Once a wrong R7 found on a "RIPE-NCC GPS interface 2.0" card was identified as the only cause of the malfunction and problem was solved after replacement. This brief will not apply to any other component mistake but can give an idea about how things can be screwed because of one single wrong component.
R7 is a 'place' on the "RIPE-NCC GPS interface 2.0" card and it is supposed to be filled by a ' 6X-2-102 ' resistor array on a correctly manufactured one. It is very close to the DB9 connector. A visual inspection on the component side should be enough to find it.
While installing check the part installed as R7. Do not use anything carrying something else than the ' 6X-2-102 '.
If you are not sure about the card for some reason keep it running with an antenna until the PPS LED on the card starts flashing and disconnect the computer end of the short serial cable. Let the card end stay there. Grab an oscilloscope or a digital multimeter (anything properly visualizing an electric signal is supposed to be enough). Connect your measurement device's GROUND to pin5 of the DB9 end of the serial cable. Try to measure something on pin1 and pin2. If you see something with one second period on pin1, this is PPS ready to apply to RS-232. Also pin2 may or may not have some activity but if both pin1 and pin2 gives 'very similar' output you are on the right track.
Do not lose time reading this, it is something else or sure in a different way.
If two signals are supposed to be different (why do they have two pins otherwise?) but seem very similar there must be some 'leak' connecting them together. Take the card out of the computer after shutting it down. On the soldering side measure the resistance between every two consequential pins of R7. They are supposed to read around a kiloohm (900-1100 or 0.9-1.1K) for each of the three pairs. If they are not, also measure between every two pin of the R7. Since resistors are bilateral components they read the same even if you change the measurement polarity. So, for 6 pins it is only 5+4+3+2=14 measurements. Go ahead. If you find any value not around infinity or 1Kohm something is sure wrong. Get rid of the card, label it as broken, get a new one.
Once a wrong component was installed as R7 on a card. The PPS signal starts a while after power up. While the PPS line is idle it was well possible to communicate with the antenna but after PPS starts this is not the case since GPS line will have PPS as a very suppressing 'noise' or vice-verse. Since every other component was fine the card was changed and analysed later.
Do not bother. This happened only once. If you catch one more examples like this, warning all tt-ops is enough for now. If you want to do a repair for exercise take out the wrong part and put something else satisfying the RIPE-NCC GPS Interface 2.0 schematics. Note that the third resistor of the array is not used by this implementation.
CO 20020314