Topics

Harmonics

 

While we are on the topic of harmonics:

If a rig does have spurs or harmonics in the output the antenna MAY not radiate much of it.

A 40M dipole will radiate 15M spurs or harmonics. The same will be greatly attenuated if harmonics/spurs
are present on say 20M or on a frequency that the antenna is not resonant.

I have not done experiments on this thought but some of you may have given it more thought or
even done some tests.

Cheers

--
Raj, vu2zap
Bengaluru, South India.

RCC WB5YYM
 

I too was wondering the same thing. I would think if you are using a resonant mono-band antenna, or using a tuner to tune the antenna that the harmonics measure at the output would have an even less chance of being radiated. With the ubitx being low power anyway, I suspect the radiated harmonic would be extremely minute, maybe in the low 1000ths of a watt. There has got to be some math out there some where to calculate this. 

Jack Purdum
 

Curtis:

I think there are two camps here: 1) those who feel that they cannot, in good conscience, operate any equipment that they know does not meet their ruling governance body (e.g., FCC), and 2) those who feel strongly that things are okay as long as I don't generate externalities that degrade or impair someone else's communications in any part of the spectrum. Al and I did a test on a stock µBITX and learned two things: 1) our $3600 commercial transceiver used for comparison was never "better" by more than 4dB on any band, and 2) the harmonics from the µBITX were so far below the noise level on all bands that it was undetectable.

Our tests, admittedly not as stringent as others, places me firmly in Camp 2 above. To me, its the Mathematician/Economist Problem (I'm an economist by training). Take a nude couple, place them in a room, and every 5 seconds, reduced the distance between them by 50%. To the mathematician, nothing ever happens. To the economist, game over in less than a few minutes. Camp 1 is the mathematician, Camp2 is the economist.

How could I choose anything other than Camp 2?

Jack, W8TEE

On Friday, August 10, 2018, 8:16:58 AM EDT, RCC WB5YYM <curtis03@...> wrote:


I too was wondering the same thing. I would think if you are using a resonant mono-band antenna, or using a tuner to tune the antenna that the harmonics measure at the output would have an even less chance of being radiated. With the ubitx being low power anyway, I suspect the radiated harmonic would be extremely minute, maybe in the low 1000ths of a watt. There has got to be some math out there some where to calculate this. 

ajparent1/KB1GMX
 

Jack,
No.  Camp one is it is and it has known problems what it is make your own choice.   
There are issues most can be solved. 

The other is a decision to deny, ignore, avoid, and continue.  The English word is denial.

To the Engineer and especially computers engineering its game over as 1 (true) is not 0 (false).

Its a hard reality and if one causes interference they are responsible for ceasing and fixing. 
On that basis I can only give facts after that it is a risk based choice. 

Allison

ajparent1/KB1GMX
 

Raj,

You have to go beyond the surface analysis as for odd harmonics and
dipole resonance at odd harmonics it fails. 

Allison

Jack Purdum
 

Allison:

I realize there are issues and, like everyone else who can fog a mirror, I want it solved. However, until a viable solution is found, do I stay off the air? What's the purpose of the spec in the first place? It's to make sure that a transmitter does not generate signals other than the fundamental frequency above certain specifications (i.e., strength). But why? It's done to ensure that my signal does not degrade someone else's signal on some harmonic frequency. If you, or anyone else, cannot detect my harmonic signal outside my back yard, I may fail the precise specs of the law, but am still within the intent of the law.

In your own words: "...if one causes interference..." implies that someone is impacted by my transmission. If my harmonics are in the noise level, am I causing interference? Legally, with measures at the source, yes. To the rest of the world, no. For my stock µBITX, I'm legal on 20M and above. On 40M, I'm 1dB out of spec. On 80M, I'm 3dB out of spec. I don't use 80M, other than contests, so that's really not a problem for me. 40M, however, is my favorite band. Not good.

I'm familiar with the word "denial" and I don't think you should imply that's what I'm doing. I'm not. I realize there's a problem, but unfortunately, I don't have the skill set to solve the problem. Your solution is for me to stay off 80M and 40M until you RF geniuses figure out a cost-effective solution. So why are you wasting your time bitching a me? Start pushing a pencil and show me the solution.

In the meantime, I promise not to use 80M...I'm pretty busy right now anyway, so I don't have much time for operating. However, if you hear my 40M harmonic, let me know and I'll cease operation immediately.

Jack, W8TEE

On Friday, August 10, 2018, 9:20:25 AM EDT, ajparent1/KB1GMX <kb1gmx@...> wrote:


Jack,
No.  Camp one is it is and it has known problems what it is make your own choice.   
There are issues most can be solved. 

The other is a decision to deny, ignore, avoid, and continue.  The English word is denial.

To the Engineer and especially computers engineering its game over as 1 (true) is not 0 (false).

Its a hard reality and if one causes interference they are responsible for ceasing and fixing. 
On that basis I can only give facts after that it is a risk based choice. 

Allison

jim
 



On Friday, August 10, 2018, 5:38:49 AM PDT, Jack Purdum via Groups.Io <jjpurdum@...> wrote:


Curtis:

 Take a nude couple, place them in a room, and every 5 seconds, reduced the distance between them by 50%. To the mathematician, nothing ever happens. To the economist, game over in less than a few minutes. Camp 1 is the mathematician, Camp2 is the economist.

How could I choose anything other than Camp 2?

Jack, W8TEE

Hmmm....Sounds like Zeno to me

Jim AB7VF

Jerry Gaffke
 

The solution has been stated here.  
A bank of external filters in-line with your antenna coax, these are available.
Alternately, hack the uBitx board to fix how out-of-band signals can jump across the LPF's
due to board layout issues, a half dozen different approaches have been proposed.
The only thing to be sorted out is how best to keep the cost and bother down.

Jerry


On Fri, Aug 10, 2018 at 06:53 AM, Jack Purdum wrote:
show me the solution.

Bill Cromwell <wrcromwell@...>
 

Hi,

As for me I am going the external lowpass filter route until this problem is tamed internally. But one question I must ask about operating the radio as-is. Of the thousands of units in circulation - and presumably on the air - how many hams have been cited by their regulatory agencies? 1000? 100? 10? 1? 0? Enquiring minds want to know.

Once again..I will be using external low pass filters at the BNC connector on the back of my uBit. The micro controller software in the Raduino cannot change those for me. Tuners and antennas beyond the filters. Tuners only to show the filters their design 50 ohm load. I don't want to be the first one cited for this. Nor the last.

When one of my daughters married Dr Murphy I though I would be given some slack on Murphy's Law because of being family. Instead, I get more than my share <wink>.

73,

Bill KU8H

On 08/10/2018 09:52 AM, Jack Purdum via Groups.Io wrote:
Allison:

I realize there are issues and, like everyone else who can fog a mirror,
I want it solved. However, until a viable solution is found, do I stay
off the air? What's the purpose of the spec in the first place? It's to
make sure that a transmitter does not generate signals other than the
fundamental frequency above certain specifications (i.e., strength). But
why? It's done to ensure that my signal does not degrade someone else's
signal on some harmonic frequency. If you, or anyone else, cannot detect
my harmonic signal outside my back yard, I may fail the precise specs of
the law, but am still within the intent of the law.

In your own words: "...if one causes interference..." implies that
someone is impacted by my transmission. If my harmonics are in the noise
level, am I causing interference? Legally, with measures at the source,
yes. To the rest of the world, no. For my stock µBITX, I'm legal on 20M
and above. On 40M, I'm 1dB out of spec. On 80M, I'm 3dB out of spec. I
don't use 80M, other than contests, so that's really not a problem for
me. 40M, however, is my favorite band. Not good.

I'm familiar with the word "denial" and I don't think you should imply
that's what I'm doing. I'm not. I realize there's a problem, but
unfortunately, I don't have the skill set to solve the problem. Your
solution is for me to stay off 80M and 40M until you RF geniuses figure
out a cost-effective solution. So why are you wasting your time bitching
a me? Start pushing a pencil and show me the solution.

In the meantime, I promise not to use 80M...I'm pretty busy right now
anyway, so I don't have much time for operating. However, if you hear my
40M harmonic, let me know and I'll cease operation immediately.

Jack, W8TEE

On Friday, August 10, 2018, 9:20:25 AM EDT, ajparent1/KB1GMX
<kb1gmx@...> wrote:


Jack,
No. Camp one is it is and it has known problems what it is make your
own choice.
There are issues most can be solved.

The other is a decision to deny, ignore, avoid, and continue. The
English word is denial.

To the Engineer and especially computers engineering its game over as 1
(true) is not 0 (false).

Its a hard reality and if one causes interference they are responsible
for ceasing and fixing.
On that basis I can only give facts after that it is a risk based choice.

Allison
--
bark less - wag more

Jack Purdum
 

Of the thousands of units in circulation - and
presumably on the air - how many hams have been cited by their
regulatory agencies?


Dollar to a doughnut it's 0. In the US, my guess is that the FCC has bigger fish to fry.

Jack, W8TEE


On Friday, August 10, 2018, 11:25:11 AM EDT, Bill Cromwell <wrcromwell@...> wrote:


Hi,

As for me I am going the external lowpass filter route until this
problem is tamed internally. But one question I must ask about operating
the radio as-is. Of the thousands of units in circulation - and
presumably on the air - how many hams have been cited by their
regulatory agencies? 1000? 100? 10? 1? 0? Enquiring minds want to know.

Once again..I will be using external low pass filters at the BNC
connector on the back of my uBit. The micro controller software in the
Raduino cannot change those for me. Tuners and antennas beyond the
filters. Tuners only to show the filters their design 50 ohm load. I
don't want to be the first one cited for this. Nor the last.

When one of my daughters married Dr Murphy I though I would be given
some slack on Murphy's Law because of being family. Instead, I get more
than my share <wink>.

73,

Bill  KU8H

On 08/10/2018 09:52 AM, Jack Purdum via Groups.Io wrote:
> Allison:
>
> I realize there are issues and, like everyone else who can fog a mirror,
> I want it solved. However, until a viable solution is found, do I stay
> off the air? What's the purpose of the spec in the first place? It's to
> make sure that a transmitter does not generate signals other than the
> fundamental frequency above certain specifications (i.e., strength). But
> why? It's done to ensure that my signal does not degrade someone else's
> signal on some harmonic frequency. If you, or anyone else, cannot detect
> my harmonic signal outside my back yard, I may fail the precise specs of
> the law, but am still within the intent of the law.
>
> In your own words: "...if one causes interference..." implies that
> someone is impacted by my transmission. If my harmonics are in the noise
> level, am I causing interference? Legally, with measures at the source,
> yes. To the rest of the world, no. For my stock µBITX, I'm legal on 20M
> and above. On 40M, I'm 1dB out of spec. On 80M, I'm 3dB out of spec. I
> don't use 80M, other than contests, so that's really not a problem for
> me. 40M, however, is my favorite band. Not good.
>
> I'm familiar with the word "denial" and I don't think you should imply
> that's what I'm doing. I'm not. I realize there's a problem, but
> unfortunately, I don't have the skill set to solve the problem. Your
> solution is for me to stay off 80M and 40M until you RF geniuses figure
> out a cost-effective solution. So why are you wasting your time bitching
> a me? Start pushing a pencil and show me the solution.
>
> In the meantime, I promise not to use 80M...I'm pretty busy right now
> anyway, so I don't have much time for operating. However, if you hear my
> 40M harmonic, let me know and I'll cease operation immediately.
>
> Jack, W8TEE
>
> On Friday, August 10, 2018, 9:20:25 AM EDT, ajparent1/KB1GMX
> <kb1gmx@...> wrote:
>
>
> Jack,
> No.  Camp one is it is and it has known problems what it is make your
> own choice.
> There are issues most can be solved.
>
> The other is a decision to deny, ignore, avoid, and continue.  The
> English word is denial.
>
> To the Engineer and especially computers engineering its game over as 1
> (true) is not 0 (false).
>
> Its a hard reality and if one causes interference they are responsible
> for ceasing and fixing.
> On that basis I can only give facts after that it is a risk based choice.
>
> Allison
>

--
bark less - wag more



Kees T
 

Jerry,

Do you have a summary of the various options available ? or being worked, with pros and cons ? or do we just wait for a "Eureka" thread ? 

It might be that everyone just takes his/her own path. 

73 Kees K5BCQ

Bill Cromwell <wrcromwell@...>
 

Hi Kees,

I expect there will be several different, useful solutions. I will wait a little to see how well they work and what will be required. I might try to take the best (for me) features of those designs and combine them into my own radio. Maybe one person's implementation of band pass filters ahead of the finals and another person's implementation of the output low pass filters. Open source is great ain't it?

73,

Bill KU8H

On 08/10/2018 11:43 AM, Kees T wrote:
Jerry,

Do you have a summary of the various options available ? or being
worked, with pros and cons ? or do we just wait for a "Eureka" thread ?

It might be that everyone just takes his/her own path.

73 Kees K5BCQ
--
bark less - wag more

Jerry Gaffke
 

Kees,

A few options that come to mind:
  1) Add external box with switchable filters
  2) Remove relays, create daughterboard to replace relays and LPF's plus additional relays at the far side
  3) Cut up current traces, add straps at board edge, use current relays only at input to LPF
  4) Same as above, but add a daughterboard with just the 3 extra relays at the board edge
  5) Clean up what goes into the final somehow so the LPF's don't see so much crud
  
My current favorite:
  6) Cut up the traces, dead bug three surface mount relays to the bottom of the board between L13,16,19,22,
      power the coils from the same source as the corresponding relays at KT1, KT2, KT3

The surface mount relays are a bit more expensive than the high-volume through hole relays.
Mouser 80-EE2-12NUX-L is $1.95 quantity one, down towards $1.00 in volume.
At 10.35mm high, they are just a hair shorter than the board standoffs provided with the uBitx,
at 7.5mm wide they fit nicely in the space available.  (There are smaller relays at extra cost.)
A grand total of $6 for an individual to clean up all but the (45mhz-FOp) spur on 15m-10m.
The coils can be powered from the same drivers as the existing relays,  
user will see no difference once the rig is put back in the enclosure. 


Jerry, KE7ER



On Fri, Aug 10, 2018 at 08:43 AM, Kees T wrote:
Do you have a summary of the various options available ? or being worked, with pros and cons ? or do we just wait for a "Eureka" thread ? 

It might be that everyone just takes his/her own path. 

73 Kees K5BCQ

Gordon Gibby
 

​Jerry, I'm with you, my plan is #4 with through-hole relays and surface mount capacitors / inductors as needed.   I don't need the higher bands that much and this is such a cool transceiver touse for winlink / ale / etc that these pesky issues just have to be fixed.   The first time I put it cycling ALE those folks were AMAZEDthat I could scan all their frequencies well.  The bandpass of my crystal filter was a bit too narrow, but now yall have taught me how to fix THAT as well and I'll do that to some rig.   


Cheers,

gordon



From: BITX20@groups.io <BITX20@groups.io> on behalf of Jerry Gaffke via Groups.Io <jgaffke@...>
Sent: Friday, August 10, 2018 1:10 PM
To: BITX20@groups.io
Subject: Re: [BITX20] Harmonics
 
Kees,

A few options that come to mind:
  1) Add external box with switchable filters
  2) Remove relays, create daughterboard to replace relays and LPF's plus additional relays at the far side
  3) Cut up current traces, add straps at board edge, use current relays only at input to LPF
  4) Same as above, but add a daughterboard with just the 3 extra relays at the board edge
  5) Clean up what goes into the final somehow so the LPF's don't see so much crud
  
My current favorite:
  6) Cut up the traces, dead bug three surface mount relays to the bottom of the board between L13,16,19,22,
      power the coils from the same source as the corresponding relays at KT1, KT2, KT3

The surface mount relays are a bit more expensive than the high-volume through hole relays.
Mouser 80-EE2-12NUX-L is $1.95 quantity one, down towards $1.00 in volume.
At 10.35mm high, they are just a hair shorter than the board standoffs provided with the uBitx,
at 7.5mm wide they fit nicely in the space available.  (There are smaller relays at extra cost.)
A grand total of $6 for an individual to clean up all but the (45mhz-FOp) spur on 15m-10m.
The coils can be powered from the same drivers as the existing relays,  
user will see no difference once the rig is put back in the enclosure. 


Jerry, KE7ER



On Fri, Aug 10, 2018 at 08:43 AM, Kees T wrote:
Do you have a summary of the various options available ? or being worked, with pros and cons ? or do we just wait for a "Eureka" thread ? 

It might be that everyone just takes his/her own path. 

73 Kees K5BCQ

KE2GKB
 

Could I implore for a write up of that with some visuals? #6 seems like my style fix. can deal with the 45Mhz issue somehow else.
73,
KE2GKB

On Fri, Aug 10, 2018 at 1:10 PM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:
Kees,

A few options that come to mind:
  1) Add external box with switchable filters
  2) Remove relays, create daughterboard to replace relays and LPF's plus additional relays at the far side
  3) Cut up current traces, add straps at board edge, use current relays only at input to LPF
  4) Same as above, but add a daughterboard with just the 3 extra relays at the board edge
  5) Clean up what goes into the final somehow so the LPF's don't see so much crud
  
My current favorite:
  6) Cut up the traces, dead bug three surface mount relays to the bottom of the board between L13,16,19,22,
      power the coils from the same source as the corresponding relays at KT1, KT2, KT3

The surface mount relays are a bit more expensive than the high-volume through hole relays.
Mouser 80-EE2-12NUX-L is $1.95 quantity one, down towards $1.00 in volume.
At 10.35mm high, they are just a hair shorter than the board standoffs provided with the uBitx,
at 7.5mm wide they fit nicely in the space available.  (There are smaller relays at extra cost.)
A grand total of $6 for an individual to clean up all but the (45mhz-FOp) spur on 15m-10m.
The coils can be powered from the same drivers as the existing relays,  
user will see no difference once the rig is put back in the enclosure. 


Jerry, KE7ER



On Fri, Aug 10, 2018 at 08:43 AM, Kees T wrote:
Do you have a summary of the various options available ? or being worked, with pros and cons ? or do we just wait for a "Eureka" thread ? 

It might be that everyone just takes his/her own path. 

73 Kees K5BCQ


RICHARD
 

To me option one  looks the best for the average ham today.

K6KWQ

 

Sent from Mail for Windows 10

 


From: BITX20@groups.io <BITX20@groups.io> on behalf of Jerry Gaffke via Groups.Io <jgaffke@...>
Sent: Friday, August 10, 2018 10:10:21 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Harmonics
 
Kees,

A few options that come to mind:
  1) Add external box with switchable filters
  2) Remove relays, create daughterboard to replace relays and LPF's plus additional relays at the far side
  3) Cut up current traces, add straps at board edge, use current relays only at input to LPF
  4) Same as above, but add a daughterboard with just the 3 extra relays at the board edge
  5) Clean up what goes into the final somehow so the LPF's don't see so much crud
  
My current favorite:
  6) Cut up the traces, dead bug three surface mount relays to the bottom of the board between L13,16,19,22,
      power the coils from the same source as the corresponding relays at KT1, KT2, KT3

The surface mount relays are a bit more expensive than the high-volume through hole relays.
Mouser 80-EE2-12NUX-L is $1.95 quantity one, down towards $1.00 in volume.
At 10.35mm high, they are just a hair shorter than the board standoffs provided with the uBitx,
at 7.5mm wide they fit nicely in the space available.  (There are smaller relays at extra cost.)
A grand total of $6 for an individual to clean up all but the (45mhz-FOp) spur on 15m-10m.
The coils can be powered from the same drivers as the existing relays,  
user will see no difference once the rig is put back in the enclosure. 


Jerry, KE7ER



On Fri, Aug 10, 2018 at 08:43 AM, Kees T wrote:
Do you have a summary of the various options available ? or being worked, with pros and cons ? or do we just wait for a "Eureka" thread ? 

It might be that everyone just takes his/her own path. 

73 Kees K5BCQ

Kees T
 

If we settle on 4 LPFs, and they must be relay switched on some kind of daughter board and we use the existing components off the uBITX board, what is the easiest way to use the TXA,TXB, and TXC outputs from the 2N3904's to decode the 4 states to drive 4 sets of 2 relays per LPF (000,001,011,111).

"Easiest" means lowest parts cost and hassle for the daughter board.

73 Kees K5BCQ 

Richard Spohn
 

I volunteer to be the guy in Camp 2. - Rich

On 8/10/18, jim via Groups.Io <ab7vf=yahoo.com@groups.io> wrote:


On Friday, August 10, 2018, 5:38:49 AM PDT, Jack Purdum via Groups.Io
<jjpurdum=yahoo.com@groups.io> wrote:

Curtis:
 Take a nude couple, place them in a room, and every 5 seconds, reduced the
distance between them by 50%. To the mathematician, nothing ever happens. To
the economist, game over in less than a few minutes. Camp 1 is the
mathematician, Camp2 is the economist.

How could I choose anything other than Camp 2?
Jack, W8TEE

Hmmm....Sounds like Zeno to me

Jim AB7VF




Glenn
 

Kees, i think a CD4028 would do it although the outputs will be active logic Lo so a transistor driver would be needed also..
glenn

Dexter N Muir
 

Cure: Witchcraft! A Hex-Inverter :)