6502.org

Isn't that oscillator circuit supposed to have a couple of capacitors from each side of the crystal to ground? I'm not seeing any...

Many years ago I made a proto-board 65802 project, and the only problem that I had with it was an initial "zombie" condition that I finally traced to improper RESET hysteresis. Playing around with a simple RC combo finally did the trick, and I was off and running!

Here's a post with a picture and brief description.

Mike

That's what I thought, but every schematic I've seen doesn't have them or specify values.

Thanks Mike. One of the fun things about accelerated AA courses are the loss of things. I vaguely remember anything about hysteria.

I wasn't exactly hysterical about it, but I remember being up against a deadline, and my blood pressure had reached an unhealthy level because of it!

Put in "layman's" terms, I found that my /RES was floating to logic high too quickly after power-up to allow the '802 and clock to settle properly, and my solution was to tie /RES to ground through a capacitor, and to +5v through a resistor. I apologize for not remembering the actual working values of those two components, but the idea was to delay the rise of /RES just long enough to do the trick. There are more efficient ways of doing it, I'm sure, but my little kludge got me through to a passing grade, and that was my immediate goal.

Best of luck,

Mike

I believe the '802 has a Schmitt-trigger reset input though, while the old NMOS 6502 does not, and just putting a capacitor on that line to let it up more slowly can cause problems on the NMOS '02 as the voltage slowly makes its way through no-man's land, possibly with noise on it from the clock and other lines that might start toggling. That's why I advocate a clean reset signal.

Ah. Yeah, I have /RES tied to ground via a 10uF cap, and to VCC with a 2.7k. I'm going to try to verify the clock before I go forward. I just hope I didn't fry my 6502. What my current worry is. Lol

I didn't see any mention of the MPU's RDY input. It should be pulled up to Vcc with a resistor (3.3K recommended).

Also, Garth noted the problems of generating a proper reset pulse to the NMOS 6502. A good solution is the Maxim DS1813 reset generator, a handy little gadget that will hold reset down for ~150ms after Vcc stabilizes above 4 volts. The DS1813 also responds to a push button on the reset line, so it is both automatic and manual in nature.

Other reset generation methods have included using a 555 timer (the Commodore 64 had one for that purpose) or half of an LM393 comparator rigged up to act as a timer (I used to do that years ago). The DS1813 is the simplest solution that I know.

Lastly, the NMOS 6502 has relatively weak fan-out. It doesn't take too much to load it to where you can no longer get adequate voltage on its outputs when they are supposed to be a logic one. This is one of several reasons why we recommend starting out with the 65C02.

Would adding 2 22pF caps between both pins of the crystal and ground force the clock to oscillate?

See the 6502 primer's page on clock generation, at http://wilsonminesco.com/6502primer/ClkGen.html . There are a couple of ap. notes linked there also on making oscillators. Using oscillator cans however is less work, more reliable, and takes less board space.

I have read both of the pdfs on your primer, and the circuit you have at the top is the one that I am using, I just didn't know if anything is omitted to simplify it. I could do a Colpitts, but I was wanting to try it around the 7404. Just for a learning experience.

kasrym,

Please take heed of Garth's warning about an inadequate /RST circuit ... the man knows his stuff! Even though it was 27 years ago, I was using a CMOS chip that was apparently much more tolerant of marginal design than your NMOS chip. I would strongly recommend implementing one of Garth's examples, rather than a simple R-C combo.

Mike

In going from 74LS04 to 7404, I think the resistor from the input (on the left) to ground will need to be about a quarter as much, since the 7404's input current is quadruple that of the 74LS04. There is no reason to use 7400 logic anymore though, unless you have a lot of it and you're in a place or situation that makes it hard to get at least 74LS if not 74HC or other CMOS. The capacitors provide phase shift; but for myself, without being strong in oscillator-related math (which can get rather complex for oscillators if you want them to be dependable across the range of all potential operating conditions), I don't know if maybe the 7404/74LS04's inputs provide a heavy enough load to omit the capacitors. I design mostly analog circuits for a living; but although I've designed various kinds of oscillators that do work, I don't consider myself to be expert enough in that department to fully trust my own crystal oscillator designs for products we will be marketing. For that, I'd rather buy an off-the-shelf one.