[Coco] OT: 63B09

[Richard E. Crisilip]

On Sunday, October 30, 2011 05:48:21 PM gene heskett wrote:
> On Sunday, October 30, 2011 05:17:55 PM Richard E. Crisilip did opine:
> > On Saturday, October 29, 2011 01:48:17 PM Louis Ciotti wrote:
> > > I can across an NTE6809 processor and thought I would give it a try
> > > as
> > > a replacement for the 63b09.  It has an identical Pinot so I figured
> > > it may work.  I plug it in the socket, and I get nothing at all. 
> > > The
> > > external clock (4mhz) crystal does not even work, which on the one I
> > > am trying to replace does.  Is the IC bad, or is it a compatibility
> > > issue.
> > > 
> > > On Thursday, October 20, 2011, Louis Ciotti <lciotti1 at gmail.com> wrote:
> > > > Off topic, but since the CoCo uses 6809s, figured someone on
> > > > here
> > > > might
> > > 
> > > know about this.
> > > 
> > > > I have a radio controller from circa late 80s that uses a 63B09
> > > > that
> > > > I
> > > 
> > > think may be bad.  What can I use in its place and is there a
> > > source?
> > > 
> > > > The problem I am seeing is RN6(Pins 24-31) has some "noise" on
> > > > the
> > > > line
> > > 
> > > and the waveform is not exactly square like the rest of the data
> > > lines
> > > when observed with an oscilloscope.
> > > 
> > > 
> > > --
> > > Coco mailing list
> > > Coco at maltedmedia.com
> > > http://five.pairlist.net/mailman/listinfo/coco
> > 
> > If I am wrong, I am sure the more knowledgeable CoCo folks will correct
> > me. The CoCo requires a CPU with an external clock as the 6809E and the
> > 6903Es are.
> 
> This is not in a coco, Richard, its some sort of a ham radio controller I
> believe if you follow the thread back far enough.
> 
> The OEM chip is an HD63B09P, CMOS but not an E model.  My guess is that
> when he plugs in the NTE part, which is likely either NMOS or PMOS, the
> biasing of the internal oscillator is so far off compared to what the
> Hitachi CMOS chip needs that the internal oscillator never starts.
> 
> One thing that would help me ascertain at least a better SWAG was a comment
> he made about noisy data lines.  Apparently the OP has a scope, so if he
> could post a properly zeroed, dc coupled, 100 mhz or so bandwidth, snap
> shot of the scope screen while looking at 1 or 2 of these noisy signals,
> put on one of the pastebin's & give us the link so I could take a look.
> That might help clear up my cloudy crystal ball quite a bit.
> 
> Something else that comes to mind also is aged, oxide blackened IC legs &
> socket contacts.  Temporary fixes can be had by lifting each end of the
> chip about 1/16" and pressing them back down, but the real fix is probably
> to remove them all, keeping track of what type goes in which socket, pour
> out 2/3rds of a cup of _fresh_ "Tarnex" in a small bowl, drop all the chips
> in and stir gently with a long q-tip stick or ?? until the chip legs are
> nice & clean, nearly white.  Pick them back out of the cup and rinse each
> one in warm running water, drop on a bath towel and dry with a hair dryer.
> Reinstall & see if that fixed it.  I have probably used something in excess
> of a 2 gallons of Tarnex for that over the last 25 years.  Sometimes I have
> resorted to wetting the sockets with it using a q-tip, let them soak for
> half an hour, then rinse the whole board with warm running water using the
> spray head in the kitchen sink, but because you can't agitate it down
> inside the socket pins, its not all that effective.  I have had to run down
> the bad socket and replace it occasionally, which is a right PIMA.  But
> when something to replace the recalcitrant device was in those days
> 25-50k$, you fixed it by whatever means you could.  When I say rinse, that
> means 20 to 30 minutes of running water, don't even leave the smell of
> tarnex in the room.
> 
> Oh oh, my soap box is crumbling. ;-)
> 
> > The chip you are trying has an internal clock. Also just
> > because the pin match up, you MUST be very careful. The 68020 and 68030
> > match pin for pin too, but that does not mean you will not let the blue
> > smoke out 8-(.
> > 
> > --
> > Coco mailing list
> > Coco at maltedmedia.com
> > http://five.pairlist.net/mailman/listinfo/coco
> 
> Cheers, Gene

Thanks for the gentle knuckle wrap 8-). You are so correct, I did not follow 
the thread far enough back. In my defense, I lost a lot when I upgrade 
(downgraded??) to Ubuntu 11.10. Killed my KMail history 8-/. Hate it when that 
happens.