New Smeter issue when Squelch set to 0 about uv-k5-firmware-custom HOT 21 CLOSED

The revised s-meter values are an improvement but still read too high on some bands.

-130dBm is a typical noise floor for this type of radio and that would be expected to coincide with S0. Removing the antenna should give readings close to those values on all bands.

I have the necessary test equipment, I will do some measurements on each band and test the firmware with suitable adjustments. I will then post the results here. There will be some variation between radios but I suspect they will not be that significant.

Colin G4EML

from uv-k5-firmware-custom.

I have now checked my radio's S-Meter display using a Marconi 2955 Communications test set.
I tested all bands with signals at -100dBm and -70dBm which are equivalent to a medium and strong signal strengths.

I then adjusted the firmware correction values to give the correct readout on the display. I also changed the definition of S0 to be -130dBm.

The S-Meter performance of my radio is now what I would expect. The dBm readout remains accurate to within a few dBs over the whole range and indicates approx -130dBm and S0 with the antenna disconnected.

What is unknown is how much variation there is between individual radios.

My new settings in ui/main.c are as follows...

const int8_t dBmCorrTable[7] = {
			-15, // band 1
			-25, // band 2
			-20, // band 3
			-4, // band 4
			-7, // band 5
			-6, // band 6
			 -1  // band 7
		};

		const int16_t      s0_dBm       = -130;                  // S0 .. base level

Colin G4EML

from uv-k5-firmware-custom.

The new s-meter is only a visual change, all the underlaying values are all the same. It all comes from the RSSI which is being read directly from BK4819.
dBm value is calculated as (rssi / 2) - 160 (don't know where it comes from), and the resulting value is used to set the s-meter as described here
To be clear, the underlaying code comes from 1o11, I'm not a radio amateur, I learn as I go, so I don't know many things, but..
It all seems wrong, the original s-meter in stock firmware is being calibrated at the factory and written to EEPROM. 4 levels are being calibrated for different values of RSSI. Each band 1-7 is calibrated separately.
Now the middle screen s-meter doesn't use calibration at all so it cannot be considered reliable I would say. I think the calibration needs to be used but how to use those 4 points to infer dBm that I don't know yet.

from uv-k5-firmware-custom.

I did change the algorithm for s-meter. I don't have many transmission around me but it seems that it works much better.
It takes first and last calibration point, it assumes it is S1 and S9 and does linear approximation from that.
Let me know how it works for you.
firmware.packed.bin.zip

from uv-k5-firmware-custom.

Yes the equation, is correct, it indeed comes from the documentation. The RSSIcal description from SI44xx sheds some light on the problem, however I don't have proper equipment, nor a necessary knowledge. There are some LNAs, I don't know which are external to BK4819.
The S-meter values were adjusted by feel in 6730746, which was discussed in #92.

Current build that includes the adjustment:
https://github.com/egzumer/uv-k5-firmware-custom/suites/18492876221/artifacts/1072961377

from uv-k5-firmware-custom.

Thank you for the measurements and for explanation. I pushed those changes, will be included in the next release.

from uv-k5-firmware-custom.

I hink its not a major isue.

from uv-k5-firmware-custom.

It turns out the RSSI levels are not calibrated at factory. They are written to eeprom but are the same on each radio. But the levels are different for 2 groups of bands. This was not taken into account in the s-meter

bands 3,4,5,6,7
lvl 1: 110
lvl 2: 120
lvl 3: 130
lvl 4: 140
bands 1,2
lvl 1: 180
lvl 2: 190
lvl 3: 200
lvl 4: 210

from uv-k5-firmware-custom.

https://github.com/egzumer/uv-k5-firmware-custom/assets/148666132/9652e7ff-029e-4aec-addd-259a941f9773
On 162.400 MHz it seems to have went tighter. Prior to this update (from the link you provided above), this NOAA signal was -80db with an S9 on the bar. This seems to have made it far less sensitive (at least what's being reflected on the display / meter.)

from uv-k5-firmware-custom.

Here's the same frequency on version .14
https://github.com/egzumer/uv-k5-firmware-custom/assets/148666132/8b48fa14-d942-4e27-8bef-06b7afabba1a

from uv-k5-firmware-custom.

As I said I'm not a HAM operator so I'm asking about opinion. Is it better or is it worse? Tighter, less sensitive means nothing.
The real solution would be if someone did measure full spectrum characteristic with a frequency generator or something.

from uv-k5-firmware-custom.

I couldn't see what the links are on the phone.
So now it says S0 like there would be no signal. So not too good, but this is on one frequency. I get better results on other frequencies, but it gets overdriven for high strength signals, when I transmit with the other radio it goes up to 200dBm :).
I have no reference on what it should be so this makes no sense unless someone do some measurements or comparison with other decent radio on multiple frequencies.

This version shows raw RSSI value. I need dBm values for different frequencies and different signal strengths
rssi.zip

from uv-k5-firmware-custom.

Maybe this article will be helpful: https://www.giangrandi.org/electronics/radio/smeter.shtml

from uv-k5-firmware-custom.

This is the same as wikipedia page I gave in 3rd post. I know this, the problem is what is the relation of RSSI to dBm. I don't have 50R load nor RF voltage meter.

from uv-k5-firmware-custom.

Related to this:

dBm value is calculated as (rssi / 2) - 160 (don't know where it comes from)

the problem is what is the relation of RSSI to dBm

Take a look at BK4819(V3) Application Note 20210428.pdf, page 13. The equation from first quotation seems to be correct. The pdf says nothing more about. However recently I was playing around with a SI44xx chip and I have found something interesting in their Si4464-63-61-60.pdf, page 31, the same equation is mentioned with a small change:

The RSSI value read by the API could be translated to dBm by the following linear equation:
RSSI (in dBm) = (RSSI_value /2) – RSSIcal
RSSIcal in the above formula depends on the matching network, modem settings, and external LNA gain (if
present). The RSSIcal value can be obtained by a simple calibration with a signal generator connected at the
antenna input. Without external LNA, the value of RSSIcal is around 130 ±30.

Does it put a bit of light to the case?

from uv-k5-firmware-custom.

4pda's user says that there is correct S-meter in 15th and 17th releases of UV_K5_playground firmware. I have tested custom FW based on it and I can say that S-meter works properly. I'm not a programmer myself. Is it possible to get a part of code responsible for S-meter from that firmware?

from uv-k5-firmware-custom.

Why S0 changed to -130?
Sources I read mention -141 for S1 for frequencies above 30 MHz. S0 is presumed to be -6dBm lower hence it was -147.

from uv-k5-firmware-custom.

from uv-k5-firmware-custom.

Great work with the firmware!
Is there a possibility to add a VOX DELAY option to the menu? I am looking to use this radio with VOX and a delay adjustment would be really useful. Thank you.

from uv-k5-firmware-custom.

Can I make a few suggestions for additions to the firmware, if they could be implemented.

Add:
VOX DELAY
STARTUP ON FM RADIO (VHF FM) default boot option
S-Meter on FM RADIO
MIC +20db (some speaker mics slightly require more gain)

Thanks Everyone

from uv-k5-firmware-custom.

@g4eml, you guys talk to each other in #174

from uv-k5-firmware-custom.