Topics

Understanding Spurious Emissions

Howard Fidel
 

There have been a few threads discussing how the uBitx has spurs that may exceed the spur amplitude allowed by law. Ham radio is regulated by Part 97 of the FCC regulations. The allowed spur amplitude for frequencies below 30 MHz is in section 307. I have reproduced paragraph D below:

(d) For transmitters installed after January 1, 2003, the mean power of any spurious emission from a station transmitter or external RF power amplifier transmitting on a frequency below 30 MHz must be at least 43 dB below the mean power of the fundamental emission. For transmitters installed on or before January 1, 2003, the mean power of any spurious emission from a station transmitter or external RF power amplifier transmitting on a frequency below 30 MHz must not exceed 50 mW and must be at least 40 dB below the mean power of the fundamental emission. For a transmitter of mean power less than 5 W installed on or before January 1, 2003, the attenuation must be at least 30 dB. A transmitter built before April 15, 1977, or first marketed before January 1, 1978, is exempt from this requirement. 

The first interesting thing about these regulations is that they are not consistent across time. If you have a transmitter that was installed before April 15, 1977 it is not even regulated for spurious emissions. Before 2003, the spurs were limited to 50 mW max and must be at least 40 dB below the carrier power. Today, the constraint is -43 dB below the carrier power. So if you have a 1 KW transmitter, you are allowed to have spurs that do not exceed 50 mW, hence the earlier 50 mW limit. To keep things in perspective, if your uBitx puts out 5 watts of power on 15 meters, 50 mW of energy would be only 20 dB down from your carrier. 
So if you use your uBitx barefoot (i.e without a linear amp) although you may exceed the allowed spur amplitude on the 15 and 10 meter bands, your radiated power level will be so low that it is virtually impossible for it to interfere with other services, and that your radiated emissions maybe in line with what other ham operators are radiating legally. Also, I might add that these spurs are not consistent in amplitude  from unit to unit. I measured mine as being in compliance, but right at the -43 dB limit. Furthermore, the testing is done into a dummy load. When connected to an antenna, your SWR at the spur frequency is probably high, so you are radiating even less energy. 
My personal take away from this is that although the uBitx may at times not be in technical compliance with the regulations, operating at frequencies above 21 MHz, it is in compliance with the intent of the law, which is to prevent interference with other services. You can help your uBitx stay in-compliance by not over driving your audio, and by only operating it barefoot above 21 MHz. 
So, the bottom line is you should not use the uBitx above 21 MHz if you are concerned with the letter of the law, but you may use it if you are concerned with complying with the intent of the law.

Howard

Arv Evans
 

Most who have performed detailed tests seem to indicate that unwanted spurs are a problem only if AF or RF drive levels are set too high.  On CW the uBITX does not use modulation and thus has low spurious output.

Arv
_-_



Sent from my Verizon, Samsung Galaxy smartphone

-------- Original message --------
From: Howard Fidel <sonic1@...>
Date: 6/20/18 8:36 AM (GMT-07:00)
To: BITX20@groups.io
Subject: [BITX20] Understanding Spurious Emissions

There have been a few threads discussing how the uBitx has spurs that may exceed the spur amplitude allowed by law. Ham radio is regulated by Part 97 of the FCC regulations. The allowed spur amplitude for frequencies below 30 MHz is in section 307. I have reproduced paragraph D below:

(d) For transmitters installed after January 1, 2003, the mean power of any spurious emission from a station transmitter or external RF power amplifier transmitting on a frequency below 30 MHz must be at least 43 dB below the mean power of the fundamental emission. For transmitters installed on or before January 1, 2003, the mean power of any spurious emission from a station transmitter or external RF power amplifier transmitting on a frequency below 30 MHz must not exceed 50 mW and must be at least 40 dB below the mean power of the fundamental emission. For a transmitter of mean power less than 5 W installed on or before January 1, 2003, the attenuation must be at least 30 dB. A transmitter built before April 15, 1977, or first marketed before January 1, 1978, is exempt from this requirement. 

The first interesting thing about these regulations is that they are not consistent across time. If you have a transmitter that was installed before April 15, 1977 it is not even regulated for spurious emissions. Before 2003, the spurs were limited to 50 mW max and must be at least 40 dB below the carrier power. Today, the constraint is -43 dB below the carrier power. So if you have a 1 KW transmitter, you are allowed to have spurs that do not exceed 50 mW, hence the earlier 50 mW limit. To keep things in perspective, if your uBitx puts out 5 watts of power on 15 meters, 50 mW of energy would be only 20 dB down from your carrier. 
So if you use your uBitx barefoot (i.e without a linear amp) although you may exceed the allowed spur amplitude on the 15 and 10 meter bands, your radiated power level will be so low that it is virtually impossible for it to interfere with other services, and that your radiated emissions maybe in line with what other ham operators are radiating legally. Also, I might add that these spurs are not consistent in amplitude  from unit to unit. I measured mine as being in compliance, but right at the -43 dB limit. Furthermore, the testing is done into a dummy load. When connected to an antenna, your SWR at the spur frequency is probably high, so you are radiating even less energy. 
My personal take away from this is that although the uBitx may at times not be in technical compliance with the regulations, operating at frequencies above 21 MHz, it is in compliance with the intent of the law, which is to prevent interference with other services. You can help your uBitx stay in-compliance by not over driving your audio, and by only operating it barefoot above 21 MHz. 
So, the bottom line is you should not use the uBitx above 21 MHz if you are concerned with the letter of the law, but you may use it if you are concerned with complying with the intent of the law.

Howard

Howard Fidel
 

Very true!

On 6/20/2018 11:30 AM, Arv Evans wrote:
Most who have performed detailed tests seem to indicate that unwanted spurs are a problem only if AF or RF drive levels are set too high.  On CW the uBITX does not use modulation and thus has low spurious output.

Arv
_-_



Sent from my Verizon, Samsung Galaxy smartphone

-------- Original message --------
From: Howard Fidel <sonic1@...>
Date: 6/20/18 8:36 AM (GMT-07:00)
Subject: [BITX20] Understanding Spurious Emissions

There have been a few threads discussing how the uBitx has spurs that may exceed the spur amplitude allowed by law. Ham radio is regulated by Part 97 of the FCC regulations. The allowed spur amplitude for frequencies below 30 MHz is in section 307. I have reproduced paragraph D below:

(d) For transmitters installed after January 1, 2003, the mean power of any spurious emission from a station transmitter or external RF power amplifier transmitting on a frequency below 30 MHz must be at least 43 dB below the mean power of the fundamental emission. For transmitters installed on or before January 1, 2003, the mean power of any spurious emission from a station transmitter or external RF power amplifier transmitting on a frequency below 30 MHz must not exceed 50 mW and must be at least 40 dB below the mean power of the fundamental emission. For a transmitter of mean power less than 5 W installed on or before January 1, 2003, the attenuation must be at least 30 dB. A transmitter built before April 15, 1977, or first marketed before January 1, 1978, is exempt from this requirement. 

The first interesting thing about these regulations is that they are not consistent across time. If you have a transmitter that was installed before April 15, 1977 it is not even regulated for spurious emissions. Before 2003, the spurs were limited to 50 mW max and must be at least 40 dB below the carrier power. Today, the constraint is -43 dB below the carrier power. So if you have a 1 KW transmitter, you are allowed to have spurs that do not exceed 50 mW, hence the earlier 50 mW limit. To keep things in perspective, if your uBitx puts out 5 watts of power on 15 meters, 50 mW of energy would be only 20 dB down from your carrier. 
So if you use your uBitx barefoot (i.e without a linear amp) although you may exceed the allowed spur amplitude on the 15 and 10 meter bands, your radiated power level will be so low that it is virtually impossible for it to interfere with other services, and that your radiated emissions maybe in line with what other ham operators are radiating legally. Also, I might add that these spurs are not consistent in amplitude  from unit to unit. I measured mine as being in compliance, but right at the -43 dB limit. Furthermore, the testing is done into a dummy load. When connected to an antenna, your SWR at the spur frequency is probably high, so you are radiating even less energy. 
My personal take away from this is that although the uBitx may at times not be in technical compliance with the regulations, operating at frequencies above 21 MHz, it is in compliance with the intent of the law, which is to prevent interference with other services. You can help your uBitx stay in-compliance by not over driving your audio, and by only operating it barefoot above 21 MHz. 
So, the bottom line is you should not use the uBitx above 21 MHz if you are concerned with the letter of the law, but you may use it if you are concerned with complying with the intent of the law.

Howard


N1EDC
 

Thanks for the concise explanation!

As for overdriving your audio, is this something that can be solved by dialing down RV1 to reasonable levels, or are the spurs created before that stage?

Tim Gorman
 

Spurs are generated in mixer stages. Overdriving your audio can
increase the amplitude of those spurs. It can also generate 3rd order
IMD products.

Adjusting RV1 down won't help with the spurs but it might help with 3rd
order IMD products.

tim ab0wr

On Thu, 21 Jun 2018 06:59:14 -0700
"KE0GYC" <@N1EDC> wrote:

Thanks for the concise explanation!

As for overdriving your audio, is this something that can be solved
by dialing down RV1 to reasonable levels, or are the spurs created
before that stage?


Howard Fidel
 

The overdrive occurs in the audio drive to the mixer.  There is no gain control for the microphone. I found on my board that the gain was too low. You can look at the mixer input with a scope and see when the audio starts clipping (I didn't validate this). Another test is if your audio reports are clean, you probably are okay, if you start to distort due to compression, you are over driving. I can't overdrive on my board because I put an AGC on the mic and its output swing is limited.

On 6/21/2018 9:59 AM, KE0GYC wrote:
Thanks for the concise explanation!

As for overdriving your audio, is this something that can be solved by dialing down RV1 to reasonable levels, or are the spurs created before that stage?


ajparent1/kb1gmx
 

Howard,

The mixer where the spur is generated is the mixer formed by T2/T3 that is about 
28db (after allowing for filter losses and all)  later in the stream and if that mixer
is overloaded you get the spur.  The driving modulator is then underdriven.

Work it backward.  If that mixer is overloaded at -7 oir -10DBM what got to be
happening well before that?   And that mixer can easily be driven to its 1DB
compression or worse without hurting the audio to the ear.

Allison

ajparent1/kb1gmx
 

Tim has it right.  By time you get to RV1, the signals are what they are.  There is no adjustment
prior to that other than audio level from the mic.

Allison

Howard Fidel
 

Got it. Thanks. I never saw on my uBitx the numbers you reported, but I don't think it is a serious problem for a barefoot uBitx, which is the main point I was trying to make. It seemed like the group was spending a lot of energy on something that is like driving 60 MPH in a 55 zone.

On 6/21/2018 7:22 PM, ajparent1/KB1GMX wrote:
Howard,

The mixer where the spur is generated is the mixer formed by T2/T3 that is about 
28db (after allowing for filter losses and all)  later in the stream and if that mixer
is overloaded you get the spur.  The driving modulator is then underdriven.

Work it backward.  If that mixer is overloaded at -7 oir -10DBM what got to be
happening well before that?   And that mixer can easily be driven to its 1DB
compression or worse without hurting the audio to the ear.

Allison


ajparent1/kb1gmx
 

No, your analogy would be right if you were on the wrong side of the road. 
If you get near even fully driving the modulator the last mixer is easily 10+ db
past saturated as result of there being enough IF gain to assure that.

For those that never go above 20Mhz its adequately filtered and not an issue.

Allison

Jerry Gaffke
 

I wouldn't operate the transmitter on 15m,12m,10m without
either a fix for the spur or some way to monitor it.

Monitoring could be done with a second receiver plus a step attenuator.
Transmit into a dummy load while monitoring the spur at 44.995mhz-Fop through a step attenuator, 
where Fop is your operating frequency.
So if you are operating at  Fop = 29.0 mhz, the spur is at 44.995-29=15.995mhz, give or take a couple khz. 
Determine the signal strength of the spur, perhaps using the receiver S-meter.
Then tune in your operating frequency, and use the step attenuator to reduce the strength of your main signal
into the dummy load until you get the same S-meter reading as you did for the spur.
The additional attenuation required tells you how many dB below your main signal that spur is.
A good source for a step attenuator is the "Attenuator Kit #13" from Kees:  http://www.qsl.net/k5bcq/Kits/Kits.html

If you can't be bothered to do the procedure above, it may not be a good idea to be transmitting 
on the 15m, 12m and 10m bands using the uBitx.
This is especially true if you have jacked up your mike gain somehow to get more power out,
or are running your uBitx into a high powered external amplifier.

Some do report that a properly adjusted uBitx can have that spur down near the legal limit
of 43 dB below the main signal.  But that is pretty much the best possible case.
Your best possible case could be much worse, the gain of the 45mhz amp in particular will be
all over the map depending on the characteristics of the mmbt3904's that happen to be in your rig.
And it will definitely get worse if you raise the mike gain, or shout into the mike.
We seem to know the best case, I would want to know my worst case 
before attempting to operate on these upper bands. 


There are several solutions to this.

We could call it a rig suitable for operation from 160m to 17m, forget the upper bands,
replacing the top transmit LPF with one suitable for 160m.

We could replace the 30mhz LPF at L1,2,3,4 with different filters for each band.
This is what pretty much all other HF transceivers do.

Or we could raise the first IF frequency from 45mhz to something well above 60mhz,
perhaps the 86.85mhz of post   https://groups.io/g/BITX20/message/52460
Since 86.85mhz - 30mhz puts the lowest possible frequency for the spur at 56.85mhz,
the spur won't get through the stock 30mhz LPF at L1,2,3,4


I'd prefer to see a fix that replaces the 45mhz IF, everything there including those two diode ring mixers.
This could be a small daughterboard that mounts into the holes remaining when the mixer transformers are removed.
All 6 mmbt3904's would be replaced with BFR106's, the 86.85 mhz PX1002 is used as the filter,
the mixers redesigned for 86mhz.  Gains would be adjusted in both IF amps to minimize the spurs,
and also minimize the carrier we have leaking through.
Parts count and topology remain the same as the stock uBitx, no added complexity,
though we could add some sort of RF AGC attenuator for some extra pennies.
The better transistors will allow an even drive to the PA, giving us a better shot at getting
the same power out to the antenna on 10m as on 80m.
Total cost of parts can be under $10, so it might be a kit that sells for $20 at sufficient volume.
If hfsignals decides to move in this direction, the only significant hit on cost or procurement
would be the $6 PX1002 (volume pricing) vs the custom built 45mhz crystal filter (whatever that costs).
This has not yet been tried.

Jerry, KE7ER


On Fri, Jun 22, 2018 at 05:18 am, ajparent1/KB1GMX wrote:
No, your analogy would be right if you were on the wrong side of the road. 
If you get near even fully driving the modulator the last mixer is easily 10+ db
past saturated as result of there being enough IF gain to assure that.

For those that never go above 20Mhz its adequately filtered and not an issue.

Allison
#############
Howard Fidel  June 21   #52728

Got it. Thanks. I never saw on my uBitx the numbers you reported, but I don't think it is a serious problem for a barefoot uBitx, which is the main point I was trying to make. It seemed like the group was spending a lot of energy on something that is like driving 60 MPH in a 55 zone. 

Jerry Gaffke
 

Correction:  use the step attenuator to reduce the strength of your main signal FROM the dummy load 
You likely want a pretty serious resistive divider between dummy load and step attenuator,
The step attenuator might be able to disspate a Watt, maybe less. 


On Fri, Jun 22, 2018 at 09:58 am, Jerry Gaffke wrote:
Then tune in your operating frequency, and use the step attenuator to reduce the strength of your main signal
into the dummy load until you get the same S-meter reading as you did for the spur.

Arv Evans
 

Caution here seems to be that if you have modified your uBITX from original,
if you are using an amplified microphone, or if you have pushed the gain
excessively you may have increased the spurious output.  Unmodified uBITX
operating within stated specification should be okay.  Tests were performed
during initial design and prototype testing to make sure that this is the case.

However, if you have applied any of the many modifications that are described
on this venue, or any other modifications, then it would be your responsibility
to insure that these changes to the original design do not cause it to operate '
outside of original specifications.

Arv
_._


On Fri, Jun 22, 2018 at 10:59 AM Jerry Gaffke via Groups.Io <jgaffke=yahoo.com@groups.io> wrote:
I wouldn't operate the transmitter on 15m,12m,10m without
either a fix for the spur or some way to monitor it.

Monitoring could be done with a second receiver plus a step attenuator.
Transmit into a dummy load while monitoring the spur at 44.995mhz-Fop through a step attenuator, 
where Fop is your operating frequency.
So if you are operating at  Fop = 29.0 mhz, the spur is at 44.995-29=15.995mhz, give or take a couple khz. 
Determine the signal strength of the spur, perhaps using the receiver S-meter.
Then tune in your operating frequency, and use the step attenuator to reduce the strength of your main signal
into the dummy load until you get the same S-meter reading as you did for the spur.
The additional attenuation required tells you how many dB below your main signal that spur is.
A good source for a step attenuator is the "Attenuator Kit #13" from Kees:  http://www.qsl.net/k5bcq/Kits/Kits.html

If you can't be bothered to do the procedure above, it may not be a good idea to be transmitting 
on the 15m, 12m and 10m bands using the uBitx.
This is especially true if you have jacked up your mike gain somehow to get more power out,
or are running your uBitx into a high powered external amplifier.

Some do report that a properly adjusted uBitx can have that spur down near the legal limit
of 43 dB below the main signal.  But that is pretty much the best possible case.
Your best possible case could be much worse, the gain of the 45mhz amp in particular will be
all over the map depending on the characteristics of the mmbt3904's that happen to be in your rig.
And it will definitely get worse if you raise the mike gain, or shout into the mike.
We seem to know the best case, I would want to know my worst case 
before attempting to operate on these upper bands. 


There are several solutions to this.

We could call it a rig suitable for operation from 160m to 17m, forget the upper bands,
replacing the top transmit LPF with one suitable for 160m.

We could replace the 30mhz LPF at L1,2,3,4 with different filters for each band.
This is what pretty much all other HF transceivers do.

Or we could raise the first IF frequency from 45mhz to something well above 60mhz,
perhaps the 86.85mhz of post   https://groups.io/g/BITX20/message/52460
Since 86.85mhz - 30mhz puts the lowest possible frequency for the spur at 56.85mhz,
the spur won't get through the stock 30mhz LPF at L1,2,3,4


I'd prefer to see a fix that replaces the 45mhz IF, everything there including those two diode ring mixers.
This could be a small daughterboard that mounts into the holes remaining when the mixer transformers are removed.
All 6 mmbt3904's would be replaced with BFR106's, the 86.85 mhz PX1002 is used as the filter,
the mixers redesigned for 86mhz.  Gains would be adjusted in both IF amps to minimize the spurs,
and also minimize the carrier we have leaking through.
Parts count and topology remain the same as the stock uBitx, no added complexity,
though we could add some sort of RF AGC attenuator for some extra pennies.
The better transistors will allow an even drive to the PA, giving us a better shot at getting
the same power out to the antenna on 10m as on 80m.
Total cost of parts can be under $10, so it might be a kit that sells for $20 at sufficient volume.
If hfsignals decides to move in this direction, the only significant hit on cost or procurement
would be the $6 PX1002 (volume pricing) vs the custom built 45mhz crystal filter (whatever that costs).
This has not yet been tried.

Jerry, KE7ER


On Fri, Jun 22, 2018 at 05:18 am, ajparent1/KB1GMX wrote:
No, your analogy would be right if you were on the wrong side of the road. 
If you get near even fully driving the modulator the last mixer is easily 10+ db
past saturated as result of there being enough IF gain to assure that.

For those that never go above 20Mhz its adequately filtered and not an issue.

Allison
#############
Howard Fidel  June 21   #52728

Got it. Thanks. I never saw on my uBitx the numbers you reported, but I don't think it is a serious problem for a barefoot uBitx, which is the main point I was trying to make. It seemed like the group was spending a lot of energy on something that is like driving 60 MPH in a 55 zone. 

Jerry Gaffke
 

I can sort of agree.
If transmitting a 1W signal out on 10m somehow and the spur is down 30 dB,
that spur is equivalent to what we get from a 100W transmitter with a spur that is down 50 dB.
Though the relevant officials may disagree.

It's a kit, it is your responsibility to ensure your transmitted signal is clean.
Regardless of mods or no mods.
This rig un-modded can transmit on 4.5mhz, don't do that either!
Keep in mind that there is considerable variation between rigs regarding available output power,
much of this is due to gain variations in the 45mhz IF which will effect the strength of the spur.
 
Increasing the mike gain on 15-10m will increase the rather meager output power
but will disproportionally increase the spur.  So don't.

Most important, do not transmit SSB from the uBitx through a high powered 
external linear amp on 15-10m unless you deal with the spur.
Put a KW amp on it and your spur would be about as strong as a stock uBitx.

There is no problem with CW transmission on any band with regard to the spur.
Though you might check out this thread regarding uBitx key clicks:  
    https://groups.io/g/BITX20/message/51673

I'm pushing a bit here because:
  Very few here have any inclination to check the spur.
  Many would be inclined to bump mike gain up somehow to the limit where IMD kicks in (the spur occurs way before that).
  Many would be inclined to use an external RF power amp, especially on the weak high bands.
  Future versions of the rig should clean this up or it will remain an issue. 

Jerry, KE7ER

ajparent1/kb1gmx
 

I have tested three with no mods that impact spurs and at 2W out on 10M the spur is
less than 38DB down on the best of them.  One has a standing 2.1mW output at
16.595mhz when no modulation at 28.4mhz due to carrier leakage.  Modulated its 
less than 29db down from 2W.  At 21.3mhz the spur is at 23.695mhz (use a really
good filter).  

To verify the power amp does not produce this I cut the trace and measure the
signal out of the 30mhz 9 element filter and its there and the uneven gain of the amp makes it worse.

A bench test using a MiniCircuits ZLW-2 connectorized mixer with 6DB pads on all ports
when 45mhz is applied to the IF port (at less than -1dbm) and 73mhz applied to the
LO port at 6dbm (all ins and outs corrected for the pads as they are there to insure the 
match to all ports are good) the RF port will have 28mhz(difference), 17mhz (major spur),
and 118mhz (sum).   Also there are the 2nd 3rd and nth order mix products and their
sums and differences..  This is normal behaviour.  Based on the datasheet the spur
will be 40db down if the levels are not excessive and no better with the test set up
I saw 41db but increase the 45mhz to -7dbm and the spur increases by more than
3DB.  Reduce the 45mhz 10db to -20dbm and the spur is there at -40dbm less than
the desired 28mhz.  Remove the 45mhz completely and the only the output is 73mhz
and its harmonics.  Its the nature of the beast as they say.

Unfiltered output will have spurs.  There is no filter when tuning above 20mhz 
that can take out any spur below the 30mhz cutoff.  The spurs are blow that.

I have two power attenuators for stuff like this.  My first is the Narda 30DB 50W,
I also have a Bird 25W 30DB.  I also test them before use.  Test setup is 30DB
narda 50W attenuator, step attenuator (20db in use) and spectrum analyser
(either my HP8568B or Rigol DSA815T).  They agree within less than a fraction
of a DB and both have been calibrated in the last year.  You can do this using 
the switched attenuator and a receiver with a meter it just takes longer as you have to
measure at two frequencies.  But with care you can be accurate to 1DB.

I have 3 samples and a test bed and exact measurements.  FYI, no power amp mods can
improve this to compliance though those that do not improve the 10M power out to equal
that of 20M will make it appear worse as the spur at 16-17mhz (for 28-29mhz tuning) will
be amplified more than the 28mhz signal!  If it were marginal or almost I'd have stayed quiet.

That is the whole story.  
There are three solutions:
1) Do not operate SSB  TX above 20mhz, CW TX is not an issue due to the way the radio operates.
2) Include some form of band pass filter for the band desired (15, 12, or  10). 
    Any suggestions will have to be completely tested by other than me. (tuners may may it worse).
3) Change the 45mhz IF to more than 60mhz (this means better transistors in the "45mhz"
IF section for RX and TX.

Sorry to be a bringer of bad news.  I've spent a lot of time making sure there was no error or
simple fix.   It is why mixers in commercial use always get filtered at the outputs.  I plan to do
something as I want 10M.  at this time I'm not sure which.

Allison

Gordon Gibby <ggibby@...>
 

Jerry, I assume just purchasing or building a bandpass filter for the band(s)  amongst 15-10 that you want to use would solve this problem,  right?   I forget exactly which band has a really close and spur, perhaps it wouldn’t work for that band.

I think we should give people more solutions here. Here is an ARRL publication on how to build shelters for yourself.   Note that he built his 15 m filter with air inductors, and it’s performance wasn’t nearly as good as the ones he built on ferrites.

And I think I bought a band pass filter for 80 m back when I was using a bit x40 switched to 80, I forget where I got that from....

Cheers,
Gordon







On Jun 22, 2018, at 18:56, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:

I can sort of agree.
If transmitting a 1W signal out on 10m somehow and the spur is down 30 dB,
that spur is equivalent to what we get from a 100W transmitter with a spur that is down 50 dB.
Though the relevant officials may disagree.

It's a kit, it is your responsibility to ensure your transmitted signal is clean.
Regardless of mods or no mods.
This rig un-modded can transmit on 4.5mhz, don't do that either!
Keep in mind that there is considerable variation between rigs regarding available output power,
much of this is due to gain variations in the 45mhz IF which will effect the strength of the spur.
 
Increasing the mike gain on 15-10m will increase the rather meager output power
but will disproportionally increase the spur.  So don't.

Most important, do not transmit SSB from the uBitx through a high powered 
external linear amp on 15-10m unless you deal with the spur.
Put a KW amp on it and your spur would be about as strong as a stock uBitx.

There is no problem with CW transmission on any band with regard to the spur.
Though you might check out this thread regarding uBitx key clicks:  
    https://groups.io/g/BITX20/message/51673

I'm pushing a bit here because:
  Very few here have any inclination to check the spur.
  Many would be inclined to bump mike gain up somehow to the limit where IMD kicks in (the spur occurs way before that).
  Many would be inclined to use an external RF power amp, especially on the weak high bands.
  Future versions of the rig should clean this up or it will remain an issue. 

Jerry, KE7ER

Gordon Gibby <ggibby@...>
 

On Jun 22, 2018, at 19:23, Gordon Gibby <ggibby@...> wrote:

Jerry, I assume just purchasing or building a bandpass filter for the band(s)  amongst 15-10 that you want to use would solve this problem,  right?   I forget exactly which band has a really close and spur, perhaps it wouldn’t work for that band.

I think we should give people more solutions here. Here is an ARRL publication on how to build shelters for yourself.   Note that he built his 15 m filter with air inductors, and it’s performance wasn’t nearly as good as the ones he built on ferrites.

And I think I bought a band pass filter for 80 m back when I was using a bit x40 switched to 80, I forget where I got that from....

Cheers,
Gordon







On Jun 22, 2018, at 18:56, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:

I can sort of agree.
If transmitting a 1W signal out on 10m somehow and the spur is down 30 dB,
that spur is equivalent to what we get from a 100W transmitter with a spur that is down 50 dB.
Though the relevant officials may disagree.

It's a kit, it is your responsibility to ensure your transmitted signal is clean.
Regardless of mods or no mods.
This rig un-modded can transmit on 4.5mhz, don't do that either!
Keep in mind that there is considerable variation between rigs regarding available output power,
much of this is due to gain variations in the 45mhz IF which will effect the strength of the spur.
 
Increasing the mike gain on 15-10m will increase the rather meager output power
but will disproportionally increase the spur.  So don't.

Most important, do not transmit SSB from the uBitx through a high powered 
external linear amp on 15-10m unless you deal with the spur.
Put a KW amp on it and your spur would be about as strong as a stock uBitx.

There is no problem with CW transmission on any band with regard to the spur.
Though you might check out this thread regarding uBitx key clicks:  
    https://groups.io/g/BITX20/message/51673

I'm pushing a bit here because:
  Very few here have any inclination to check the spur.
  Many would be inclined to bump mike gain up somehow to the limit where IMD kicks in (the spur occurs way before that).
  Many would be inclined to use an external RF power amp, especially on the weak high bands.
  Future versions of the rig should clean this up or it will remain an issue. 

Jerry, KE7ER

Jerry Gaffke
 

Gordon,

I agree, easiest solution for current owners is an external filter in-line at the antenna port,
no mods to the rig.

As Allison says, "At 21.3mhz the spur is at 23.695mhz (use a really good filter)."  
So 15m would be difficult to filter when using a 45mhz IF.
The formula is 44.995mhz - 21.3 = 23.695mhz.

Something like this at the antenna port might be sufficient on 15m, I don't really know.
Seems plenty sharp to me, maybe a bit too sharp to fit all of 15m.   
This is a bandpass, could be a very sharp lowpass.
Tuning will be critical, may need trimmer caps.
The T50 toroids are bigger than the transmit LPF's in the uBitx, so my guess is it will handle the power:
    http://www.kitsandparts.com/univbpfilter.php
 
10m is rather broad, probably easiest to use a high-pass filter of some sort.

I would sweep any such attempts at filters with the si5351, determine attenuation
at the operating frequency and spur with an ad8307.

Others with more experience building filters will undoubtedly have better advice
on filter design choices and testing methods.  The above is just a first guess
at what direction to go.

Building shelters is a good idea too.
Winter will be coming on soon enough.
.
Jerry


On Fri, Jun 22, 2018 at 04:22 pm, Gordon Gibby wrote:
Jerry, I assume just purchasing or building a bandpass filter for the band(s)  amongst 15-10 that you want to use would solve this problem,  right?   I forget exactly which band has a really close and spur, perhaps it wouldn’t work for that band.
 
I think we should give people more solutions here. Here is an ARRL publication on how to build shelters for yourself.   Note that he built his 15 m filter with air inductors, and it’s performance wasn’t nearly as good as the ones he built on ferrites.
 
And I think I bought a band pass filter for 80 m back when I was using a bit x40 switched to 80, I forget where I got that from....
 
Cheers,
Gordon

Gordon Gibby <ggibby@...>
 

Jerry , that is a great solution.
Reading carefully through it I see that for the higher bands they all include three 40pF trimmer capacitors to tunethe filter.

Although they don’t include instructions how to tune,I assume you just adjust for highest output at the center of the band area you wish.   


If the 3DB points are only 90 kHz at 40 m, there should be plenty Narrow for getting rid of spurs.


On Jun 22, 2018, at 20:17, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:

Gordon,

I agree, easiest solution for current owners is an external filter in-line at the antenna port,
no mods to the rig.

As Allison says, "At 21.3mhz the spur is at 23.695mhz (use a really good filter)."  
So 15m would be difficult to filter when using a 45mhz IF.
The formula is 44.995mhz - 21.3 = 23.695mhz.

Something like this at the antenna port might be sufficient on 15m, I don't really know.
Seems plenty sharp to me, maybe a bit too sharp to fit all of 15m.   
This is a bandpass, could be a very sharp lowpass.
Tuning will be critical, may need trimmer caps.
The T50 toroids are bigger than the transmit LPF's in the uBitx, so my guess is it will handle the power:
    http://www.kitsandparts.com/univbpfilter.php
 
10m is rather broad, probably easiest to use a high-pass filter of some sort.

I would sweep any such attempts at filters with the si5351, determine attenuation
at the operating frequency and spur with an ad8307.

Others with more experience building filters will undoubtedly have better advice
on filter design choices and testing methods.  The above is just a first guess
at what direction to go.

Building shelters is a good idea too.
Winter will be coming on soon enough.
.
Jerry


On Fri, Jun 22, 2018 at 04:22 pm, Gordon Gibby wrote:
Jerry, I assume just purchasing or building a bandpass filter for the band(s)  amongst 15-10 that you want to use would solve this problem,  right?   I forget exactly which band has a really close and spur, perhaps it wouldn’t work for that band.
 
I think we should give people more solutions here. Here is an ARRL publication on how to build shelters for yourself.   Note that he built his 15 m filter with air inductors, and it’s performance wasn’t nearly as good as the ones he built on ferrites.
 
And I think I bought a band pass filter for 80 m back when I was using a bit x40 switched to 80, I forget where I got that from....
 
Cheers,
Gordon

ajparent1/kb1gmx
 

If we use round numbers...

45mhz - dial frequency= spur

For 28 to 29mhz the spur is 17 to 16mhz
For 24.830 to 24.990mhz the spur is 20.17 to 20.01 mhz 
For 21.2 to 21.45mhz the spur is 23.8 to 23.55mhz 

For 10 and 12 if one can devise a suitable high pass  that attenuates
everything below 21mhz it can work....  I ran a 9pole filter and it did not
work well enough as in it either kills 24.9 or it doesn't kill the 20.17
spur much.  Possibly a Caur or elliptical filter would but they will
be touchy to make.

Not matter what one filter can't do it all.  Though 28mhz is the easiest
as the spur is further away.

Triple tuned band pass can, models were tried and maybe a double tuned might.

I do know that a triple tuned filter for 28 to 29 mhz was effective.  Phone (SSB)
is frm 28.3 to not much over 28.8 and for techs its only 28.3 to 28.5.  It would
have to pass down to 28.0 for CW.

The 15M filter is a tough nut as the spur is close at 23.55 when at 21.45.
That would require a triple tuned filter of some quality.  If it lowers the
spur by 20db it would do the job with a little excess.  It would be a very
good filter to do that.

Allison