wwv calibration and plus or minus specs

  • 2
  • Question
  • Updated 2 years ago
After reading a previous thread about WWV and Tim's response on the calibration specs, I was wondering what the range of specification is? I have noticed that the PPB amount that shows when I do the calibration has changed from when using v1.6 to v1.9. (The PPB # has gone down) I also see that it changes from different WWV frequencies. I calibrated last night on 5 Mcs and got a figure of 52 and then this morning I did a calibration on 10 Mcs and a figure of 92. At both times of calibration the signal strength of WWV was at least S 9 or better. So why is there such a difference in the calibration figure? 

One other question...how do I calculate the actual frequency offset by the calibration error number? An example would help!
Photo of Rick  WN2C

Rick WN2C

  • 211 Posts
  • 29 Reply Likes

Posted 2 years ago

  • 2
Photo of WA2SQQ

WA2SQQ

  • 394 Posts
  • 84 Reply Likes
I might be wrong, but I think I recall reading somewhere that the WWV calibration should be done at the highest frequency you can reliably receive. Hopefully someone who is more knowledgable can respond.
(Edited)
Photo of Bob Brown - N8OB

Bob Brown - N8OB

  • 89 Posts
  • 6 Reply Likes
The only way to know the correct frequency is with measurement using a frequency standard.  The auto frequency calibration procedure will put you very close but to know the exact frequency requires specialized equipment.  I participate in Frequency Measurement Tests and in the last FMT, 4 out of my 5 readings were within less than 1 hz from a measured standard.  So there is no way to calculate the offset from the calibration error number
Photo of Michael Coslo

Michael Coslo

  • 753 Posts
  • 172 Reply Likes
Isn't this what we call being accurate to several insignificant digits? 8^)
Photo of Bob Brown - N8OB

Bob Brown - N8OB

  • 89 Posts
  • 6 Reply Likes
That's why they call us FMT Nuts
Photo of Michael Coslo

Michael Coslo

  • 753 Posts
  • 172 Reply Likes
I thought that was Fix My Transceiver?
Photo of Bob Brown - N8OB

Bob Brown - N8OB

  • 89 Posts
  • 6 Reply Likes
Would you want a nut fixing your transceiver?
Photo of Michael Coslo

Michael Coslo

  • 753 Posts
  • 172 Reply Likes
Well, theres always been a nut operating - me!
Photo of Lee - N2LEE

Lee - N2LEE

  • 268 Posts
  • 138 Reply Likes
On a semi related topic. Does the 6300 support 10mhz reference on start up ?
Photo of k3Tim

k3Tim

  • 817 Posts
  • 155 Reply Likes
That's a good question..  From the H/W manual:

http://www.flexradio.com/downloads/flex-6000-hardware-reference-manual-pdf/

Page: 27, paragraph 7.5  sub-paragraph 7.5.1   (verse II, except on Tuesday....)

I read that as NO for 6300. The 14 pin connectors for 6300/6500/6700  are documented in this area of the manual and it not identical pin for pin operation there should be a table with a check box as to the feature described being available on said model.


Tim
Photo of Rick  WN2C

Rick WN2C

  • 211 Posts
  • 29 Reply Likes
Well I am not sure my Question was answered so again I ask what are the plus / minus specs for calibration? Tim in a previous post said this...

"A correction value of -6128 ppb is within spec and I presume the radio is on frequency so I wouldn't worry too much about it ."

That number seems like alot to me. So what would be the plus side of the PPB number?
If I am getting a figure of +52 PPB on 5 Mcs and +92 PPB on 10 Mcs, this is good. But why is there a difference between the 2 bands especially if the signal level of WWV is approximately the same?
Does this mean that the receiver is not as accurate the higher you go in frequency?
I know in some other post somewhere on here it was explained how to use the figure to determine how close the receiver is to actual displayed frequency using the PPB number.

Tim please way in!
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9031 Posts
  • 3431 Reply Likes
That number seems like alot to me. So what would be the plus side of the PPB number? 

If the calibration value is out of spec, the software will inform you.

If I am getting a figure of +52 PPB on 5 Mcs and +92 PPB on 10 Mcs, this is good.

The value of the offset is not indicative of anything good or bad.  As long as the frequency calibration routine generates an offset value, then all is good. This is a parameter that is used to compensate for frequency variance that exists in all oscillators.  Using a fixed reference (WWV) we run a routine that corrects for the variance to ensure that the radio's frequency accuracy is within specification (1 Hz) .  In general, using our frequency accuracy routine can achieve a frequency accuracy that is an order of magnitude better (several tenths of a Hz).

But why is there a difference between the 2 bands especially if the signal level of WWV is approximately the same? 

Because the calculated offset is dependent on the reference frequency.  You should notice a relationship between the 2 sets of values.  A doubling of the reference frequency produced an offset that is approximately doubled.  Since a WWV reference signal is perturbed by several factors, like multi-pathing and Doppler effects, it can vary which interjects error into the measurements.

I know in some other post somewhere on here it was explained how to use the figure to determine how close the receiver is to actual displayed frequency using the PPB number.

You don't use the offset to determine the frequency accuracy of the radio.  It is an oscillator correction value.

The bottom line is, run the frequency correction routine.  Use the default value for WWV (10 MHz).  If you do not get any errors displayed after running the routine, then you are good to go.  Your radio is frequency calibrated within spec and you need not worry about it anymore.
(Edited)
Photo of k3Tim

k3Tim

  • 817 Posts
  • 156 Reply Likes
To check accuracy : 

Fire up fldigi in Freq Analysis mode and check "how accurate WWV is".  At this station a 6500 with internal GPS shows WWV on 10.000Mc  bouncing around about 0.1 cycle (not bad). The error, as Tim points out is propagation and I would also suspect sound card/ CPU clock.  Fldigi is writing a CVS file and the values, I would hope / suspect, will sum to 10.000Mc although the result was:

  9999999.9815 cycles or a 0.0185 cycle (20ppb) error .  Chock that up to CPU clock in PC (audio channel of DAX) and/or propagation.  Not bad. 

Makes running the FMT pretty easy.

Best,
k3Tim/6