[time-nuts] Spectracom 8170 -> SWCC clock
John Ackermann N8UR
jra at febo.com
Wed Feb 22 19:55:11 UTC 2012
The 8170 isn't smart enough to calculate the leapsecond immediately.
Instead, it uses its error correction routine which takes about 4
minutes after the event to realize that it is one second off, then
updates the time to match. Details quite far down the page at
On 2/22/2012 2:48 PM, Jim Hickstein wrote:
> On 2012/01/23 15:28, Jim Hickstein wrote:
>> Reviewing the schematic for the 8170 (I'll get back to the setting
>> coil in the
>> clock a bit later), my unit does not seem to have Option 18 and A4-U4,
>> the 8255A
>> at its heart, is not populated. A pity: it would have lovely BCD
>> outputs and an
>> on-time pulse, just what I need. (The recent discussion of multiplexed
>> LED outputs tells me I don't want to try to detect ":00:00" or
>> ":59:59" there.)
>> What are the chances my EEPROMs would DTRT if I just added an 8255A to
> Looking better today. I dragged a scope up from the basement. CS is
> getting hit, about 11 times, near the beginning of each second (image).
> I guess this means I have to design the circuit, now, and place an order
> for some parts.
> > What other components should I add to complete the Option 18
> A3U4, the 8255A, seems to be the only thing missing, apart from A3J4,
> the 50-pin header that connects to a cable to bring these BCD signals
> out to the rear panel DD-50, also missing. For my purposes I can stop at
> the header. It'll be a nice place to mount a little daughterboard. This
> is on sheet 4 of A3, the uP board schematic, not on a separate sheet as
> some other options are.
>> If I need different EEPROMs to enable Option 18, where could I get the
>> (The parts and the programmer I can probably get.)
> Perhaps I'll do a socket and the 8255A first, just to see the output
> pins move as expected. This means paying Digikey for shipping twice. Oh,
> well. (Anyone near http://www.airnav.com/airport/KMSP got an 8255A lying
> around? I have some Augat 40-pin sockets left over from 1983, packrat
> that I am.)
> Maybe I can get it done in time to have it send a pulse to the clock
> during :59:59 _or_ :59:60. :-) I don't recall if an 8170 is quite that
> savvy, to produce this on its output, but the BCD encoding could do it.
> We'll find out, eh? And the clock will be set properly the next hour,
>> On 2012/01/20 16:23, Jim Hickstein wrote:
>>> I have a Spectracom 8170 in the living room (who doesn't?), and a
>>> Western Union
>>> time-service clock, a.k.a SWCC clock -- a nice one, in a 3-foot-high
>>> wood case.
>>> I've been watching TV with this combination for years and years but
>>> never got
>>> around to feeding a pulse from the 8170 to discipline the other one.
>>> Now that
>>> Western Union no longer provides the service. :-) But it just begs to
>>> be done.
>>> I did draw up a TTL circuit, once (on a napkin, naturally, which I
>>> have now
>>> misplaced), that could live inside the 8170. I figured out the
>>> minimum number of
>>> inputs needed to detect when the MM:SS LED displays said 00:00 (for
>>> one second).
>>> It would close a relay, which could feed the winding-battery power to
>>> hour-set solenoid down a pair of wires from 10 feet away. But I never
>>> built it.
>>> A little over a year ago the TS clock was getting gummy and
>>> free-running slower
>>> and slower, so I sent it to the clock hospital. It's back and
>>> nicely, so maybe it's finally time I did this. Anyone got a better
>>> idea than my
>>> little TTL circuit, on a breadboard inside the 8170? I'd like to get
>>> it across
>>> the rear panel without cutting a new hole, if I can avoid it. But
>>> maybe the
>>> right connector would do. Another time-code receiver in the TS clock,
>>> e.g. a GPS
>>> module that provides a relay closure for 1 second on the hour (if
>>> such exists)
>>> might be neater. But the living room faces north.
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to
>>> and follow the instructions there.
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to
>> and follow the instructions there.
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts