Knucker Question
Knucker Question
I have finally got my Knucker built and working after having some issues programming the FT2232 Module. (I am not sure how it happened but the EEPROM got scrambled and needed to be erased to get it working again.)
Although it seems to be receiving OK there is a very long delay while the Tuner is initialised (10 seconds) both with Portsdown and Ryde. Also Ryde initially says the Knucker is not responding (for a few seconds). Is this normal or is it a sign of another problem I need to look for.
Colin.
Although it seems to be receiving OK there is a very long delay while the Tuner is initialised (10 seconds) both with Portsdown and Ryde. Also Ryde initially says the Knucker is not responding (for a few seconds). Is this normal or is it a sign of another problem I need to look for.
Colin.
Re: Knucker Question
When it powers up from cold the NIM has to have it's firmware downloaded via I2C that is probably why you are seeing such a long delay.
- Charles
- Charles
Re: Knucker Question
All perfectly normal Colin
Dave
Dave
Re: Knucker Question
Thanks Charles and Dave.
I hadn’t seen anyone else mention the delay so I thought it was worth checking.
The reception seems fine on initial testing.
73
Colin.
I hadn’t seen anyone else mention the delay so I thought it was worth checking.
The reception seems fine on initial testing.
73
Colin.
Re: Knucker Question
That is the way moine has always behaved
Dave
G4FRE
Dave
G4FRE
g4eml wrote: ↑Sat Apr 10, 2021 2:56 pm
Although it seems to be receiving OK there is a very long delay while the Tuner is initialised (10 seconds) both with Portsdown and Ryde. Also Ryde initially says the Knucker is not responding (for a few seconds). Is this normal or is it a sign of another problem I need to look for.
Colin.
Re: Knucker Question (init delay)
Dear all,
I have built my knucker three days ago and also noticed the long delay (not for the first initialisation but whenever starting RX).
You must not forget that once Knucker is initialised, it searches for a signal (even if the frequency has not changed) and when found, it has to sync to it. Some delays cannot be avoided because they are due to the way DVB-T works but 20 seconds until the picture is displayed in my opinion is too long. When you start a QSO, you need to tell the OM on the other side that he has to wait at least 20 seconds before you can receive him.
I have been playing around with DVB-T since almost two years now and I have not found the ideal solution for a (stand alone) DVB-T transceiver yet. Until now I have used HiDes hardware (a transmitter similar to the HV-310 and a HV-122 DCA receiver) but to me this is not an ideal solution. I have developed a software for the RPi (still under construction), based on Qt, that controls the HiDes RX and TX via the serial port (to avoid the use of the remote control which is the same for RX and TX if I am right so you can imagine what happens) but for several reasons this is more difficult than I thought at the beginning. But that is not the only problem. The HV-122 DCA receiver is not able to display every received signal correctly due to DDR access problems (this is mentioned in the manual). That's why I was curious to see how Knucker behaves.
Does somebody know if and how the delay can be shortened? What about leaving Knucker running in the background when switching to TX? Would this be possible?
73 Ralf, DL5EU
I have built my knucker three days ago and also noticed the long delay (not for the first initialisation but whenever starting RX).
You must not forget that once Knucker is initialised, it searches for a signal (even if the frequency has not changed) and when found, it has to sync to it. Some delays cannot be avoided because they are due to the way DVB-T works but 20 seconds until the picture is displayed in my opinion is too long. When you start a QSO, you need to tell the OM on the other side that he has to wait at least 20 seconds before you can receive him.
I have been playing around with DVB-T since almost two years now and I have not found the ideal solution for a (stand alone) DVB-T transceiver yet. Until now I have used HiDes hardware (a transmitter similar to the HV-310 and a HV-122 DCA receiver) but to me this is not an ideal solution. I have developed a software for the RPi (still under construction), based on Qt, that controls the HiDes RX and TX via the serial port (to avoid the use of the remote control which is the same for RX and TX if I am right so you can imagine what happens) but for several reasons this is more difficult than I thought at the beginning. But that is not the only problem. The HV-122 DCA receiver is not able to display every received signal correctly due to DDR access problems (this is mentioned in the manual). That's why I was curious to see how Knucker behaves.
Does somebody know if and how the delay can be shortened? What about leaving Knucker running in the background when switching to TX? Would this be possible?
73 Ralf, DL5EU
Re: Knucker Question
I am sure that in time we can speed up the locking. The frequency search seems to take longer the narrower the channel bandwidth
and it also has to look to see whether the signal is DVB-T2 or DVB-T. Although the source code we have is not open source the Linux kernel driver
is almost identical and can be found here if anyone wants to spin their own code.
https://github.com/availink/dvb-fronten ... nk/avl68x2
- Charles
and it also has to look to see whether the signal is DVB-T2 or DVB-T. Although the source code we have is not open source the Linux kernel driver
is almost identical and can be found here if anyone wants to spin their own code.
https://github.com/availink/dvb-fronten ... nk/avl68x2
- Charles
Re: Knucker Question
Hello Charles,
I think there is much more to it than the demodulator driver.
Unfortunately, other than for the FTS-4334, I have not been able to find any information about the FTM-4762 NIM (I2C addresses, commands and so on). And it seems that something has to be loaded into the NIM on startup because the first initialisation takes quite a long time.
73 Ralf
I think there is much more to it than the demodulator driver.
Unfortunately, other than for the FTS-4334, I have not been able to find any information about the FTM-4762 NIM (I2C addresses, commands and so on). And it seems that something has to be loaded into the NIM on startup because the first initialisation takes quite a long time.
73 Ralf
Re: Knucker Question
Hi Ralf,
Here is a link to the NIM datasheet https://www.dropbox.com/s/j4z2fe2timcie ... 3.pdf?dl=0
Yes firmware is downloaded, the files are in that repository somewhere. The part that needs to be changed for
narrowband operation is the FFT table.
There is a lot of work involved but there is enough information in the public domain if someone was desperate enough.
- Charles
Here is a link to the NIM datasheet https://www.dropbox.com/s/j4z2fe2timcie ... 3.pdf?dl=0
Yes firmware is downloaded, the files are in that repository somewhere. The part that needs to be changed for
narrowband operation is the FFT table.
There is a lot of work involved but there is enough information in the public domain if someone was desperate enough.
- Charles
Re: Knucker Question
I too have now finished my Knuckler and connected it to the Ryde.
I transmitted from the PD4 a DVB-T signal and almost thought it was not working with the 20 second delay.
Maybe it is the beginnings of a time machine hi hi.
Happy Christmas to all
73
Nick
I transmitted from the PD4 a DVB-T signal and almost thought it was not working with the 20 second delay.
Maybe it is the beginnings of a time machine hi hi.
Happy Christmas to all
73
Nick