[TowerTalk] Beverage impedance change in cold WX solved

Wed Jan 22 17:07:00 EST 2020

On 1/22/20 1:18 PM, VE6WZ_Steve wrote:
> Thanks Jim!
> 
> I didn't look at the Fair Rite catalog, but yes…its there on page 8.  I am using 73 material Binocular cores.
> 
> The temp. coefficient for 73 material is .65 % per deg C.  My temp change is about 50 deg C, therefore the permeability will change 33 % !!
> That seems like a decent amount.
> 

Go to page 28, and there's a curve of properties vs temperature. It 
looks like mu is 1500 at -40C and 2100 at 0C  (that's at 10kHz, where 
mu'' (the loss) is very low.

The impedance at 10 MHz changes from about 80% at -40 to about 90% at 0 
to a peak of 100% at 30C.

https://www.fair-rite.com/73-material-data-sheet/

unfortunately, they don't give you a way to download permeability vs 
temperature.

> I am going to do an experiment today.
> 
> Wind a 9:1 transformer, terminate it with 600 ohm resistor, and sweep with analyzer. Save file.  Should be pretty flat at 75 Ohms.
> Put the transformer in the freezer for an hour or so, take out and remeasure. (I wont cool the resistor)  Unfortunately my freezer dose not get to -41 deg C like it does outside, but the results may be telling.
> 
> Yes I could calculate it based on the catalog info, but that is above my pay-grade.
> 
> de steve ve6wz
> 
>> On Jan 22, 2020, at 12:10 PM, jimlux <jimlux at earthlink.net> wrote:
>>
>> On 1/22/20 10:23 AM, VE6WZ_Steve wrote:
>>> Hi Mike
>>> I have not considered the temperature coefficient of the core.  That is a great suggestion.
>>> I did find some generic plots of permeability vs temperature and there can be a big change.  The data sheet from Fair Rite does not quote any temperature coefficient values.  I’m not sure how to to test that easily other than heating the core in the field on a cold day.  Or perhaps cooling it with an air can upside down in the summer.
>>
>> Which material is the core made out of?  The Fairrite catalog has properties vs temperature data in it for some materials.  Pages 6 and 7 have the tempco of initial permeability  (1.6 %/C for 31 mix, 1.25 for 43 mix, 0.10 for 33 mix, for instance)
>>
>> The pages for 31 mix (page 17) show mu vs temperature from -50 to 150C, measured at 100 kHz (about a 1:3 variation from -50 to 25, BTW)
>> and percent impedance vs temperature at 4 frequencies. (10,25, 50, 100 MHz), where the loss is dominant.
>>
>> The other mixes are the same.
>> Some of the mixes that are more intended for power supply usage (79,77) have power loss density vs temperature at a couple frequencies, for instance
>>
>>
>>
>>> The other option is to sweep the raw wire in the field, but at -38C that would be rather difficult.
>>> I do plan to experiment with changing the termination resistor value this summer to see if I can replicate the cold weather sweep.
>>> 73, de steve ve6wz
>>>> On Jan 21, 2020, at 6:50 PM, Michael Tope <W4EF at dellroy.com> wrote:
>>>>
>>>> Steve,
>>>>
>>>> Have you eliminated the core material in the 9:1 transformer as a possible source of the problem?
>>>>
>>>> I don't see how the resistance of the galvanized wire changes the surge impedance significantly. When viewing the beverage as a transmission line (single round conductor over ground plane), I would expect the distributed resistance of the steel wire would to make the transmission line loss/unit length higher, but other than adding a slightly reactive component to the characteristic impedance, the steel wire resistance change shouldn't make the real component of the transmission line characteristic impedance change much.
>>>>
>>>> 73, Mike W4EF...............
>>>>
>>>> On 1/21/2020 3:55 PM, VE6WZ_Steve wrote:
>>>>> A few days ago I posted a video explaining some experiments I did adding short radials to my feed and termination grounds on the Beverage.
>>>>> During that week temperatures went as low as -41 C at the station.
>>>>>
>>>>> The mystery has been solved.
>>>>>
>>>>> I wondered if the frozen ground was limiting conductivity and upsetting the ground resistance part of the impedance. The experiments showed that adding radials to stabilize the frozen ground had no effect.
>>>>>
>>>>> In the video I show how the real resistance of the 920 foot galvanized steel wire might drop 10-20 Ohms when the temperature drops 50 deg C from the summertime.  This would mean that I need a LOWER termination resistor to match the surge impedance.
>>>>> However, I also show that the ceramic termination resistor I use has a very high negative temperature coefficient (-1300 C). This results in an INCREASE of  30 to 40 ohms in my 470 Ohm termination resistor.  Therefore, the termination could be 40 to 60 Ohms too low!!  I show in the video that the resistor does indeed increase by 40 Ohms when I chucked it out my back door at -30C for an hour.
>>>>>
>>>>> Today I was at the remote and it has warmed up to -1C (35 deg C warmer) and the analyzer sweep of the wire has mostly returned to it summer time reading.
>>>>>
>>>>> This seems to indicate that the main cause of the Beverage miss match is the temperature coefficient of the wire and the termination resistor. Obviously the ground is still just as frozen today as it was last week at the feed and termination and under the wire length!  Perhaps I will look into changing the term resistor with a metal foil unit that has a positive temperature coefficient.  Carbon composition units also have a negative TC, but not as severe as the ceramic units.  Those ceramic units are almost like thermistors! (well….not really)
>>>>>
>>>>> However, as I said in the video, this is like a “solution looking for a problem” because few Hams experience these extreme temperature changes, and even when we do, it's just for a short time.  Also, Beverage modelling shows that the termination resistor can effect the F/B, but it has limited effect on the RDF which is what really matters.  This exercise was done mostly out of interest and to try and solve the puzzle.
>>>>>
>>>>> I made a new video showing the sweeps I made today if interested.  The video includes the original content so if you've seem that just jump to the end.
>>>>>
>>>>> YouTube video here: https://youtu.be/3Oft826Q8tA <https://youtu.be/3Oft826Q8tA>
>>>>>
>>>>>
>>>>>
>>>>> 73, de steve ve6wz
>>>>> _______________________________________________
>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> TowerTalk mailing list
>>>>> TowerTalk at contesting.com
>>>>> http://lists.contesting.com/mailman/listinfo/towertalk
>>>>
>>>> _______________________________________________
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> TowerTalk mailing list
>>>> TowerTalk at contesting.com
>>>> http://lists.contesting.com/mailman/listinfo/towertalk
>>> _______________________________________________
>>> _______________________________________________
>>> TowerTalk mailing list
>>> TowerTalk at contesting.com
>>> http://lists.contesting.com/mailman/listinfo/towertalk
>>
>> _______________________________________________
>>
>>
>>
>> _______________________________________________
>> TowerTalk mailing list
>> TowerTalk at contesting.com
>> http://lists.contesting.com/mailman/listinfo/towertalk
> 