TX power levels are not always correct on any releases after 2020.10.16

VK3KYY
Posts: 7484
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: TX power levels are not always correct on any releases after 2020.10.16

Post by VK3KYY » Fri Oct 30, 2020 5:30 am

KI5GZK wrote:
Fri Oct 30, 2020 4:55 am
Well, I know you are tired of hearing from me, and I hate to be the bearer of bad news all the time, but there is still an issue with the power levels.

For example:
Channel1 is set to 400Mhz at 1 watt.
Channel2 is set to 475Mhz at 1 watt.
These frequencies are at the bottom and top of the range used for calibrating the UHF band as I'm sure you know.

When I key up on channel1 it measures 1 watt as expected.
When I key up on channel2 it measures 0.5 watts.
If I power cycle the radio while still on channel2 and then key up, it now measures 1 watt as it should.
If I key up on channel1 again it now measures 1.5 watts.

WTF :?
I'm convinced that my GD-77 hates me, and I'm beginning to think you might too!
Thats OK.

I assume you calibrated the radio and looked at the calibration table values and they all looked OK, i.e the calibration value should not change a lot between each 5Mhz segment.

I'll try power cycling the radio, because it sounds like something is not being initialised correctly

Also, so it have this problem on frequencies well within the ham bands e.g 430Mhz and 440Mhz, or is it only on the limits of the calibration range where this is occurring ?

VK3KYY
Posts: 7484
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: TX power levels are not always correct on any releases after 2020.10.16

Post by VK3KYY » Fri Oct 30, 2020 5:41 am

I've tested my radio on 2m and the 70cm band, and I can't replicate those issues, i.e turning the radio on and off, on different channels with the same power and different power and also turning on and off on the VFO, and also changing from the VFO to the Channel and also changing channel after powering up, when the 2 channels have the same or different power levels.

I'll have to make a special codeplug with your configuration, because I just can't replicate the bug under normal operating conditions.

I presume you have build a codeplug, just to make it easy to calibrate the radio.

VK3KYY
Posts: 7484
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: TX power levels are not always correct on any releases after 2020.10.16

Post by VK3KYY » Fri Oct 30, 2020 5:55 am

OK.

Arrghhh

I over optimised some code, which resulted in the calibration data not getting read under some circumstances.

i.e the circumstance you found.

Try this fix

https://github.com/rogerclarkmelbourne/ ... 2020.10.30

(also available via the CPS)

KI5GZK
Posts: 129
Joined: Sat Apr 11, 2020 6:27 pm

Re: TX power levels are not always correct on any releases after 2020.10.16

Post by KI5GZK » Fri Oct 30, 2020 6:19 am

VK3KYY wrote:
Fri Oct 30, 2020 5:30 am

Thats OK.

I assume you calibrated the radio and looked at the calibration table values and they all looked OK, i.e the calibration value should not change a lot between each 5Mhz segment.

I'll try power cycling the radio, because it sounds like something is not being initialised correctly

Also, so it have this problem on frequencies well within the ham bands e.g 430Mhz and 440Mhz, or is it only on the limits of the calibration range where this is occurring ?
My calibration settings on the VHF band range from 66 to 76 at 1 watt, and from 136 to 163 at 5 watts.
My calibration settings on the UHF band range from 74 to 87 at 1 watt, and from 172 to 182 at 5 watts.

The closer together the frequencies are together the smaller the difference. When I tested the entire range of 16 frequencies on the UHF calibration table, the power output slowly dropped from 1 watt at the low end to .5 watts at the high end.
This issue also seems to affect the VHF band to a lesser degree.

When I test using the 2020.10.16 firmware, the power output is consistent across the entire range.

VK3KYY
Posts: 7484
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: TX power levels are not always correct on any releases after 2020.10.16

Post by VK3KYY » Fri Oct 30, 2020 6:21 am

Did you try the latest bug fix version ?

KI5GZK
Posts: 129
Joined: Sat Apr 11, 2020 6:27 pm

Re: TX power levels are not always correct on any releases after 2020.10.16

Post by KI5GZK » Fri Oct 30, 2020 6:39 am

Okay, I just finished testing. It looks like we have a winner! :D

I tested the entire range of calibration frequencies, both VHF and UHF at 1 watt and 5 watts.
The power level is very consistent through all ranges.
I can cancel my appointment with my psychoanalyst now!

Excellent work! You have saved my sanity!

VK3KYY
Posts: 7484
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: TX power levels are not always correct on any releases after 2020.10.16

Post by VK3KYY » Fri Oct 30, 2020 6:43 am

KI5GZK wrote:
Fri Oct 30, 2020 6:39 am
Okay, I just finished testing. It looks like we have a winner! :D

I tested the entire range of calibration frequencies, both VHF and UHF at 1 watt and 5 watts.
The power level is very consistent through all ranges.
I can cancel my appointment with my psychoanalyst now!

Excellent work! You have saved my sanity!
LOL.

No worries.

I'm relieved I've now finally killed all the bugs.

BTW. I just looked in the calibration table in my radio, and for some reason I seem to have set all the values to the same for 400 to 475
I think I must have been doing some testing of the calculations in the firmware, months ago..

So all my 1W settings are 108 and all my 5W settings are 217

These seem to give the correct power within the 70cm band, but are not good at 400Mhz, as I get 1.5W into my dummy load on 400Mhz.

When I get time, I may calibrate the radio again, but I'm never going to use it on 400Mhz, so its not a high priority

KI5GZK
Posts: 129
Joined: Sat Apr 11, 2020 6:27 pm

Re: TX power levels are not always correct on any releases after 2020.10.16

Post by KI5GZK » Fri Oct 30, 2020 6:50 am

Yes, it seems that at the extreme low and high ends of the range don't calibrate as well. It's probably due to the filtering in the RF section of the radio.

VK3KYY
Posts: 7484
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: TX power levels are not always correct on any releases after 2020.10.16

Post by VK3KYY » Fri Oct 30, 2020 6:55 am

KI5GZK wrote:
Fri Oct 30, 2020 6:50 am
Yes, it seems that at the extreme low and high ends of the range don't calibrate as well. It's probably due to the filtering in the RF section of the radio.
Yep. Its probably at the edge of the band pass filters.

I found the power output is also quite dependant on batter voltage, as the PA drive is a constant voltage high power OpAmp, driven via a DAC pin of the CPU.

In theory it may be possible to read the battery voltage and compensate the DAC output value, but each radio may behave slightly differently and I think its common on a lot of radios, for power output to decrease with battery voltage

Also, the power levels below 1W are just an approximation, as the PA drive to Output power varies considerably from radio to radio, as does the lower power threshold of the PA FET.

So getting within 10% - 20% is as good as I could achieve across a range of different radios.

The only way to have accurate power below 1W would be to have multiple calibration points below 1W. Probably a 50mW data point, but there is no space for this in the official calibration table data that is in the radio.

KI5GZK
Posts: 129
Joined: Sat Apr 11, 2020 6:27 pm

Re: TX power levels are not always correct on any releases after 2020.10.16

Post by KI5GZK » Fri Oct 30, 2020 7:26 am

Yeah, that's why I do most of my testing at 1 watt. It keeps the battery level higher longer.

My under 1 watt settings seem to be fairly accurate, at least according to my relatively inexpensive SWR/Power tester.

I've found the power levels between 1 and 5 watts to be less accurate on my radio. I actually calibrated it at 3 watts so that the 2 and 4 watt settings would be more accurate, and my 5 watt settings average around 4.5 watts.

Post Reply