P25RX-II P25RX-II Firmware Updates / Testing
- Thread starter maus92
- Start date
- Status
I received a report that testing version (11-16_2303) was a step in the wrong direction for mobile operation. I suspected that might be the case as I have been busy working on code and not doing a lot of mobile testing the past couple of days. If the version you are running is working well, then please wait for the next release. If you are running version 11-16_2303, you may want to revert back to the experimental version 11-15_0731. This version is known to work well in most environments including mobile. I'm currently working on rewriting large portions of the decoder based on what I'm seeing from the I/Q files sent in by those with receive P25 receive issues. It is very clear that the issue is firmware. Feel free to send more I/Q files if you are having issues. As the code is re-written, changes will be verified to maximize performance across all available I/Q files. Thank you!
Yes, I confirmed that _2303 was not good at all on a drive this morning. It turned out to be due to a change that caused intermittent interruptions to the ADC sample buffering. I received some mobile I/Q samples today. Thank you! I'm making progress on generating statistics for all available I/Q files. It has already enabled improvements to the symbol synchronization and symbol decisions. I'm working on improving framing / decoding now.
2021-11-20 07:01 Jittering Froze go to advanced tab and reset to defaults seems to fix it
Testing version _0843 installed very smoothly with none of the prior hiccups with TG Loads or failure to discover device. After install, Alias Tab and Advanced Tab settings had been returned to default. Sounds great
freqseeker
Member
Big improvement here. I monitor the WSP700 Tacoma simulcast site of this system WSP 700
I let Todd know that all previous versions with both the RX and RX-2 worked with good audio. This system has a very good IQ plot, but the BLKS only run in the high teens to low twenties. Other systems run in the mid 50's for comparison.
On the latest release 11-20_0843 the WSP system runs in the mid 50's! This has never happened with this sytem. Audio is great so far.
I use this stationary on a external ST2 antenna so can't comment on mobile decode.
Great job Todd!
I let Todd know that all previous versions with both the RX and RX-2 worked with good audio. This system has a very good IQ plot, but the BLKS only run in the high teens to low twenties. Other systems run in the mid 50's for comparison.
On the latest release 11-20_0843 the WSP system runs in the mid 50's! This has never happened with this sytem. Audio is great so far.
I use this stationary on a external ST2 antenna so can't comment on mobile decode.
Great job Todd!
I haven't been able to verify P25 P2 improvements yet, but please update to version 2021-11-20_1446 (even if you already updated to versions from earlier today). I can guarantee that P25 P1 is working better than ever! Much of the code for P25 has been rewritten. One thing that has become clear from analyzing the IQ files that have been sent in, is that the hardware is doing great. The IQ file data all looks good. The other thing that became apparent from looking at the stats of all the IQ files is that the largest issues was when frame sync words were not detected. In addition to improving frame sync detection and symbol synchronization in general, it appeared that it might be worth trying to detect missing syncs and try to recover. The cool thing is that this actually works! I haven't been able to test P25 P2 with the latest release yet. DMR will benefit from the improved sync detection as well. Looking forward to hearing how it does for you.
No, I haven't had the time to port these massive changes to the P25RX yet. I will, but I need to take a break for a bit. It has been a coding marathon for days now!
In my opinion, you’ve hit it out of the ballpark with the 2021-11-20 version for the P25RX-II. P1 and P2 performance sounds amazing from a fixed location. I think this offers the biggest leap in performance ever from a single release. Have yet to test in a mobile environment.
Thank you for testing and the feedback! Even with the improvements, some people are still having some issues. It appears that some of the changes I made to the sync detection yesterday weren't a good thing for everyone. Today, I confirmed that by running the stats on all the IQ files sent in (more than 40 files). The latest release 11-21_0734 should be better. Currently, I can listen to the decoded audio for P25 P1 from the submitted IQ files, but not P2. Today I hope to get the P2 audio output working. This will give me a better idea of what P2 sounds like for everyone submitting files. Note that the logic is a bit different from the firmware running on the receiver because the receiver has to actually decode the control channel, switch frequencies and sample rates, etc. As I make changes to this, I'm going to leave various versions up as I go for the next few releases. We are definitely making progress.
I certainly concur with all these comments about the continuous improvement. Every time I think it's sounding really great, it gets better. I don't really have signal issues since I'm in a fixed location only 2 miles from a high profile site for a very large P25 system. And my location is at the "intersection" of the 3 large counties making up Metro Detroit so I'm easily able to pick up just about everything.
Could someone tell me what EVM_P means?
Error-vector-magnitude (percent). It can also be represented in dB.
Thank you
2021-11-21_1328 is available. If you have a large number of talk groups and you noticed that reading/writing those was painfully slow in recent releases, that has been fixed. I was able to get actual audio output for P2 working when analyzing the IQ files. Quite a few changes were made and everyone's IQ file playback sounds good now (P1 and P2) except for one person having extreme issues. Two new values were added to the advanced page (and quite a few removed). The new values, P1/P2 sync thresh, allow you to specify how many bit errors are allowed during frame sync detection. The values can range from 0 to 6. P1 is 4 by default and P2 is 0. If you are one of the people having lots of reception issues, then set them both to 0. Experiment with these values and report if you find something that works better. These values can make a difference! There were many more internal changes related to getting the best performance audio possible for all the IQ files sent in.
- Status
Similar threads
BCD325P2/BCD996P2:
BCD996P2-Bad data cable
