GB3EY status update: delayed
Posted: Mon Nov 02, 2020 11:56 pm
The repeater was taken off-site shortly after it was first put in due to an issue with the receiver. Unfortunately my rotator had failed at the time and as a result I wasn't able to get any test signals into it for diagnostics. It was therefore necessary to get it home to put on the bench.
Bench testing showed that the receivers needed a very strong signal to switch between H262 and H264. A marginal signal would create a 'signal locked' condition which was passed to the logic which duly switched the transmitter video path to the receiver. However, despite being locked there was no video at all, just a black screen from the receiver.
Another problem found with the little bit of on-site testing that I was able to do showed that the receivers were still being bothered by the Claxby radar despite over 120dB of filtering. Not only did they freeze or drop the video altogether but they also took far too long to recover; sometimes they hadn't got the picture back before the next burst of radar came along eight seconds later. A desense issue was seen that needed bench checks.
After a lot of experimentation with different receivers, where I deliberately injected the radar into a combiner with a weak DATV signal (the peak radar signal from my antenna and preamp combination is almost 10mW when beaming at Claxby), I concluded that set top boxes are just not suitable for our site.
I then tried the Ryde and noticed a substantial improvement. It copes with the radar much better than any of the set top boxes because it mainly freezes briefly and then carries on whereas the set top boxes drop lock. A bonus was that the logic needed no modification to use the 'Locked' output from the Ryde.
The symbol rate is not restricted to 1MS/s or above using the Ryde. Tests with the Ryde using DVB-S2 @ 333kS/s, 8PSK, H264, FEC 3/4 have been very encouraging with good quality audio and video. Swapping between H262 and H264 does not cause a problem and provided the number of symbol rates scanned is kept to two the lock time is short. Another advantage is that the Ryde has remotely accessible menus for configuration changes and diagnostics.
The desense issue has been traced to wideband 'crud' from the transmitter when modulated. Specifically, it's the result of IMD products in the driver and PA stages. The filtering on the transmitter output appeared initially to be adequate but there's about 5µV of noise near the input frequency when the transmitter is relaying. That's enough to make it seem deaf although the problem is much less noticeable with a 333kS/s QPSK FEC1/2 input that wasn't possible with the set top box receivers.
To fix the desnse issue we are going to need one of the duplex filters from here: http://www.id-elektronik.de/en/; further progress is on hold until funds can be raised for its purchase.
Clive G3GJA
Bench testing showed that the receivers needed a very strong signal to switch between H262 and H264. A marginal signal would create a 'signal locked' condition which was passed to the logic which duly switched the transmitter video path to the receiver. However, despite being locked there was no video at all, just a black screen from the receiver.
Another problem found with the little bit of on-site testing that I was able to do showed that the receivers were still being bothered by the Claxby radar despite over 120dB of filtering. Not only did they freeze or drop the video altogether but they also took far too long to recover; sometimes they hadn't got the picture back before the next burst of radar came along eight seconds later. A desense issue was seen that needed bench checks.
After a lot of experimentation with different receivers, where I deliberately injected the radar into a combiner with a weak DATV signal (the peak radar signal from my antenna and preamp combination is almost 10mW when beaming at Claxby), I concluded that set top boxes are just not suitable for our site.
I then tried the Ryde and noticed a substantial improvement. It copes with the radar much better than any of the set top boxes because it mainly freezes briefly and then carries on whereas the set top boxes drop lock. A bonus was that the logic needed no modification to use the 'Locked' output from the Ryde.
The symbol rate is not restricted to 1MS/s or above using the Ryde. Tests with the Ryde using DVB-S2 @ 333kS/s, 8PSK, H264, FEC 3/4 have been very encouraging with good quality audio and video. Swapping between H262 and H264 does not cause a problem and provided the number of symbol rates scanned is kept to two the lock time is short. Another advantage is that the Ryde has remotely accessible menus for configuration changes and diagnostics.
The desense issue has been traced to wideband 'crud' from the transmitter when modulated. Specifically, it's the result of IMD products in the driver and PA stages. The filtering on the transmitter output appeared initially to be adequate but there's about 5µV of noise near the input frequency when the transmitter is relaying. That's enough to make it seem deaf although the problem is much less noticeable with a 333kS/s QPSK FEC1/2 input that wasn't possible with the set top box receivers.
To fix the desnse issue we are going to need one of the duplex filters from here: http://www.id-elektronik.de/en/; further progress is on hold until funds can be raised for its purchase.
Clive G3GJA