6551 internal clock osc won't tick

R65C02P3 · Post by **R65C02P3** » Tue Oct 18, 2005 11:39 pm

Howdy everyone,

Have any of you been able to get the NMOS 6551 part to work
reliably (or at all) using a crystal rather than an ext. oscillator?
The CMOS parts that I've tested all work OK with just a xtal and
a couple of 10-20pF caps. According to the Rockwell data sheet,
the NMOS part should work, but I've not been able to get the
oscillator to start on it's own. I'm sure I'm not the first to have
noticed this; just wondering if anyone has gotten it to work.
(The NMOS 6551's I have are all Rockwell R6551AP, mixed date
codes)

Thanks in advance,

Brian

dclxvi · Post by **dclxvi** » Thu Oct 20, 2005 5:52 am

I have a couple of SSCs (Super Serial Card) for the Apple II that (I think) use an NMOS 6551, that work great to this day. The schematic in the manual shows a 1.8432 MHz crystal and a 10 pf cap (in parallel with each other) across the 6551 XTAL pins. I've never really looked at the oscillator circuit, since its never been any trouble, but I can take a closer look at the actual card itself if you'd like additional details.

coredump · Post by **coredump** » Thu Oct 20, 2005 9:01 am

The power supply switch-on transient can affect crystal oscillator start-up. I've seen a crystal oscillator circuit that would start up just fine on one type of PSU, but refuse to work at all on another. There was a modified circuit that had to be used in that case.

Have you tried the 6551 on a different power supply? Or maybe try connecting the 6551 board to the PSU after the PSU has been switched on (that'll avoid any "soft start" in the PSU).

R65C02P3 · Post by **R65C02P3** » Fri Oct 21, 2005 8:22 pm

Quote:

1.8432 MHz crystal and a 10 pf cap (in parallel with each other)

Hmmm. I don't think I tried that. I had a cap from each side of the xtal
to ground, tried with one cap, no caps, different value caps, etc.
In some cases, the osc would start if I "pinged" it after power-up.
If indeed a cap ACROSS the xtal is a solution (I'll try that soon),
that suggests to me that the NMOS '51 part needs a bit more feedback
from xtlo to xtli pins. It occurs to me now: I have seen resistors used
for the purpose (I'll try that, too...).

Quote:

The power supply switch-on transient can affect crystal oscillator start-up.

Yes...I had ruled that out (to my satisfaction) early-on: I ran the parts
stand-alone with xtal (and caps) only and a dry switch between the
Vcc pin and +5v supply...no luck. (The CMOS parts I've tested seem
rather forgiving with regard to power-up time vs osc start-up.)

I'll post here what I find. Thanks for the tips!

regards,
Brian

R65C02P3 · Post by **R65C02P3** » Sat Oct 22, 2005 3:55 pm

AHA!
Found the problem: a faulty crystal.
I tried the cap/res across xtal idea, it didn't help.
I found a few 1.8432MHz xtals in my junque box, and they all worked
just fine, with or without the caps.
I opened the "can" of the suspect and found the conductive coatings
on the quartz were damaged by what appears to be corrosion (!?).
Upon further inspection, I noticed that one of the lead-in wires had
come loose from the glass header. I guess moisture had gotten in
and caused the damage. I should have followed my own troubleshooting
methods: rule-out the obvious stuff first. Oh well, live and learn.

Thanks for the help!

Regards,
Brian