Thanks to all: 8820 now working

[Arno Kletzander]

Sat, 15 Apr 2006, "Bruce Lane" <kyrrin at bluefeathertech.com> wrote:
>    My thanks to all who offered assistance on my issues with the 'tech 
> special' Trak Systems 8820 GPS station clock. The unit was successfully 
> repaired, and has been working for nearly a full week without any 
> further signs of problems.
>
>    For the curious: The problem turned out to be that one of the 
> firmware EPROMs developed a broken internal bond wire on the output-
> enable lead. This caused the chip to appear completely blank to both the 
> Unisite programmer and the Trak device. My contact at Trak was kind 
> enough to send over image files to do a fresh set of EPROMs.
>
>    The only other adjustment I found myself making was a fine-tune 
> alignment on the 10MHz ovenized oscillator, to bring its center 
> frequency back to a point where the reference circuitry could discipline 
> it down to full accuracy. That was also accomplished without incident, 
> and I now have a second Stratum-1 level clock and frequency standard for 
> my lab.
>
>    The moral of the story: EPROMs can fail too! Just not in the way we 
> might expect. ;-)

Congratulations to getting this thing back into functional state. I wish you
a good time and many years of satisfying operation for your device.

Incidentally, I've been having a bit of fun with a MEINBERG GPS166 Satellite
Controlled Clock lately - it had been set aside as defective at the
computing center of Erlangen University and I happened to get it for the
asking, together with its bullet-shaped "radome" antenna/amplifier.

When I finally built myself a coaxial cable to connect the two (SMA <->
N-Type), I had a working setup! But after a relocation a bit later, I got
some smelly smoke and an "antenna fault" error when I powered it on. Turned
out to be a small cylindrical choke in the shielded RF box which had gone
open circuit - possibly involved with supplying power to the antenna
amplifier. 

I checked the coaxial cable for shorts, couldn't find one at that time,
replaced the choke and had the next one burn through. I disassembled and
reassembled the N connector very carefully, replaced the choke once again
and have been operating the clock without problems since then. It too can
give out fixed and adjustable frequency pulses - along with second and
minute signals, which you could amplify adequately to drive daughter clocks
like that falling-leaf wall clock I have (needs a 24V pulse every minute).

The special plus of the unit I have is the so-called "Erlangener Firmware"
which was custom-written for Erlangen University. The usual variety only
used the serial port for giving the time - either in intervals, when a ? was
entered or when a pulse occurred on a logic input - but this one also puts
out the geographical location, so one could even use it for a navigation
system, in theory at least.

So long,

-- 
Arno Kletzander
Stud. Hilfskraft Informatik Sammlung Erlangen
www.iser.uni-erlangen.de

Echte DSL-Flatrate dauerhaft für 0,- Euro*!
"Feel free" mit GMX DSL! http://www.gmx.net/de/go/dsl