BER Bit Error Rate and Hotspot Mode OpenGD77

Post Reply
User avatar
kd2lh
Posts: 312
Joined: Mon Dec 02, 2019 2:44 pm

BER Bit Error Rate and Hotspot Mode OpenGD77

Post by kd2lh » Mon Jan 27, 2020 3:28 pm

In speaking with Roger yesterday, he mentioned a detail about the BER being reported to PiStar by the OpenGD77 hotspot mode.

The BER is not an average over the entire conversation, but rather the last BER calculated for the last RF packet handled inbound by OpenGD77. It's the last BER reported by the C6000.

Since Automatic Frequency Control AFC is implemented on the GD77 radio, it compensates for the small few hundred cycle difference between various handheld radios. It's not necessary to use the calibration utility to create RXOffset or TXOffset adjustments with OpenGD77 in HotSpot mode. The exact frequency entered into PiStar is fine.

If you've used the MMDVMCAL utility, you know that instantaneous BER varies a great deal from incoming packet to packet. Just remember that the BER being reported to PiStar is not the average of all voice packets in the conversation, but just reflects the last one.

kt4lh
Posts: 272
Joined: Sun Jan 12, 2020 4:27 am

Re: BER Bit Error Rate and Hotspot Mode OpenGD77

Post by kt4lh » Mon Jan 27, 2020 5:09 pm

For my two radios I just use the mmdvmcal program on PiStar, documented both RX and TX values where the signal falls out, did the math to find the center frequency. I was lucky in that the RX (TX on the hotspot) was 0 interestingly, I had it at 400 previously, and the TX (RX on the hotspot) was 100. I think I had these two backwards before, oops! But it was still well within working range.

But for mine, I found that the extremes were about a total of 1800hz, so that's a pretty wide range that work. Divide it by two, add it to the bottom, and get the 'ideal' spot and for me that wound up being almost dead on, as mentioned above. But I've also heard some mmdvm's have the offset pre-set in the firmware, but I've updated the firmware a few times on it and it has not changed so I dunno.

So I still think the best way to minimize dropouts and BER are doing both RX and TX tests on PiStar and if you have multiple radios, then average a bit to get a spot that's good for both radios. Use low power, I even did all the testing that I could do a few feet from the hotspot, don't have them right beside each other.

http://www.k9npx.com/2019/02/hotspot-of ... n.html?m=1 has really good instructions on calibration in PiStar.

That all said, my BER has always been pretty good, but my tests above also demonstrate that the RX side on the hotspot matched very closely with my radios, so I shouldn't have had any BER issues. Interesting though, what you say about the frames. I swear I've noticed that quick key/unkey the BER is unreliable, but hold it for a few seconds and it is. But maybe a quick key/unkey isn't enough data in the buffer.

User avatar
kd2lh
Posts: 312
Joined: Mon Dec 02, 2019 2:44 pm

Re: BER Bit Error Rate and Hotspot Mode OpenGD77

Post by kd2lh » Mon Jan 27, 2020 9:59 pm

With the OpenGD77, one of the advantages is the use of a fully implemented receiver and transmit exciter. The receiver has AFC automatic frequency control implemented in the GD77 itself, and it should accommodate small receive offsets from a variety of HTs.

Roger says that the use of offsets should not be necessary when setting up a GD77 based hotspot.

The BER is informational only, and it turns out it is not an average over the conversation, but just the BER calculated on the last received voice frame.

kt4lh
Posts: 272
Joined: Sun Jan 12, 2020 4:27 am

Re: BER Bit Error Rate and Hotspot Mode OpenGD77

Post by kt4lh » Tue Jan 28, 2020 4:15 am

Welp, I'm an idiot, I missed the context of the GD77 being the hotspot. Oops.

Post Reply