Welcome, Guest
Username: Password: Remember me
Please offer suggestions via our GitHub page - github.com/ExpertSDR3/ExpertSDR3-SUGGESTIONS
Before publishing a suggestion, please check that the same idea was not suggested before.

TOPIC: What's the true resolution of the frequency correction calibration coefficient?

What's the true resolution of the frequency correction calibration coefficient? 24 Mar 2023 01:18 #1

  • K3CDY
  • K3CDY's Avatar
  • OFFLINE
  • Posts: 12
  • Thank you received: 3
In preparation for the April 2023 ARRL Frequency Measurement Test, I executed a frequency calibration of my Sun SDR2Dx using an app called "Spectrum Lab" and the WWV time standard. I am running ESDR3 with the latest firmware, at 312 kHz sample rate, and no external frequency reference.

During my adjustments of the frequency correction coefficient, I noticed that very small changes (for example, 0.000210 to 0.000205) made no measurable difference in calibrating the radio.

In ESDR3 settings, the coefficient is displayed as a value with an accuracy to 10e-7 (i.e., 0.000001), but my adjustments seem to indicate this is not the true level of precision, and perhaps the actual adjustment precision is significantly more granular.

Can anyone at EE comment as to the true accuracy of the coefficient and how it really works?
73
Bill
K3CDY
The administrator has disabled public write access.

What's the true resolution of the frequency correction calibration coefficient? 24 Mar 2023 14:18 #2

  • Rome
  • Rome's Avatar
  • OFFLINE
  • Quod Licet Jovi Non Licet Bovi
  • Posts: 291
  • Thank you received: 140
Hello.
The correct way to adjust frequency is described in the UM, eesdr.com/images/Document/SunSDR2_DX/SunSDR2_DX_User_Manual_ENG.pdf
Page 66.
The main idea is that you should set sample rate 32 kHz, max zoom, with as high frequency as possible, preferably with high precision signal generator, closer to 80 MHz.
Роман, Roman
Expert Electronics
The administrator has disabled public write access.

What's the true resolution of the frequency correction calibration coefficient? 24 Mar 2023 14:41 #3

  • CASDR
  • CASDR's Avatar
Rome wrote:
Hello.
The correct way to adjust frequency is described in the UM, eesdr.com/images/Document/SunSDR2_DX/SunSDR2_DX_User_Manual_ENG.pdf
Page 66.
The main idea is that you should set sample rate 32 kHz, max zoom, with as high frequency as possible, preferably with high precision signal generator, closer to 80 MHz.

That is NOT what he was asking,

Using the coefficent he is asking what the relationship of the coefient cumber mean in regards to frequency

Example if using coefient of say 0.0000112 ( that .000112 means what in terms of how many KC in change was made in terms of a 0.000000)
Cody I think this is what your asking corect? does 0.000100 mean 1 kHz or 10 or 100 -please explain Rome

And yes your correct a low number like 50/ 80 etc won't make much if any change, I didnt noitce a change when doing this a couple years ago until I hit over 100 in the last 3 digits, and found that 0.000112 put me spot on frequency

I also DISAGREE with Rome's comment, about using as high freq as possible- that means your LOW frequencice will be inaccurate

If you chose a frequency closer toward middle of the bands (say 20/40 meters) you'll find that your accuracy is more even across all the bands,

If you choose 10 meters for example then yes 6-30 maybe40 are good and fairly accutate but 60M+ is not going to be.

I also disagree with using the lowest kHz rate, I would think that a higher resolution will provide a more accurate outcome, as well as using the tightest skirting as possible (Think filter taps)

Using 3,000 for taps and a 312 Khz sample rate- when I use 20 or 40 meters, I can by ear adjust to damn near spot on, but it takes a good ear and time.
When I got my 2DX it was bad and far off, a good ear can easily tell 1-2 Khz of difference and it drives me nuts.

Im sure I'll get flamed for the above in disagrement and thats fine, I use my ears and compare that to test equipment and the above has given damn near spot on results every time. Not only with the 2DX but other SDR rigs, Flex- Annan, Hermes, SDR kits etc.
The administrator has disabled public write access.

What's the true resolution of the frequency correction calibration coefficient? 24 Mar 2023 17:35 #4

  • K3CDY
  • K3CDY's Avatar
  • OFFLINE
  • Posts: 12
  • Thank you received: 3
Rome - thank you for your response. I will try the procedure described in the user manual, and see if minor changes in the frequency coefficient make any difference.

CASDR is correct in clarifying my question. That is, I am curious about the relationship of the frequency coefficient value to the displayed VFO frequency. This is definitely not a high priority question. I am curious about the theory of the frequency coefficient adjustment and what other factors influence adjustment (e.g., reference frequency, sample rate, number of taps, etc.). Looks like I'll be doing some general research on SDR's to see what more I can learn.

ALSO - THANK you Rome for the link to the new user manual for ESDR3! I had no idea one was even available at this time!
73
Bill
K3CDY
The administrator has disabled public write access.

What's the true resolution of the frequency correction calibration coefficient? 24 Mar 2023 19:52 #5

  • CASDR
  • CASDR's Avatar
Thats no new manual for V3 just the header and some of headings, if you look it over its a V2 manual- with an edit here or there like the title page
Last Edit: 24 Mar 2023 19:52 by CASDR.
The administrator has disabled public write access.

What's the true resolution of the frequency correction calibration coefficient? 24 Mar 2023 20:22 #6

  • K3CDY
  • K3CDY's Avatar
  • OFFLINE
  • Posts: 12
  • Thank you received: 3
Gotcha. Well at least they're making progress in that direction...
73
Bill
K3CDY
The administrator has disabled public write access.

What's the true resolution of the frequency correction calibration coefficient? 25 Mar 2023 00:03 #7

CASDR wrote:
When I got my 2DX it was bad and far off, a good ear can easily tell 1-2 Khz of difference and it drives me nuts.
Im sure I'll get flamed for the above in disagrement and thats fine, I use my ears and compare that to test equipment and the above has given damn near spot on results every time. Not only with the 2DX but other SDR rigs, Flex- Annan, Hermes, SDR kits etc.

One has to have very bad ears to hear only 1-2 Khz frequency deviation. For me, this is already possible with a deviation of 1-3 Hz. And who If confuse the MIC input with the LAN socket, even an up-to-date instruction manual is of no use. Anyone who uses high-tech equipment should have a certain basic knowledge.
The administrator has disabled public write access.

What's the true resolution of the frequency correction calibration coefficient? 25 Mar 2023 18:20 #8

  • CASDR
  • CASDR's Avatar
Typo I meant Hz in the above , not kHz when listening

Yes you can hear 1-2 Hz with a good ear..

I wouldnt say the 2DX is High TECH, the design has been around a while,
for SDR's software will make or kill a radio, and lately this is stuck in the middle
Sure EESDR layout is easy to use and looks nice, But features arent there compared to oter offerings, then ANAN has a crappy layout but all the functions + are there
Flex has piss poor customer service, and doesnt listen to its users
Last Edit: 25 Mar 2023 18:24 by CASDR.
The administrator has disabled public write access.

What's the true resolution of the frequency correction calibration coefficient? 28 Mar 2023 13:43 #9

CASDR
Good synthesis on the SDR
I totally agree!
Last Edit: 28 Mar 2023 13:46 by F6CTT.
The administrator has disabled public write access.

What's the true resolution of the frequency correction calibration coefficient? 31 Mar 2023 01:49 #10

  • CASDR
  • CASDR's Avatar
Rome, were still curious as too the resosution

For ever 1kHz change requires what offset in the coeffient
The administrator has disabled public write access.

What's the true resolution of the frequency correction calibration coefficient? 31 Mar 2023 07:36 #11

  • Rome
  • Rome's Avatar
  • OFFLINE
  • Quod Licet Jovi Non Licet Bovi
  • Posts: 291
  • Thank you received: 140
Hello everyone.
I've discussed this with the programmer, so the following information is coming from the source.
Frequency coefficient is a multiplier, the higher the coefficient, the higher the deviation.
For example, based on the 1,8 MHz with multiplier 1.001 the deviation will be: 1'800'000 * 1.001 = 1'801'800
Based on the 144,100 MHz: 144'100'000 * 1.001 = 144'244'100
The higher the calibration frequency, the higher the adjustment precision.
Роман, Roman
Expert Electronics
The administrator has disabled public write access.
The following user(s) said Thank You: K3CDY, CASDR

What's the true resolution of the frequency correction calibration coefficient? 31 Mar 2023 11:07 #12

A self or auto calibration feature would be nice like the other SDR brands have would be verry nice!
Transceiver: MB1 Prime 2021
Transceiver: SunSDR2 dx
GPSDIO: Leo Bodnar 10mhz osc
HF Amplifier: Ameritron AL-1200
HF Antennas:10-80 Vertical, 10-80 Fan Dipole
VHF Antenna: Commet GP-9
Last Edit: 31 Mar 2023 11:07 by John-KC2QMA.
The administrator has disabled public write access.
The following user(s) said Thank You: Rome

What's the true resolution of the frequency correction calibration coefficient? 31 Mar 2023 19:09 #13

  • K3CDY
  • K3CDY's Avatar
  • OFFLINE
  • Posts: 12
  • Thank you received: 3
This is precisely the information I was looking for. This detail should be included in the new user manual for ESDR3. Thank you Rome!
73
Bill
K3CDY
The administrator has disabled public write access.
The following user(s) said Thank You: Rome

What's the true resolution of the frequency correction calibration coefficient? 04 Apr 2023 20:14 #14

John-KC2QMA wrote:
A self or auto calibration feature would be nice like the other SDR brands have would be verry nice!
I Should explain the way the other brands implement the “Auto Calibration” feature is to have the transceiver go to WWV on one of there frequencies and the radio checks it Oscillator against the precise WWV signal. Then it comperes the difference and applies to the coefficient offset to put the receiver exactly on frequency.
Transceiver: MB1 Prime 2021
Transceiver: SunSDR2 dx
GPSDIO: Leo Bodnar 10mhz osc
HF Amplifier: Ameritron AL-1200
HF Antennas:10-80 Vertical, 10-80 Fan Dipole
VHF Antenna: Commet GP-9
Last Edit: 04 Apr 2023 20:15 by John-KC2QMA.
The administrator has disabled public write access.
The following user(s) said Thank You: K3CDY

What's the true resolution of the frequency correction calibration coefficient? 04 Apr 2023 21:14 #15

  • K3CDY
  • K3CDY's Avatar
  • OFFLINE
  • Posts: 12
  • Thank you received: 3
Great idea John. For the best chances of getting this implemented, I suggest you go to GitHub and post this as a suggestion. The software developers definitely keep an eye on GitHub, but generally don't follow the forum posts.

github.com/ExpertSDR3/ExpertSDR3-BUG-TRACKER
73
Bill
K3CDY
Last Edit: 04 Apr 2023 21:16 by K3CDY. Reason: Clarification
The administrator has disabled public write access.
Time to create page: 0.139 seconds