<snip> Second issue: The loudest playback sounds cause the URI to go into saturation and distort. I may not have calibrated the levels very carefully, but even if I manually lower the txlevel setting from 800 down to, say, 300 it still distorts - it's just a quieter distortion. What can cause this? <snip>
I have the same problem with a modified CM108 usb installation.
After setting all the 'radio tune' parameters, there is serious distortion on voice modulation peaks, irrespective of the source being Iaxrpt or another node.
However, the 'radio tune txvoice' generates a perfect sine wave at 3 KHz deviation, as seen on an oscilloscope at the exciter modulator or on a service monitor off-air. I am using duplex=0 in rpt.conf since the node operates simplex.
Dave VE7MQ 2140
_______________________________________________ App_rpt-users mailing list App_rpt-users@qrvc.com http://qrvc.com/mailman/listinfo/app_rpt-users
I have not seen this on a properly configured system. Chances are it has to do with de-emphasis configuration on the receive side, or pre-emphasis/limiting (prelim) in the transmit side) Where is the audio getting distorted? Transmit or receive? so you see the distortion on the TX line or RX line if you look with a scope? You might be able to determine where it is getting distored by connecting your node to a working node you can monitor, then send audio both ways. Disclaimer: I offer URI support and suggestions as a courtesy only as I do not derive any income from sales of the URI. I am in no way obligated to provide support at any level. Steve WA6ZFT Steve: Thank you for your response. Much appreciated. This distortion also appears when accessing the transmitter digitally via Iaxrpt with no radio receivers involved. Ditto for when my node is connected elsewhere, such as Calnet. The problem is entirely on transmit, somewhere in the route from apt_rpt internals to the MICOR exciter audio input. The receive audio is perfect. I have two independent nodes on separate computers. I followed your suggestion and linked them together. Using a service monitor, I fed 1 KHz tones at various deviations into the receiver of Node A and monitored the transmit output of problem Node B. There is no distortion until reaching 4 KHz. At 5 KHz deviation, there is serious distortion of the sine wave. I want to emphasize that this problem remains constant no matter if the modulation source originates in a radio receiver, another node or Iaxrpt. As Ken mentioned earlier, adjusting txvoice moves the deviation up and down but has no effect upon the distortion. FWIW, the distortion (on voice) sounds like listening to a loudspeaker with a ripped cone. After reading Ken's last post, I tried 'txprelim=yes' and it was a great improvement. It would be nice if the limiting could be selected independently. To summarize: it seems that with 'txprelim=no', audio from a receiver seeing 5 KHz peaks or the equivalent from Iaxrpt or from other nodes, is going to be distorted by the URI being overdriven. Unfortunately, the distortion strikes right at the deviation level that most users hit regularly. I am not using the DMK product. I understand it's a very professional unit. I rolled my own some time ago, following the original construction details. Thanks again for a great system. 73, Dave VE7MQ.
David Powell-Williams wrote:
<snip> Second issue: The loudest playback sounds cause the URI to go into saturation and distort. I may not have calibrated the levels very
carefully,
but even if I manually lower the txlevel setting from 800 down to, say,
300
it still distorts - it's just a quieter distortion. What can cause this? <snip>
I have the same problem with a modified CM108 usb installation.
After setting all the 'radio tune' parameters, there is serious distortion on voice modulation peaks, irrespective of the source being Iaxrpt or another node.
However, the 'radio tune txvoice' generates a perfect sine wave at 3 KHz deviation, as seen on an oscilloscope at the exciter modulator or on a service monitor off-air. I am using duplex=0 in rpt.conf since the node operates simplex.
Dave VE7MQ 2140
_______________________________________________ App_rpt-users mailing list App_rpt-users@qrvc.com http://qrvc.com/mailman/listinfo/app_rpt-users
I have not seen this on a properly configured system. Chances are it has to do with de-emphasis configuration on the receive side, or pre-emphasis/limiting (prelim) in the transmit side)
Where is the audio getting distorted? Transmit or receive? so you see the distortion on the TX line or RX line if you look with a scope?
You might be able to determine where it is getting distored by connecting your node to a working node you can monitor, then send audio both ways.
Disclaimer: I offer URI support and suggestions as a courtesy only as I do not derive any income from sales of the URI. I am in no way obligated to provide support at any level.
Steve WA6ZFT
Steve:
Thank you for your response. Much appreciated.
This distortion also appears when accessing the transmitter digitally via Iaxrpt with no radio receivers involved. Ditto for when my node is connected elsewhere, such as Calnet. The problem is entirely on transmit, somewhere in the route from apt_rpt internals to the MICOR exciter audio input. The receive audio is perfect.
I have two independent nodes on separate computers. I followed your suggestion and linked them together. Using a service monitor, I fed 1 KHz tones at various deviations into the receiver of Node A and monitored the transmit output of problem Node B. There is no distortion until reaching 4 KHz. At 5 KHz deviation, there is serious distortion of the sine wave.
I want to emphasize that this problem remains constant no matter if the modulation source originates in a radio receiver, another node or Iaxrpt. As Ken mentioned earlier, adjusting txvoice moves the deviation up and down but has no effect upon the distortion. FWIW, the distortion (on voice) sounds like listening to a loudspeaker with a ripped cone.
After reading Ken's last post, I tried 'txprelim=yes' and it was a great improvement. It would be nice if the limiting could be selected independently.
To summarize: it seems that with 'txprelim=no', audio from a receiver seeing 5 KHz peaks or the equivalent from Iaxrpt or from other nodes, is going to be distorted by the URI being overdriven. Unfortunately, the distortion strikes right at the deviation level that most users hit regularly.
I am not using the DMK product. I understand it's a very professional unit. I rolled my own some time ago, following the original construction details.
Thanks again for a great system.
73, Dave VE7MQ.
I would check and see if you have already or need dc blocking between the URI and the mic input as the mic input most likely has dc bias for the mic cartridge... the URI does NOT like DC back fed. There are DC blocked inputs and outputs on the other row of pins.. I found that DC was an issue on most radios I am hooking to on both the TX and RX sides... It causes the audio to get unsymetrical...IN general I use those pins as they are capacitively coupled and most my issues like you are indicating.. stopped.. YMMV Just a thought to check... Doug KD8B
I would check and see if you have already or need dc blocking between the URI and the mic input as the mic input most likely has dc bias for the mic cartridge... the URI does NOT like DC back fed. There are DC blocked inputs and outputs on the other row of pins.. I found that DC was an issue on most radios I am hooking to on both the TX and RX sides... It causes the audio to get unsymmetrical...IN general I use those pins as they are capacitively coupled and most my issues like you are indicating.. stopped.. YMMV Just a thought to check... Doug KD8B
I am using the URI which has both DC-blocked and solidly-coupled. I have wired to the pin that has the DC block. For the moment, I am just feeding it to some amplified speakers for testing. They present a high impedance to the URI. I can't prove it, but it sounds like Asterisk is coming up with some audio that is bigger than the 16-bit number that is supposed to hold the value for the D/A and, of course, the next bigger number after full scale is zero or minus full scale or something equally horrible. It is strange that the pre-recorded messages would be subjected to this though. They are, by definition, good-sounding 16-bit audio data. When I listened to one offending phrase off-line (Apple Quicktime) and it sounded fine. I guess the recorded audio passes through additional DSP that alters the levels. Anyway, the limiting feature seems like a good thing to have both with and without pre-emphasis. Ken
_______________________________________________ App_rpt-users mailing list App_rpt-users@qrvc.com http://qrvc.com/mailman/listinfo/app_rpt-users
David Powell-Williams wrote:
<snip> Second issue: The loudest playback sounds cause the URI to go into saturation and distort. I may not have calibrated the levels very carefully, but even if I manually lower the txlevel setting from 800 down to, say, 300 it still distorts - it's just a quieter distortion. What can cause this? <snip>
I have the same problem with a modified CM108 usb installation.
After setting all the 'radio tune' parameters, there is serious distortion on voice modulation peaks, irrespective of the source being Iaxrpt or another node.
However, the 'radio tune txvoice' generates a perfect sine wave at 3 KHz deviation, as seen on an oscilloscope at the exciter modulator or on a service monitor off-air. I am using duplex=0 in rpt.conf since the node operates simplex.
Dave VE7MQ 2140
_______________________________________________ App_rpt-users mailing list App_rpt-users@qrvc.com http://qrvc.com/mailman/listinfo/app_rpt-users
I have not seen this on a properly configured system. Chances are it has to do with de-emphasis configuration on the receive side, or pre-emphasis/limiting (prelim) in the transmit side)
Where is the audio getting distorted? Transmit or receive? so you see the distortion on the TX line or RX line if you look with a scope?
You might be able to determine where it is getting distored by connecting your node to a working node you can monitor, then send audio both ways.
Disclaimer: I offer URI support and suggestions as a courtesy only as I do not derive any income from sales of the URI. I am in no way obligated to provide support at any level.
Steve WA6ZFT
Steve:
Thank you for your response. Much appreciated.
This distortion also appears when accessing the transmitter digitally via Iaxrpt with no radio receivers involved. Ditto for when my node is connected elsewhere, such as Calnet. The problem is entirely on transmit, somewhere in the route from apt_rpt internals to the MICOR exciter audio input. The receive audio is perfect.
I have two independent nodes on separate computers. I followed your suggestion and linked them together. Using a service monitor, I fed 1 KHz tones at various deviations into the receiver of Node A and monitored the transmit output of problem Node B. There is no distortion until reaching 4 KHz. At 5 KHz deviation, there is serious distortion of the sine wave.
I want to emphasize that this problem remains constant no matter if the modulation source originates in a radio receiver, another node or Iaxrpt. As Ken mentioned earlier, adjusting txvoice moves the deviation up and down but has no effect upon the distortion. FWIW, the distortion (on voice) sounds like listening to a loudspeaker with a ripped cone.
After reading Ken's last post, I tried 'txprelim=yes' and it was a great improvement. It would be nice if the limiting could be selected independently.
To summarize: it seems that with 'txprelim=no', audio from a receiver seeing 5 KHz peaks or the equivalent from Iaxrpt or from other nodes, is going to be distorted by the URI being overdriven. Unfortunately, the distortion strikes right at the deviation level that most users hit regularly.
I am not using the DMK product. I understand it's a very professional unit. I rolled my own some time ago, following the original construction details.
Thanks again for a great system.
73, Dave VE7MQ.
_______________________________________________ App_rpt-users mailing list App_rpt-users@qrvc.com http://qrvc.com/mailman/listinfo/app_rpt-users
This probably explains why I never experienced this. I never use the mic input with a URI as I like to send the pre-emphasized and limited audio directly to the modulator. This works quite well as long as you have a a transmitter using direct FM and not phase modulation. Steve WA6ZFT
participants (4)
-
David Powell-Williams -
Doug Bade -
Ken -
Stephen Rodgers