those of us who have eyeball issues.

The unused general purpose inputs (GPIs) can be safely left floating. However, maybe you should bring them out to a header and pull them up to Vcc with a resistor network, which is small and inexpensive. That way the GPIs are usable if you decide you want to pick up a signal from somewhere and have your computer do something with it.

Similarly, if PCB real estate isn't at too much a premium I'd put the GP outputs on a header so you can play around with them.


It depends on what is going to be communicating through the 65SPI. A clock source you may not have thought about is either OP2 or OP3 (pins 31 and 15, respectively) of the DUART. These pins can be rigged up to output the channel transmitter clocks (channel A on OP2, channel B on OP3) or in the case of OP3, the counter/timer (C/T) square wave when it is operating as a free-running timer (see bits 4, 5 and 6 in the auxiliary control register). The C/T in timer mode runs at a programmable fraction of the X1 clock, up to a maximum frequency of X1 clock ÷ 2. If you use OP3 to drive EXTC and have the C/T output driving OP3, changing the 65SPI clock rate is as simple as loading new values into the C/T's preset registers, followed by stopping and then starting the C/T.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

OK...
On thinking about it, I might just break out the unused I/O on the duart, because I can probably use it.
I'm tempted to break out the address and data pins too, given that this is a prototype, but maybe I should just install a set of SMD blinkenlights on the board.

I would also like to have NMI on a pushbutton. What is a good way of doing that?

Remember to debounce, something like the 2nd reset circuit at http://wilsonminesco.com/6502primer/RSTreqs.html .

_________________
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?

Garth answered that. The DS1813 is a very easy way to debounce an NMI push button, and is what I do in my POC units.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

Does that cause a spurious NMI on startup? Or is that masked by the reset sequence?