Input Voltage Levels for 65xx

[sburrow]

Howdy! I'm posting a question here in the newbie section because I think it belongs here.

I see on the W65CxxS datasheets that the minimum high input voltage is 0.7 x VDD. I've seen that on the 6502 and 65816 datasheets. (The W65C22S datasheet says 0.8 x VDD!) Simple math: 5V * 0.7 = 3.5 V.

Now on my last board PHI2 is produced from the 3.3V CPLD, pushed into the 6502 powered by 5V. That clock signal is definitely 0V to 3.3V. Yet the 6502 runs very well without any fuss!

Am I to expect the same behavior if I use all other 65xx chips? Does this mean that the 6522 would accept 3.3V input signals on it's PA and PB pins? Would the 65134 and 65265 I/O pins act similarly? Should I not expect it to work on every chip, and I just got lucky this time?

Just wondering why this works even when it's not following the datasheet.

Thank you everyone!

Chad

[gfoot]

Officially the 65C02 actually wants at least VDD-0.4 on PHI2 and some other pins. I think in general while you can get away with it, your mileage may very, so it's up to you.

In particular I imagine it's possible that it could draw more current than you'd like (from VDD, not the input) if the signal is outside the range the pin is designed for. Remember that a CMOS input is typically connected simultaneously to the gates of two MOSFETS which bridge VDD to VSS, the idea being only one allows current flow at a time, but a middling input voltage can allow some significant current to flow through both MOSFETs.

But this might not bother you for hobby use - it's one of those cases where maybe it will gradually damage the part or something, or later start acting up, but you also might get away with it and never need to look back. These parts seem pretty resilient.

[Dr Jefyll]

Hi, Chad. The following thread examines the issue of input voltage levels for modern WDC parts. Spoiler alert: it's as George said. Ie; while it's often possible to ignore the issue, your mileage may vary, so it's up to you. And BTW it's worth noting that, in regard to the required input voltage levels, WDC parts including their 'C02 differ markedly from those of most other 65xx manufacturers (such as Rockwell, CMD etc).

TTL Compatible... NOT! ( modern WDC CPU's )

-- Jeff

[sburrow]

Wow, those are very good replies. Thank you George and Jeff!

Jeff, that link is VERY detailed, very good to read. Thank you for that.

On my next board design, I'm planning on using a TXB0104 as a level-shifter. It seems easy to use, and the propagation delay seems very small for 3V3 to 5V conversions.

Thank you both again, this has been very helpful.

Chad

[AndrewP]

sburrow wrote:

I'm planning on using a TXB0104 as a level-shifter. It seems easy to use, and the propagation delay seems very small for 3V3 to 5V conversions.

Oooh, that's an interesting IC. I wish I'd found that before I ran off and tried to use an LSF0104 / LSF0204 with which I did not have good high frequency results.

A glance at the datasheet seems to show that I could use this to interface a 5V 65C22 with a 3.3V bus. That would be pretty sweet as my current solution is to just run the 65C22 at 3.3V.

[and3rson]

sburrow wrote:

Now on my last board PHI2 is produced from the 3.3V CPLD, pushed into the 6502 powered by 5V. That clock signal is definitely 0V to 3.3V. Yet the 6502 runs very well without any fuss!

Be careful though: I recently shot myself in a leg by driving Ф2 with ATF16V8. It seemed to work in the beginning, and everything was 5V. But when I touched Ф2 line with my finger or a probe, my '02 went berserk.
Eventually it was easily fixed by adding an additional 74HC14 in between my PLD & '02 Ф2 input. Other lines were totally fine though.
More info, as well as valuable inputs from other forum members: viewtopic.php?f=4&t=7636