Required Calibrations - am I overlooking anything?

[JJ4SDR]

Thanks to Doug's patience and dedication I am looking forward to getting on the air again with no anomaly in the noise floor that I had reported on very extensively in an earlier posting. The Orion board of my 7000 DLE MKII was replaced (the PA/Filter board was replaced too, months ago) by Doug so I am starting from "scratch" in a way.

After I asked Doug about the difference in noise floor that may result from my radio having a replacement Orion board in it he responded by writing":
"There might be a few dB difference in the noise floor. They [he is talking about the 2 Orion boards] were within a few dB of each other. You would need to calibrate the amplitude if you have a sig gen available. Those boards used a couple of different preamps to drive the adc. I recall the idea was to try and reduce some distortion."

Now, apart from TX Power and Frequency and Level Calibrations, are there any other calibrations that I may have missed bringing up?

I guess I will need to dive back into the Manual to make sure that what I am reporting back to him early next week really aren't "issues", after all these replacements.

Juha
NI2M

[JJ4SDR]

S meter calibration is an important parameter that I have never done before, but I will be getting help by someone who has done it before, using a signal generator. I have watched Richie's brief video on how to do this.

I would imagine that even if you do a good job at establishing S9 level that the scale isn't linear so reading below and above the S9 level may be off quite a bit. And, then there is the frequency dependency too, but my friend is fully aware of this of course.

Presumably, the calibration needs to be done for all the bands as well.

[w-u-2-o]

S meter calibration is an important parameter that I have never done before, but I will be getting help by someone who has done it before, using a signal generator. I have watched Richie's brief video on how to do this.

The procedure is simplicity itself:

- Attach signal generator. Ideally this will be set to a frequency that is near mid-band, e.g. 15MHz, and at an input level that is near the top of the ADC dynamic range, e.g. -20dBm.

- Set 0dB atten, temporarily increase FFT bin width to maximum.

- Use USB mode, set receiver passband to 500Hz

- Make sure the receiver is tuned so that the cal signal is in the middle of the passband (250Hz off from the cal. freq.)

- On the Calibration tab set the freq. and level being used. Push start.

- Don't forget to set FFT bin width back to where it was.

That's it!

I would imagine that even if you do a good job at establishing S9 level that the scale isn't linear so reading below and above the S9 level may be off quite a bit.

That's not at all true. Receive power measurement on our radios is very linear. And S9 is not where you want to calibrate, you want to calibrate near the top of the ADC dynamic range as mentioned above.

And, then there is the frequency dependency too, but my friend is fully aware of this of course.

Presumably, the calibration needs to be done for all the bands as well.

I have not investigated how flat the response is, but there is only a single calibration value in the software and it is used on all frequencies/bands.

[JJ4SDR]

Scott,

Thank you for your consistent support and a detailed response .

The ANAN never ceases to amaze me.

Juha

[JJ4SDR]

Scott,

I will be unpacking a frequency analyzer that features a reasonably good signal generator tomorrow.

In the run-up to conducting the actual level calibration, when you write " temporarily increase FFT bin width to maximum ", are you talking about
the "Slider" that governs the FFT "size"? If I max out on the slider the BIN width (Hz) is reduced.....Oh, I believe I addressed my own concern here......I need to move the slider all the way to the left for max bin width. No need to mess with the "Window" (it is set to Hann now). Will look into the window business as I don't really understand that concept and how it affect the radio and in which way it does that.

Juha

[JJ4SDR]

Why is it that when I inject -20dBm into RX antenna (RX2), I see the signal in RX1? Also, the S meter indicates a little shy of +50dB signal although -20dBm is equivalent to +53dB/S-9.

Update:
Disabled VAC1 and VAC2 and the S meter now corresponds pretty well with +53dB. Might as well use -23dBm level for S9/+50dB which should be good enough. However, I am still seeing the -20dBm signal in RX1 (injecting signal into RX2 BNC) not RX1.

Juha

[JJ4SDR]

ANAN GUI CALIB.jpg (667.9 KiB) Viewed 13999 times

ANAN DISPLAY CAL.jpg (119.01 KiB) Viewed 13999 times

ANAN CAL Form.jpg (95.52 KiB) Viewed 13999 times

Signal generator output is 14,999.75 MHz

[JJ4SDR]

[w-u-2-o]

:?

What issues are you having?

I would never calibrate frequency using a signal generator unless I was absolutely certain that sig. gen. was itself calibrated by a professional calibration lab to <<1Hz. Use WWV instead. Or go all-in and get a GPSDO.

[JJ4SDR]

Scott,

I am NOT calibrating for frequency, but I am calibrating the S meter ("Level Cal").

In the mean-time, I had RX2 tuned to the wrong frequency which explains why I didn't see the signal in the Panadapter although I injected a signal into RX2.

The Signal generator (TinySA Ultra) compares surprisingly well with MUCH more expensive pro gear.

Looked at the Thetis database (Database.xml) after calibration and noticed that the "Cal" figures, "display_cal_offset" and "meter_cal_offset" were much higher for both RX1 and RX2 than they were before! I suppose this is normal as I had never performed the Level (S meter) before and both the Orion board and PA/Filter boards were replaced by Doug.

UPDATE: All done and I am good. I understand that RX2 actually gets its cue from RX1 as the receiver chains are the same so there was no reason to mess with RX2 in the first place.

Juha

[w-u-2-o]

So all is ok then?

[JJ4SDR]

Yup.

Well, I thought I was OK, but now that I look at my TX with the DUP on, I see a MUCH larger RF signature with the max of my TX being off the screen. Never saw that before. PS works as it should, but I am seeing a bottom that is VERY wide. Nobody will see it on the air though, but this bothers me.

PS skirts.jpg (657.48 KiB) Viewed 13857 times

I guess I will need to adjust display settings.

Update: Found setting for TX Grid !

Juha

[n2gq]

Scott,

In the run-up to conducting the actual level calibration, when you write " temporarily increase FFT bin width to maximum ", are you talking about
the "Slider" that governs the FFT "size"? If I max out on the slider the BIN width (Hz) is reduced.....Oh, I believe I addressed my own concern here......I need to move the slider all the way to the left for max bin width. No need to mess with the "Window" (it is set to Hann now). Will look into the window business as I don't really understand that concept and how it affect the radio and in which way it does that.

Juha

Hi
Question does the fft size goes to min on the slider?

[w-u-2-o]

Set the slider so the largest bin size in Hz is obtained.

[n2gq]

Okay Thank you Scott.