P25RX Firmware Updates / Testing

btt

Active Member
Premium Subscriber
Joined
Mar 11, 2020
Messages
836
Location
Wa State
I've noticed a bug with the skip function from the panel button. I tried the 2157 and 1909 builds, but the device seems to ignore skip commands. I can hold on TG and enter BT pairing just fine, but skip no longer seems to work. I have not tried from BTConfig yet.
If the firmware update reset the configuration to defaults, you may need to reconfigure the button for the skip function. I tested the function here. It appears to be working with version 2157.
 

dispatchgeek

Member
Joined
Feb 29, 2004
Messages
72
Location
Fightin' River, Michigan
I did verify the Button was assigned correctly. Even moved it to single click with the same result. Skip works when I hit it in BTconfig, but not from the button. I copied a log where I attempted multiple skips from the front button, TG continued along until I finally hit skip in BTConfig, where it finally followed. I have attached a snip of the log in question.
 

Attachments

  • Like
Reactions: btt

btt

Active Member
Premium Subscriber
Joined
Mar 11, 2020
Messages
836
Location
Wa State
Thank you for attaching the log file. The line that shows "starting discovery" indicates that it was starting the Bluetooth pairing. Can you attach a screenshot of the Button CFG tab so I can try to reproduce here.
 

dispatchgeek

Member
Joined
Feb 29, 2004
Messages
72
Location
Fightin' River, Michigan
Todd- I made a clearer log. The first "Skip" in the log was attempted from the hardware button. It appears to be in slightly different formatting to the BTConfig skip button. Audio continued after pressing the hardware button, but stopped immediately after pressing the soft skip button. I may have been mashing the button too much and inadvertently hit the BT Pairing command. (As for how this hospital talkgroup from several counties away keeps lighting up our towers, that's something else.)
 

Attachments

  • Like
Reactions: btt

btt

Active Member
Premium Subscriber
Joined
Mar 11, 2020
Messages
836
Location
Wa State
Thanks. I think I see what is happening now. The skip command didn't immediately return to the control channel. It wouldn't be active until the current voice call ended. Release 2021-06-09_1141 will immediately return to the cc when you skip. The button skip and console skip use exactly the same function, so they should operate identically. Let me know if you continue to have issues with this.
 

TomTN

Member
Joined
Jan 26, 2003
Messages
60
Location
Nashville, TN
I really want to love the newest experimental versions as the reboot issue has long been a thorn. Unfortunately both 0743 and 1141 have extremely choppy audio on a Harris P2 system. Most of the audio cannot be understood. Hopefully this can be resolved as the audio was great in previous versions.
 

btt

Active Member
Premium Subscriber
Joined
Mar 11, 2020
Messages
836
Location
Wa State
I really want to love the newest experimental versions as the reboot issue has long been a thorn. Unfortunately both 0743 and 1141 have extremely choppy audio on a Harris P2 system. Most of the audio cannot be understood. Hopefully this can be resolved as the audio was great in previous versions.
I can't think of anything that would be causing this except for the channel bandwidth. The default used to be 18.2 kHz. Try that setting for both P1 and P2 on the advanced tab. What was the last version that worked for you?
 

btt

Active Member
Premium Subscriber
Joined
Mar 11, 2020
Messages
836
Location
Wa State
@TomTN
Harris uses more bandwidth for the WCQPSK "wide eyes" modulation. I know Harris systems work well with the 18.2 kHz bandwidth setting, so would be worth trying.
 

btt

Active Member
Premium Subscriber
Joined
Mar 11, 2020
Messages
836
Location
Wa State
What does it mean when the plot basically looks like an X?
It means that the symbol synchronizer is locked onto the incoming signal with the amplitude varying. The amplitude changes are probably due to simulcast reception, but can be due to being mobile, etc. Remember that this is not a "constellation display". This is a plot of the differential symbols right before they are passed to the decoder. You are seeing the effects of digital signal processing, RF AGC, etc. As long as the symbol is in the correct quadrant at symbol time, then it will be decoded correctly.
 

W0RS

Member
Premium Subscriber
Joined
Dec 9, 2003
Messages
195
Location
Nixa, MO
My symbol plotter here at my residence forms a circle rather then a X but when in the car most of the time it is an X.....
 

Cheeseburgers

Member
Joined
Jan 9, 2018
Messages
388
One site is 4 dots. 1 in each quadrant. the other site has the plotter going nuts and it stretches from the corners into the center to form an x. the 325 ALWAYS struggled at my office, SDS did too and the G4/5 also had a tough time.. the p25rx is on par with the Unication maybe even out performing by a hair.
 
  • Like
Reactions: btt

btt

Active Member
Premium Subscriber
Joined
Mar 11, 2020
Messages
836
Location
Wa State
Any Ideas regarding CC scanning coming?
Adding more features to the roaming option is still planned, but it is going to be a while.

My symbol plotter here at my residence forms a circle rather then a X but when in the car most of the time it is an X.....
I assume you mean there are 4 clusters of symbols on the circumference of a circle in their respective quadrants? If you see an actual circle drawn out that would indicate that there is a constant amplitude with the frequency error being too large to decode properly. (symbol phase spinning around on the unit circle). The solution for this would be to make sure the reference frequency is correct.

One site is 4 dots. 1 in each quadrant. the other site has the plotter going nuts and it stretches from the corners into the center to form an x. the 325 ALWAYS struggled at my office, SDS did too and the G4/5 also had a tough time.. the p25rx is on par with the Unication maybe even out performing by a hair.
Thanks for pointing that out. When you see the X, you know that would be a problem area for old style FM demodulation where the symbol decode decision is dependent on the absolute instantaneous phase at symbol time. With the P25RX (and I assume SDS, G4), where quadrature is used (I/Q), the symbol decode decision is the phase change between the previous symbol time and the current symbol time. (relative phase change). This way ,if the absolute instantaneous phase at symbol time moves around as received, the symbols will still be decoded correctly as long as the sync is good. The way CQPSK works is that you have a QPSK constellation (4 dots in quadrants) that is rotated by pi/4 every symbol time. This is what makes the signal compatible with FM demod. If you were to plot the "constellation" of this, it would look like 8 clusters of dots around the unit circle (two QPSK constellations separated by pi/4 rotation).
 

W0RS

Member
Premium Subscriber
Joined
Dec 9, 2003
Messages
195
Location
Nixa, MO
Adding more features to the roaming option is still planned, but it is going to be a while.


I assume you mean there are 4 clusters of symbols on the circumference of a circle in their respective quadrants? If you see an actual circle drawn out that would indicate that there is a constant amplitude with the frequency error being too large to decode properly. (symbol phase spinning around on the unit circle). The solution for this would be to make sure the reference frequency is correct.


Thanks for pointing that out. When you see the X, you know that would be a problem area for old style FM demodulation where the symbol decode decision is dependent on the absolute instantaneous phase at symbol time. With the P25RX (and I assume SDS, G4), where quadrature is used (I/Q), the symbol decode decision is the phase change between the previous symbol time and the current symbol time. (relative phase change). This way ,if the absolute instantaneous phase at symbol time moves around as received, the symbols will still be decoded correctly as long as the sync is good. The way CQPSK works is that you have a QPSK constellation (4 dots in quadrants) that is rotated by pi/4 every symbol time. This is what makes the signal compatible with FM demod. If you were to plot the "constellation" of this, it would look like 8 clusters of dots around the unit circle (two QPSK constellations separated by pi/4 rotation).
Yes, 4 clusters of symbols in each of the quadrants and the only time I see the circle is just before and after a transmission not during a transmission. The reference freq is always in the green as is the sync state....
 
  • Like
Reactions: btt
Top