Topics

Simple spur fix

 


This fix reduced the spurs by up to 10 db  and requires ONLY ONE part to be added.
There is big change above 10MHz in the board. There is some improvement below also.

The way I did it, CW may not work anymore will need some more mods for CW:

1. T2 - desolder the transformer wires that go to pin 3 and 5. Pin 1 has a square pad.
2. Bring out the two wires above board and join them together and solder.
3. Take a 45Mhz filter- 45M15 or  similar 2 pole - one xtal only. Solder one end of filter
to the wires of T2 pulled out. The center filter wire to ground at one end of R26. You will
see a ground via there.
4. Solder the third wire of filter to C10/R27 junction.

Thats it! This prevents the leaked TX signal that gets amplified by the 1st BiDi from getting into
the first mixer and creating havoc.

Farhan method of the same..much simpler and CW will work.

1. Remove R27
2. Solder the 45Mhz filter two extreme ends to the pads of the resistor.
3. Solder the center lead of the filter to the nearest ground. R13 is very near with a ground via.

The first method the extra filter will work in RX mode also and may help! In the second the
filter is only in the TX path..

Folks with DSA815 or better please share your feed back. The filter may work better properly
terminated etc.

Have fun!

--
Raj, vu2zap
Bengaluru, South India.

Lev
 

Would that be a good solution to make the IF amplifier tuned? Like this? Also, change the transistors to BFS17P, this is more suited to RF.

73s de HA5OGL


On Wed, Sep 5, 2018 at 10:46 AM Raj vu2zap <rajendrakumargg@...> wrote:

This fix reduced the spurs by up to 10 db  and requires ONLY ONE part to be added.
There is big change above 10MHz in the board. There is some improvement below also.

The way I did it, CW may not work anymore will need some more mods for CW:

1. T2 - desolder the transformer wires that go to pin 3 and 5. Pin 1 has a square pad.
2. Bring out the two wires above board and join them together and solder.
3. Take a 45Mhz filter- 45M15 or  similar 2 pole - one xtal only. Solder one end of filter
to the wires of T2 pulled out. The center filter wire to ground at one end of R26. You will
see a ground via there.
4. Solder the third wire of filter to C10/R27 junction.

Thats it! This prevents the leaked TX signal that gets amplified by the 1st BiDi from getting into
the first mixer and creating havoc.

Farhan method of the same..much simpler and CW will work.

1. Remove R27
2. Solder the 45Mhz filter two extreme ends to the pads of the resistor.
3. Solder the center lead of the filter to the nearest ground. R13 is very near with a ground via.

The first method the extra filter will work in RX mode also and may help! In the second the
filter is only in the TX path..

Folks with DSA815 or better please share your feed back. The filter may work better properly
terminated etc.

Have fun!

--
Raj, vu2zap
Bengaluru, South India.

Ashhar Farhan
 

I have tried the spur-zap fix. It really works. We need a big shout out to Raj, VU2ZAP. He has endlessly measured, tabulated but never given up on hunting this spur down.
I have tried it this morning too. Here are my results. I took out the R27, it is the 47 ohm series resistor that connects the tx side IF p to the front end mixer. In its place, i soldered the two ends of a spare 45 mhz filter. The ground pin of the filter was soldered to the ground end of C11.

73, f. 


On Wed, 5 Sep 2018, 14:28 Lev, <leventelist@...> wrote:
Would that be a good solution to make the IF amplifier tuned? Like this? Also, change the transistors to BFS17P, this is more suited to RF.

73s de HA5OGL

On Wed, Sep 5, 2018 at 10:46 AM Raj vu2zap <rajendrakumargg@...> wrote:

This fix reduced the spurs by up to 10 db  and requires ONLY ONE part to be added.
There is big change above 10MHz in the board. There is some improvement below also.

The way I did it, CW may not work anymore will need some more mods for CW:

1. T2 - desolder the transformer wires that go to pin 3 and 5. Pin 1 has a square pad.
2. Bring out the two wires above board and join them together and solder.
3. Take a 45Mhz filter- 45M15 or  similar 2 pole - one xtal only. Solder one end of filter
to the wires of T2 pulled out. The center filter wire to ground at one end of R26. You will
see a ground via there.
4. Solder the third wire of filter to C10/R27 junction.

Thats it! This prevents the leaked TX signal that gets amplified by the 1st BiDi from getting into
the first mixer and creating havoc.

Farhan method of the same..much simpler and CW will work.

1. Remove R27
2. Solder the 45Mhz filter two extreme ends to the pads of the resistor.
3. Solder the center lead of the filter to the nearest ground. R13 is very near with a ground via.

The first method the extra filter will work in RX mode also and may help! In the second the
filter is only in the TX path..

Folks with DSA815 or better please share your feed back. The filter may work better properly
terminated etc.

Have fun!

--
Raj, vu2zap
Bengaluru, South India.

Henning Weddig
 

Raj,

Your mod verifies my suspicion: apart form the mixing products within the mixer harmonics of the TX Bidi amp in front of the mixer also contribute to the problem, i.e. the first harmonic of the BiDI amp (2*45 MHz = 90 MHz!!!).

Placing an xtal filter directly into the path without any impedance transformation is a bad idea, the filter itself wants to see a higher impedance than 50 ohms. If this filter is not terminated correctly its insertion loss and passband ripple will be "ugly".

In addition: the connection from R27 to the transformer T2 will produce a source impedance close to 50 ohms. The output of the emitter follower Q21 is low impedance. PLUS: any complex load (especially capacitive) can cause Q21 to oscillate! A coure would then be a series resistor ( e.g. 100 ohms or so) to the base of Q21 .�

As the bidiamp produces harmoncis it would be easier to place a low pass filter (corner freq. a bit larger than 45 Mhz but "good"� attenaution at 90 MHz). A small pcb with good grounding should do--� please test this small pcb for real attenuation @90 MHz or if available a "commercial low pass filter e.g. from Mini Circuits could also do the trick for just a test.

A" fast design" using RFSIM99 for a 50 MHz five pole low pass filter (chebychev, 0.1 dB ripple, 50 ohm impedance)� C-L-C-L-C gives the following component values: 73 pF 218 nH 125 pF 218 nH 73 pf. The simulated attenuation at 90 MHz is 30 dB.�

73

Henning Weddig

DK5LV�

Am 05.09.2018 um 10:46 schrieb Raj vu2zap:


This fix reduced the spurs by up to 10 db� and requires ONLY ONE part to be added.
There is big change above 10MHz in the board. There is some improvement below also.

The way I did it, CW may not work anymore will need some more mods for CW:

1. T2 - desolder the transformer wires that go to pin 3 and 5. Pin 1 has a square pad.
2. Bring out the two wires above board and join them together and solder.
3. Take a 45Mhz filter- 45M15 or� similar 2 pole - one xtal only. Solder one end of filter
to the wires of T2 pulled out. The center filter wire to ground at one end of R26. You will
see a ground via there.
4. Solder the third wire of filter to C10/R27 junction.

Thats it! This prevents the leaked TX signal that gets amplified by the 1st BiDi from getting into
the first mixer and creating havoc.

Farhan method of the same..much simpler and CW will work.

1. Remove R27
2. Solder the 45Mhz filter two extreme ends to the pads of the resistor.
3. Solder the center lead of the filter to the nearest ground. R13 is very near with a ground via.

The first method the extra filter will work in RX mode also and may help! In the second the
filter is only in the TX path..

Folks with DSA815 or better please share your feed back. The filter may work better properly
terminated etc.

Have fun!

--
Raj, vu2zap
Bengaluru, South India.

 

Hi Henning,

I did try a filter but did not help much, I guess this issue needed a filter with greater end stop.
Allison said she tried a filter and found improvement.

I originally visualized a pad with LC matching with xtal conditioning filter. The pad would keep
the mixer peaceful and the BiDi will be happy with LC. Have to try this in the next days.

What I felt is that the BiDi was amping the TX signal that it picked up and sending to the
first mixer. When I lowered th power the spurs came down much more..
The spurs were moving together and apart in 3x 4x and 5x of the change in VFO freq.
They came together every 2.5 Mhz and were worst behaved at 15MHz and 22.5MHz

Raj


At 05-09-18, you wrote:

Raj,

Your mod verifies my suspicion: apart form the mixing products within the mixer harmonics of the TX Bidi amp in front of the mixer also contribute to the problem, i.e. the first harmonic of the BiDI amp (2*45 MHz = 90 MHz!!!).

Placing an xtal filter directly into the path without any impedance transformation is a bad idea, the filter itself wants to see a higher impedance than 50 ohms. If this filter is not terminated correctly its insertion loss and passband ripple will be "ugly".

In addition: the connection from R27 to the transformer T2 will produce a source impedance close to 50 ohms. The output of the emitter follower Q21 is low impedance. PLUS: any complex load (especially capacitive) can cause Q21 to oscillate! A coure would then be a series resistor ( e.g. 100 ohms or so) to the base of Q21 .�

As the bidiamp produces harmoncis it would be easier to place a low pass filter (corner freq. a bit larger than 45 Mhz but "good"� attenaution at 90 MHz). A small pcb with good grounding should do--� please test this small pcb for real attenuation @90 MHz or if available a "commercial low pass filter e.g. from Mini Circuits could also do the trick for just a test.

A" fast design" using RFSIM99 for a 50 MHz five pole low pass filter (chebychev, 0.1 dB ripple, 50 ohm impedance)� C-L-C-L-C gives the following component values: 73 pF 218 nH 125 pF 218 nH 73 pf. The simulated attenuation at 90 MHz is 30 dB.�

73

Henning Weddig

DK5LV�
Am 05.09.2018 um 10:46 schrieb Raj vu2zap:

This fix reduced the spurs by up to 10 db� and requires ONLY ONE part to be added.
There is big change above 10MHz in the board. There is some improvement below also.

The way I did it, CW may not work anymore will need some more mods for CW:

1. T2 - desolder the transformer wires that go to pin 3 and 5. Pin 1 has a square pad.
2. Bring out the two wires above board and join them together and solder.
3. Take a 45Mhz filter- 45M15 or� similar 2 pole - one xtal only. Solder one end of filter
to the wires of T2 pulled out. The center filter wire to ground at one end of R26. You will
see a ground via there.
4. Solder the third wire of filter to C10/R27 junction.

Thats it! This prevents the leaked TX signal that gets amplified by the 1st BiDi from getting into
the first mixer and creating havoc.

Farhan method of the same..much simpler and CW will work.

1. Remove R27
2. Solder the 45Mhz filter two extreme ends to the pads of the resistor.
3. Solder the center lead of the filter to the nearest ground. R13 is very near with a ground via.

The first method the extra filter will work in RX mode also and may help! In the second the
filter is only in the TX path..

Folks with DSA815 or better please share your feed back. The filter may work better properly
terminated etc.

Have fun!

--
Raj, vu2zap
Bengaluru, South India.

 

It is a idea to try Lev.. will give it a thought.

Raj


At 05-09-18, you wrote:
Would that be a good solution to make the IF amplifier tuned? Like this? Also, change the transistors to BFS17P, this is more suited to RF.

73s de HA5OGL

On Wed, Sep 5, 2018 at 10:46 AM Raj vu2zap < rajendrakumargg@...> wrote:

This fix reduced the spurs by up to 10 db  and requires ONLY ONE part to be added.
There is big change above 10MHz in the board. There is some improvement below also.

The way I did it, CW may not work anymore will need some more mods for CW:

1. T2 - desolder the transformer wires that go to pin 3 and 5. Pin 1 has a square pad.
2. Bring out the two wires above board and join them together and solder.
3. Take a 45Mhz filter- 45M15 or  similar 2 pole - one xtal only. Solder one end of filter
to the wires of T2 pulled out. The center filter wire to ground at one end of R26. You will
see a ground via there.
4. Solder the third wire of filter to C10/R27 junction.

Thats it! This prevents the leaked TX signal that gets amplified by the 1st BiDi from getting into
the first mixer and creating havoc.

Farhan method of the same..much simpler and CW will work.

1. Remove R27
2. Solder the 45Mhz filter two extreme ends to the pads of the resistor.
3. Solder the center lead of the filter to the nearest ground. R13 is very near with a ground via.

The first method the extra filter will work in RX mode also and may help! In the second the
filter is only in the TX path..

Folks with DSA815 or better please share your feed back. The filter may work better properly
terminated etc.

Have fun!

--
Raj, vu2zap
Bengaluru, South India.

Content-Type: application/pdf; name="if_amp-if_amp.pdf"
Content-Disposition: attachment; filename="if_amp-if_amp.pdf"
Content-ID: <f_jlowtrh20>
X-Attachment-Id: f_jlowtrh20

 

A a big thanks to Farhan, Allison and earlier to Jerry!

Cheers
Raj


At 05-09-18, you wrote:
I have tried the spur-zap fix. It really works. We need a big shout out to Raj, VU2ZAP. He has endlessly measured, tabulated but never given up on hunting this spur down.
I have tried it this morning too. Here are my results. I took out the R27, it is the 47 ohm series resistor that connects the tx side IF p to the front end mixer. In its place, i soldered the two ends of a spare 45 mhz filter. The ground pin of the filter was soldered to the ground end of C11.

73, f.Â


On Wed, 5 Sep 2018, 14:28 Lev, <leventelist@...> wrote:
Would that be a good solution to make the IF amplifier tuned? Like this? Also, change the transistors to BFS17P, this is more suited to RF.

73s de HA5OGL

On Wed, Sep 5, 2018 at 10:46 AM Raj vu2zap < rajendrakumargg@...> wrote:

This fix reduced the spurs by up to 10 db  and requires ONLY ONE part to be added.
There is big change above 10MHz in the board. There is some improvement below also.

The way I did it, CW may not work anymore will need some more mods for CW:

1. T2 - desolder the transformer wires that go to pin 3 and 5. Pin 1 has a square pad.
2. Bring out the two wires above board and join them together and solder.
3. Take a 45Mhz filter- 45M15 or  similar 2 pole - one xtal only. Solder one end of filter
to the wires of T2 pulled out. The center filter wire to ground at one end of R26. You will
see a ground via there.
4. Solder the third wire of filter to C10/R27 junction.

Thats it! This prevents the leaked TX signal that gets amplified by the 1st BiDi from getting into
the first mixer and creating havoc.

Farhan method of the same..much simpler and CW will work.

1. Remove R27
2. Solder the 45Mhz filter two extreme ends to the pads of the resistor.
3. Solder the center lead of the filter to the nearest ground. R13 is very near with a ground via.

The first method the extra filter will work in RX mode also and may help! In the second the
filter is only in the TX path..

Folks with DSA815 or better please share your feed back. The filter may work better properly
terminated etc.

Have fun!

--
Raj, vu2zap
Bengaluru, South India.

 

Further experimentation with Farhan's method,

Instead of removing R27 47Ohms and putting the filter there. I removed C22
and soldered the filter there.

The spurs on 20 dropped even lower @5W the spurs were -55 to -60dbm!
With proper termination the results may be better!

Raj


At 05-09-18, you wrote:

This fix reduced the spurs by up to 10 db  and requires ONLY ONE part to be added.
There is big change above 10MHz in the board. There is some improvement below also.

Farhan method of the same..much simpler and CW will work.

1. Remove R27
2. Solder the 45Mhz filter two extreme ends to the pads of the resistor.
3. Solder the center lead of the filter to the nearest ground. R13 is very near with a ground via.

Warren Allgyer <allgyer@...>
 

Filters suitable (I believe) for this mod are available here for $5.99 for a pair: 

https://www.ebay.com/itm/Tellurian-Technologies-45MHz-Crystal-Bandpass-Filter-T45U15BG-Matched-Pair/181956516605?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

If this solves the spur problem then it is also a very positive step toward solving the overall PA filter issue. Absent the spurs, and with the board layout problems addressed, the existing four filters should be adequate. I am about to receive a prototype 4 filter board that can accept the existing components for testing. 

Very nice work!

WA8TOD

 

At 05-09-18, you wrote:

Filters suitable (I believe) for this mod are available here for $5.99 for a pair:

https://www.ebay.com/itm/Tellurian-Technologies-45MHz-Crystal-Bandpass-Filter-T45U15BG-Matched-Pair/181956516605?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

If this solves the spur problem then it is also a very positive step toward solving the overall PA filter issue. Absent the spurs, and with the board layout problems addressed, the existing four filters should be adequate. I am about to receive a prototype 4 filter board that can accept the existing components for testing.

Very nice work!

WA8TOD

Gordon Gibby <ggibby@...>
 

​I ordered the pair.   Interesting and exciting development.

gordon



From: BITX20@groups.io <BITX20@groups.io> on behalf of Warren Allgyer <allgyer@...>
Sent: Wednesday, September 5, 2018 7:50 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Simple spur fix
 

Filters suitable (I believe) for this mod are available here for $5.99 for a pair: 

https://www.ebay.com/itm/Tellurian-Technologies-45MHz-Crystal-Bandpass-Filter-T45U15BG-Matched-Pair/181956516605?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

If this solves the spur problem then it is also a very positive step toward solving the overall PA filter issue. Absent the spurs, and with the board layout problems addressed, the existing four filters should be adequate. I am about to receive a prototype 4 filter board that can accept the existing components for testing. 

Very nice work!

WA8TOD

Ashhar Farhan
 

If you guys want to try out the mod, i can have some spares shipped out.
- f

On Wed, 5 Sep 2018, 19:06 Gordon Gibby, <ggibby@...> wrote:

​I ordered the pair.   Interesting and exciting development.

gordon



From: BITX20@groups.io <BITX20@groups.io> on behalf of Warren Allgyer <allgyer@...>
Sent: Wednesday, September 5, 2018 7:50 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Simple spur fix
 

Filters suitable (I believe) for this mod are available here for $5.99 for a pair: 

https://www.ebay.com/itm/Tellurian-Technologies-45MHz-Crystal-Bandpass-Filter-T45U15BG-Matched-Pair/181956516605?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

If this solves the spur problem then it is also a very positive step toward solving the overall PA filter issue. Absent the spurs, and with the board layout problems addressed, the existing four filters should be adequate. I am about to receive a prototype 4 filter board that can accept the existing components for testing. 

Very nice work!

WA8TOD

ajparent1/KB1GMX
 

Farhans method...

1. Remove R27
2. Solder the 45Mhz filter two extreme ends to the pads of the resistor.
3. Solder the center lead of the filter to the nearest ground. R13 is very near with a ground via.

Sounds like something that may seriously help. I tried a low pass filter there and it
did help but not enough.  Time and other pressing things never got back to try a
band pass filter.

One I'd liked to have tried is just moving the filter to the other end of the amp.

What is the impact on RX performance?

Allison

Jerry Gaffke
 

Raj,

Very cool that you found such an easy and cheap solution!
Though not so sure I was a part of it in any way.

So that 2.5mhz thing had nothing to do with the si5351, was caused by feedback from the power amp?
Very curious.  I have no idea how that 2.5mhz factor came about.

The extra 45mhz filter would solve the power amp feedback issue.
That was probably all the low frequency rumble in Warren's plots of a few weeks ago.
This would not be fixed by a LPF.

I had assumed that the 45mhz-DialFreq spur was generated inside the mixer at D1,D2.
But apparently the IF amp at Q20,21,22 was generating enough of a second harmonic 
that the resultant 90mhz energy minus the clk2 VFO was going straight on through
the mixer in the expected manner.  As Henning has stated previously.

I'm sure termination impedances around the new filter could be an issue somehow.
But if it works, I'm ok with that!
Passband ripple might not be much of an issue, as the filter is much wider than what we actually use.
Extra insertion loss may not be much of an issue either, especially with the crystal only in
the TX path since we have gain to burn there.  Less TX signal into the D1,D2 mixer is a good thing.

Good job!

Jerry, KE7ER


On Wed, Sep 5, 2018 at 04:11 AM, Raj vu2zap wrote:
A a big thanks to Farhan, Allison and earlier to Jerry!

jim
 


On Wednesday, September 5, 2018, 6:49:10 AM PDT, Ashhar Farhan <farhanbox@...> wrote:


If you guys want to try out the mod, i can have some spares shipped out.
- f

I'd like one if there are enough to go around

Jim (ab7vf)

 

Allison,

the filter will be only in the TX path, the roofing filter should also be in RX path.
why amplify all the mixes from the first mixer. A good change in the architecture of
uBitx.

Raj

At 05-09-18, you wrote:
Farhans method...

1. Remove R27
2. Solder the 45Mhz filter two extreme ends to the pads of the resistor.
3. Solder the center lead of the filter to the nearest ground. R13 is very near with a ground via.

Sounds like something that may seriously help. I tried a low pass filter there and it
did help but not enough. Time and other pressing things never got back to try a
band pass filter.

One I'd liked to have tried is just moving the filter to the other end of the amp.

What is the impact on RX performance?

Allison

R. E. Klaus
 

How could I obtain one of the mods?
Thanks
Roxie -- K1AUS

Kees T
 

The cheaper ebay guy says he has 62 of them left plus postage. Someone could make him a deal for all and reship them to uBITX owners for a $0.72 stamp on an envelope.

73 Kees K5BCQ

ajparent1/KB1GMX
 

Farhan,

Mine is a mess due to other work.
>>If you guys want to try out the mod, i can have some spares shipped out.

However I have a few friends around me I can help using this mod and also a relay fix
already posted.


Allison

Jerry Gaffke
 

Raj, 

With the new filter placed in both the TX and RX paths as you originally proposed,
should be able to restore CW operation by lifting the bottom end of R105, tie it to T2 pins 3,5

I suspect that's already occurred to you, and it's a matter of finding the time to try it.

If we can have a clean all-of-HF transmitter without adding switchable bandpass filters 
in front of the first mixer, this is a significant advance to the art.

Jerry, KE7ER



On Wed, Sep 5, 2018 at 08:05 AM, Raj vu2zap wrote:
the filter will be only in the TX path, the roofing filter should also be in RX path.
why amplify all the mixes from the first mixer. A good change in the architecture of
uBitx.