WUB WUB WUB sound and drop-outs, but only at 20m #FT8 #txaudio #signalink


Christian Sandvig
 

Hi, I'm a new ham and I wondered if anyone can help me figure out a WSJTX/FT8 problem I am having. I've tried searching for the answer but I am so new at this that I think I don't know the right keywords to describe the issue.

I have a laptop connected to a SignaLink USB, to an Icom M802 (150W max), Ameritron SWR/Wattmeter, then a 1:1 balun, then an Icom AT-140 antenna tuner, and a ~34ft random wire antenna with a ~34ft counterpoise.

When I try to transmit FT8 at 30m and up everything works fine and I am able to make contacts. The wattmeter is showing 6-10W forward power depending on how I set the TX dial of the SignaLink. However when I transmit FT8 at 20m usually no one responds. I noticed some weird fluctuations on the wattmeter like a rhythmic movement and forward power periodically drops to zero. This does not happen at other frequencies.

To diagnose this problem I have used an SDRplay RSP1A connected to a front end isolator and a long piece of wire. I am also running RCForb and I can monitor my own signal from a generous Ham who has an open access rig in my town.  When I monitor myself broadcasting FT8 at 20m I hear a sound that seems like a deep-voiced man saying "WUB WUB WUB WUB WUB WUB" over my signal. If I turn the TX dial on the SignaLink down all the way this improves but it never goes away. 

In addition, during an FT8 transmission there are 1-2 periodic drops when it seems like my signal cuts out entirely, sort of like when your hand slips off of the PTT button for a second and you press it again.  However, the SignaLink PTT light remains on the whole time and the Icom "TX" indicator remains on the whole time. The wattmeter does drop to zero during the drop-outs. So I am wondering if the WSJTX software is producing these drop-outs somehow by sending the PTT signal but briefly not sending any sound?  Again I am only seeing this on 20m.

I can transmit directly from the radio (no computer) on 20m or any band using my voice and I do not see these problems. 

I have carefully followed the setup instructions for all of the software and for the SignaLink (to set levels, etc.) if that helps. I'm not sure what to try now!  Any ideas?

Here is a screenshot of the SDRuno waterfall showing a problem transmission. I think the wavy waterfall is the man saying WUB WUB WUB and I marked the 2 drop-outs with arrows.  

Really hoping someone can help!
KB0ATY
(brand new ham)



Bill Somerville
 

On 10/04/2021 23:59, Christian Sandvig wrote:
Hi, I'm a new ham and I wondered if anyone can help me figure out a WSJTX/FT8 problem I am having. I've tried searching for the answer but I am so new at this that I think I don't know the right keywords to describe the issue.

I have a laptop connected to a SignaLink USB, to an Icom M802 (150W max), Ameritron SWR/Wattmeter, then a 1:1 balun, then an Icom AT-140 antenna tuner, and a ~34ft random wire antenna with a ~34ft counterpoise.

When I try to transmit FT8 at 30m and up everything works fine and I am able to make contacts. The wattmeter is showing 6-10W forward power depending on how I set the TX dial of the SignaLink. However when I transmit FT8 at 20m usually no one responds. I noticed some weird fluctuations on the wattmeter like a rhythmic movement and forward power periodically drops to zero. This does not happen at other frequencies.

To diagnose this problem I have used an SDRplay RSP1A connected to a front end isolator and a long piece of wire. I am also running RCForb and I can monitor my own signal from a generous Ham who has an open access rig in my town.  When I monitor myself broadcasting FT8 at 20m I hear a sound that seems like a deep-voiced man saying "WUB WUB WUB WUB WUB WUB" over my signal. If I turn the TX dial on the SignaLink down all the way this improves but it never goes away.

In addition, during an FT8 transmission there are 1-2 periodic drops when it seems like my signal cuts out entirely, sort of like when your hand slips off of the PTT button for a second and you press it again.  However, the SignaLink PTT light remains on the whole time and the Icom "TX" indicator remains on the whole time. The wattmeter does drop to zero during the drop-outs. So I am wondering if the WSJTX software is producing these drop-outs somehow by sending the PTT signal but briefly not sending any sound?  Again I am only seeing this on 20m.

I can transmit directly from the radio (no computer) on 20m or any band using my voice and I do not see these problems.

I have carefully followed the setup instructions for all of the software and for the SignaLink (to set levels, etc.) if that helps. I'm not sure what to try now!  Any ideas?

Here is a screenshot of the SDRuno waterfall showing a problem transmission. I think the wavy waterfall is the man saying WUB WUB WUB and I marked the 2 drop-outs with arrows.

Really hoping someone can help!
KB0ATY
(brand new ham)
Hi Christian,

audio issues on one band while transmitting are almost certainly due to RFI. Long wire aerials can lead to considerable RF currents in the shack and such currents can disrupt signals on interconnecting cables between equipment. Good bonding between the equipment chassis can help but taking steps to reduce the amount of RF returning to the shack on the outside of feeders is better still.

73
Bill
G4WJS.


Karl Beckman
 

You may think your antenna length is random, but 34 feet per side makes it almost a full-wave dipole on the 20M band, therefore it has a VERY high impedance at its feedpoint, certainly NOT 50 ohms, more like 2500. 
First, I would suggest that If that 1:1 balun is labeled as a 50 ohm device, remove it completely when operating on 14MHz or above. Otherwise (and also when working lower frequency bands) move  the 1:1 balun to the output side of the tuner.
Now be sure you have a solid ground on both the tuner and transmitter.  Do not rely on the shield of the coax going throiugh your SWR meter.
Next, you need to loop the USB cable through a Type 72 ferrite "donut" core several times to reduce RF getting into the computer's USB audio port circuitry.  It doesn't hurt to put a snap-on ferrite on the analog audio & control cable going from the Signalink USB into the rig too. 
Last, reduce your transmit output power to 50W or less.  FT8 is a low power digital mode that decodes signals at least 24 dB below the noise floor.  I'll look for you on 14073.
--
Karl  WA8NVW  OH  EN91bi
WA8NVW@...
in WSJTX@groups.io


Jim Brown
 

On 4/11/2021 10:34 AM, Karl Beckman wrote:
Next, you need to loop the USB cable through a Type 72 ferrite "donut" core
What's the source of this advice? Who makes a #72 material? It's not Fair-Rite. Can you point us to a technical data sheet for it?

Before resorting to ferrites on cables, the station should include proper grounding and bonding. Study N0AX's ARRL book on the topic, to which I contributed, and/or slides for my tutorial talk.

http://k9yc.com/GroundingAndAudio.pdf

73, Jim K9YC


William Smith
 

https://www.tdk-electronics.tdk.com/download/528858/f166f3fdc24325c6b592a75777608210/pdf-n72.pdf

Says #72 ferrite is good for 25 to 300 KHz, so there may be a better choice. 🤷‍♂️

73, Willie N1JBJ

On Apr 11, 2021, at 4:00 PM, Jim Brown <k9yc@...> wrote:

On 4/11/2021 10:34 AM, Karl Beckman wrote:
Next, you need to loop the USB cable through a Type 72 ferrite "donut" core

What's the source of this advice? Who makes a #72 material? It's not Fair-Rite. Can you point us to a technical data sheet for it?

Before resorting to ferrites on cables, the station should include proper grounding and bonding. Study N0AX's ARRL book on the topic, to which I contributed, and/or slides for my tutorial talk.

http://k9yc.com/GroundingAndAudio.pdf

73, Jim K9YC




Larry Banks
 

25 to 300 KHz is a bit low for the ham bands.

73 -- Larry -- W1DYJ

 

Sent: Sunday, April 11, 2021 16:16
Subject: Re: [WSJTX] WUB WUB WUB sound and drop-outs, but only at 20m #FT8 #signalink #txaudio
 
https://www.tdk-electronics.tdk.com/download/528858/f166f3fdc24325c6b592a75777608210/pdf-n72.pdf
 
Says #72 ferrite is good for 25 to 300 KHz, so there may be a better choice. 🤷‍♂️
 
73, Willie N1JBJ

On Apr 11, 2021, at 4:00 PM, Jim Brown <k9yc@...> wrote:

On 4/11/2021 10:34 AM, Karl Beckman wrote:
Next, you need to loop the USB cable through a Type 72 ferrite "donut" core

What's the source of this advice? Who makes a #72 material? It's not Fair-Rite. Can you point us to a technical data sheet for it?

Before resorting to ferrites on cables, the station should include proper grounding and bonding. Study N0AX's ARRL book on the topic, to which I contributed, and/or slides for my tutorial talk.

http://k9yc.com/GroundingAndAudio.pdf

73, Jim K9YC








Jim Brown
 

On 4/11/2021 1:16 PM, William Smith wrote:
https://www.tdk-electronics.tdk.com/download/528858/f166f3fdc24325c6b592a75777608210/pdf-n72.pdf <https://www.tdk-electronics.tdk.com/download/528858/f166f3fdc24325c6b592a75777608210/pdf-n72.pdf>
Says #72 ferrite is good for 25 to 300 KHz, so there may be a better choice. 🤷‍♂️
Yes, far better choices. Fair-Rite #31 is the weapon of choice from 160M to 2M. Study my tutorial on killing RX noise to boost contest scores. k9yc.com/publish.htm

73, Jim K8YC


Christian Sandvig
 


Thanks to all; this is very helpful. I wonder if you will entertain two follow-up questions.

First question: How are WSJTX users bonding a laptop? Jim's slide presentation identifies the connection to the computer as the most important bonding step to achieve to reduce noise reduction, and in my setup I don't see the distortion unless I am using the laptop. But for years now my laptops have had plastic cases. The advice I see in many places online does not seem up-to-date. For example -- I see that I am supposed to attach a ground wire to the screw for a serial port or VGA port. Modern laptops have neither.

Second question: Can anyone direct me to a discussion of bonding and grounding for a 2nd floor ham shack? I see that just running a long ground wire downstairs will not have the desired effect, but I'm not sure what to do instead.

I am doing my reading on this but I am so new I feel like only a small percentage of it gets through -- I thought that a resonant antenna was good!  I appreciate your patience and help.

KB0ATY


Jim Brown
 

On 4/12/2021 7:24 AM, Christian Sandvig wrote:
First question: How are WSJTX users bonding a laptop? Jim's slide presentation identifies the connection to the computer as the most important bonding step to achieve to reduce noise reduction, and in my setup I don't see the distortion unless I am using the laptop. But for years now my laptops have had plastic cases. The advice I see in many places online does not seem up-to-date. For example -- I see that I am supposed to attach a ground wire to the screw for a serial port or VGA port. Modern laptops have neither.
USB ports SHOULD include a metallic shell that SHOULD be connected to the chassis. Worst case, buy a quality USB cable and sacrifice it. Even better, buy a decent USB interface that has exposed chassis or bonding screw and bond to it. You're more likely to find that on an interface for the semi-pro audio small home studio market, and Tascam is a leading player there. I'd feel fine buying a used vintage box from a highly rated auction vendor. The Tascam US100 I bought new around 2010 is still on my operating desk, and several years ago I bought a second on the auction site to have as a spare. The Numark unit in the report below is also a fine choice; it was discontinued for a while, then back into production. Again, I'd buy one used from a good vendor.

http://k9yc.com/USB_Interfaces.pdf
Second question: Can anyone direct me to a discussion of bonding and grounding for a 2nd floor ham shack? I see that just running a long ground wire downstairs will not have the desired effect, but I'm not sure what to do instead.
N0AX's ARRL Book is a very good resource. I'm currently working with Ward on a second version, and that's one of the things that is being expanded.

The short answer is bond everything within your shack using the methods and logic in my slide deck. From there, run a bonding wire to at least one driven rod as close as practical to directly below the shack, and bond that to all the other grounds (main power entry, telco, CATV, your antenna bonds). If practical, add rods to form a perimeter ground (or a partial one).

Route all feedlines to a bonding panel at ground level, bonded to those rods, and bond their shields in that panel (using feed-throughs for coax lines). Extend from that panel to a panel in the shack, where arrestors are installed. The principle here is that arrestors work by shorting the center to the shield, and they should be as close as practical to the equipment they're protecting, because strike current in the shield can induce a differential voltage inside the coax.

73, Jim K9YC


Jim Shorney
 

I can vouch for the US-100. I bought one after reading the K9YC papers (thanks Jim!) and did not regret it. I have three of them now, all found on eBay for around $40 each. Nice little box! You will see sellers asking more for them but you can find a deal if you are patient or you can spend more if you think it is worth it for you.

73

-Jim
NU0C


On Mon, 12 Apr 2021 12:48:46 -0700
"Jim Brown" <k9yc@audiosystemsgroup.com> wrote:

On 4/12/2021 7:24 AM, Christian Sandvig wrote:
First question: How are WSJTX users bonding a laptop? Jim's slide
presentation identifies the connection to the computer as the most
important bonding step to achieve to reduce noise reduction, and in my
setup I don't see the distortion unless I am using the laptop. But for
years now my laptops have had plastic cases. The advice I see in many
places online does not seem up-to-date. For example -- I see that I am
supposed to attach a ground wire to the screw for a serial port or VGA
port. Modern laptops have neither.
USB ports SHOULD include a metallic shell that SHOULD be connected to
the chassis. Worst case, buy a quality USB cable and sacrifice it. Even
better, buy a decent USB interface that has exposed chassis or bonding
screw and bond to it. You're more likely to find that on an interface
for the semi-pro audio small home studio market, and Tascam is a leading
player there. I'd feel fine buying a used vintage box from a highly
rated auction vendor. The Tascam US100 I bought new around 2010 is still
on my operating desk, and several years ago I bought a second on the
auction site to have as a spare. The Numark unit in the report below is
also a fine choice; it was discontinued for a while, then back into
production. Again, I'd buy one used from a good vendor.

http://k9yc.com/USB_Interfaces.pdf


Karl Beckman
 
Edited

Sorry for the simple typo that seems to have upset the entire rather vocal ham community.  I meant to type '73' instead of '72', but then I missed reading table note 7 regarding the non-availability of larger core sizes and substitution of type 75 for many uses.
Below are the usage recommendations copied from the Palomar Engineer web page https://palomar-engineers.com/ferrite-products/ferrite-cores/ferrite-mix-selection
...

The table below gives our recommended applications for various mixes and effective frequency ranges

Mix # Material Initial Permeability RFI/EMI Common Mode Suppression Range Tuned Circuits – Coil
Wide Band Transformer
31  (1) MnZn 1500 1-300 MHz 1:1 only, <300 MHz
43 (2) NiZn 800 25-300 MHz < 10 MHz 3-60 MHz
52 (6) NiZn 250  200-1000 MHz < 20 MHz 1-60 MHz
61 (3) NiZn 125 200-1000 MHz  <100 MHz 1-300 MHz
73 (7) MnzN 2500 < 50 MHz < 2 MHz <10 MHz
75/J (4) MnZn 5000 150 KHz – 10 MHz  < .75 MHz  .1-10 MHz


Notes
...

(4) Mix 75 (also known as Mix J) is a high permeability MnZn ferrite intended for a range of broadband and pulse transformer applications and common-mode inductor designs. Excellent for common mode suppression on AM broadcast frequencies from 500 KHz-1.8 MHz.  Also very useful for medical instrument transducer isolation, inverter assemblies, inductive motors and control units. Curie temperature >140 C. Mix 75 is available in TOROIDS,   SLIP ON BEADS, ,and SNAP ON SPLIT BEADS,

  • Optimized to solve low-frequency EMI issues between 150 kHz and 10 MHz
  • Suppresses common-mode noise up to 30 MHz
  • Replaces expensive line filters to mitigate conducted noise
  • Split core performance closely resemble solid cores, allowing seamless transitions from test to production

Applications:

  • Automotive: Inverter assemblies, inductive motors
  • Consumer Electronics: Power supplies, peripheral cables
  • Medical: Transducer isolation, human-machine interfaces
  • White Goods: Control units, motors

(7) Mix 73 only available in small bead size, for larger inside diameter requirements, use mix 75 or 77

Feel free to discuss among yourselves the relative performance  of your favorite blend, remembering that the RF freq inducing trouble is 14 MHz.  And a toroid from Fair-Rite is NOT the same as a ferrite toroid.

[p.s.  JIM K9YC - I've read and downloaded your excellent papers and understand the principles involved in Pin 1.  I'm also one of (numerous) contributors to the Motorola R56 Site Standards document.]


- - 
Karl  WA8NVW  OH
WA8NVW@...
in WSJTX@groups.io


Jim Brown
 

On 4/13/2021 9:33 AM, Karl Beckman wrote:
Below are the usage recommendations copied from the Palomar Engineer
Palomar Engineering is hardly an authority on this. They are a MERCHANT.

73, Jim K9YC