Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: Phase Shift network for HB noise canceller  (Read 16115 times)
G8JNJ
Member

Posts: 734


WWW

Ignore
« on: June 21, 2011, 12:44:23 PM »

Hi All,

I'm currently experimenting with null steering / noise cancelling circuits. These use a second 'noise' antenna which is combined with differing amplitude and phase into the main receive antenna path. So that local interference can be cancelled out.

Can anyone help with a circuit for a broadband variable phase shift circuit, which will operate over the 1.5 to 30 MHz frequency range?

I've tried a few designs such as used in the ANC-4, MFJ & SEM / DK9NL X-Phase

http://www.timewave.com/support/ANC-4/ANC_4TW8x11a.pdf

http://www.mfjenterprises.com/pdffiles/MFJ-1025.pdf

http://www.mydarc.de/dh3wl/dk9nl/X-Phase-Dateien/X-Phase.htm


These all tend to use RC bridges, which seem to suffer from large amplitude variations when adjusted.

I've heard that the DXE NCC-1

http://www.dxengineering.com/pdf/NCC-1-Rev5b.pdf

Uses a different method to provide the required variable phase shift. But I'm not sure what principle is used other than it is dc controlled. So maybe it’s a 90 degree hybrid and inductance / varicap network ?

Any suggestions gratefully received. Send me a private e-mail if you wish.

 martinmail2007-g8jnj@yahoo.com

Thanks,

Martin - G8JNJ

www.g8jnj.webs.com
Logged
KA4POL
Member

Posts: 2409




Ignore
« Reply #1 on: June 21, 2011, 10:14:28 PM »

Hi Martin,

some time ago I discussed this with DK9NL and he admitted that the results with his circuit are very shaky. The signal is usually blanked out as well, at least in part. You can only cancel the noise that comes from other directions than the wanted signal. I can't tell you anything about the other circuits but assume similar difficulties.

73 Dieter
Logged
G8JNJ
Member

Posts: 734


WWW

Ignore
« Reply #2 on: June 22, 2011, 12:39:46 AM »

Hi Dieter,

Yes, the whole principle depends upon having enough discrimination between the 'noise' and 'main' antennas. However I'm fairly confident that I can achieve this.

The main problem from an operational point of view is having too many interactive controls that require adjustment each time the frequency is changed.

I like the look of the DXE NCC-1 beacuse of the wide range, nearly constant amplitude, phase shift control.

Regards,

Martin - G8JNJ

www.g8jnj.webs.com
Logged
G3RZP
Member

Posts: 1280




Ignore
« Reply #3 on: June 22, 2011, 01:00:32 AM »

Once you have the I and Q channels, a goniometer is what is needed. However, getting a goniometer to cover the HF band isn't easy. I came up with a scheme where the HF band is upconverted to around 60 MHz, and then a wideband quadrature pahse shift network is used t- just needs to cover 60 to 90 MHz. Then you need a goniometer for that range, which isn't so difficult to do, especially as you are not interested in semi-circular, qudarantel and octanatal errors. It does need some careful mechanics though, as you would want to shield it to minimise amplitude variations.
Logged
G8JNJ
Member

Posts: 734


WWW

Ignore
« Reply #4 on: June 22, 2011, 09:32:19 AM »

Hi Peter,

OK thanks for this - I had thought of upconverting and using a 90 degree hybrid and IQ vector modulator. But it seemed to be getting way too complicated. I like things to be simple......

Regards,

Martin - G8JNJ

www.g8jnj.webs.com
Logged
G3RZP
Member

Posts: 1280




Ignore
« Reply #5 on: June 22, 2011, 02:43:33 PM »

I too like things to be simple. In fact, I could only really use a canceller on 160 and 80. Now if I could get the goniometer from an old Marconi Marine Lodestar II, it would be easy.... But the goniometer is the devcie you need.
Logged
G8JNJ
Member

Posts: 734


WWW

Ignore
« Reply #6 on: June 23, 2011, 12:50:49 AM »

Hi Peter,

Did you attend the Newbury rally last weekend ? I saw at least two of different flavours amongst the junk.

However I don't think it would actually be that difficult to make one. Especially if it was fairly small.

I wouldn't bother trying to tune the thing. I'd just drive a small number of turns from a decent pair of amplifiers. I assume I'd get quite a wide operating frequency range by doing this as the phase relationship is purely determined by the coil position ?

Thanks for the suggestion. I'll build up a simple prototype and see if it will do what I require.

Regards,

Martin - G8JNJ

www.g8jnj.webs.com

Logged
G8JNJ
Member

Posts: 734


WWW

Ignore
« Reply #7 on: June 23, 2011, 01:18:59 AM »

Hi Peter,

I just saw a previous discussion on RSGBtech regarding Goniometers. They don't work quite as I'd imagined.

The device normally only provides a method of varying amplitudes of signals relative to each other (except for a very rapid phase transition just after the rotor and one stator coil pass through the null) . In order to obtain a variable phase shift the coils have to be fed in quadrature (or tuned which produces a similar effect).

So I'm back to the original problem of how to obtain either a variable 0-180 degree or fixed 90 degree phase shift over the frequency range 1 - 30MHz (or greater). Without introducing a large amplitude variation in the process.

Regards,

Martin - G8JNJ

www.g8jnj.webs.com
Logged
G3RZP
Member

Posts: 1280




Ignore
« Reply #8 on: June 23, 2011, 03:51:02 AM »

OK, the two field coils have to be fed in phase quadrature. For D/F, this is done by having two loops at right angles (the Bellini-Tosi system) or 4 verticals at the corners of a square, with the antennas at diagonal corners feeding the goniometer (the Adcock system).

Making reasonable quadrature networks to cover a 2:1 frequency range isn't too difficult - there was an article some 40 or 50 years ago in either the IEEE or the IRE by Reed Fisher,  (who happens to be W2CQH) on it. So for the HF amateur bands, you would only need three networks ( 160 and 80, 40/30 and 20, 17/15/12 and 10). Or upconvert, as quadrature LO drive is easy, and run at fixed frequency. Then down convert with the same oscillator.

I would go for the switched quadrature networks myself.
Logged
G8JNJ
Member

Posts: 734


WWW

Ignore
« Reply #9 on: June 23, 2011, 04:02:08 AM »

Hi Peter,

If I have to add bandswitching then the designs using an input bandpass filter network which also provides the required range of phase shift may be a better option.

The G4LNA and VK5BR designs uses this method.

http://www.qsl.net/g4lna/pages/noisdes.html

http://users.tpg.com.au/users/ldbutler/NoiseCancelling.htm

Regards,

Martin - G8JNJ

www.g8jnj.webs.com



« Last Edit: June 23, 2011, 04:05:50 AM by G8JNJ » Logged
PA1ZP
Member

Posts: 688




Ignore
« Reply #10 on: June 24, 2011, 12:33:20 PM »

Hi everyone

I build a noise canceler to and got help from a friend PA3ERT.

I like the design and it works very well.

But the BIG BUT is, your noise antenna has to be as noise free as possible and your main antenna has to be even better.

I do not know abt gain influences etc.
I am glad the thing works.

All parts were found in junk-box.
Only had to pay 13 euros for cabinet material.

If you want the schematics you are free to look at:
http://blogs.hamstart.net/blog02/archives/2043

Do not be distracted by cabinet and tube .
This is just being big fun.


I have 5 different bandtransformers in use now and it helps to eliminate allmost all the noise.
But more switches meens more things to set and try, I took me some time to get used to canceling out some noises.

73 Jos
Logged
G8JNJ
Member

Posts: 734


WWW

Ignore
« Reply #11 on: June 24, 2011, 01:13:34 PM »

Hi Jos,

Nice construction !

Yes the basic circuit is the same as the G4LNA design.

http://www.qsl.net/g4lna/pages/noisdes.html

Personally I think the input tuning arrangment used by VK5BR is better.

The coils are easier to implement and the twin tuned circuits provide a greater range of phase adjustment, compaired to just having one tuned network.

http://users.tpg.com.au/users/ldbutler/NoiseCancelling.htm

However I'm sure both designs work well with a properly sited noise antenna.

Regards,

Martin - G8JNJ

www.g8jnj.webs.com




 
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #12 on: June 27, 2011, 03:29:06 AM »

Martin,

The German link you posted will act just like a goniometer with only one phase (this includes 180 shifts as being one phase) on field coils. It will not shift phase, other than an abrupt transition to 180 flip. I believe the German link is a copy or takeoff of the DeMaw QRN squasher, which also did not change phase except for the abrupt 180 flip at zero voltage crossing.

The problem with L/C networks for phase shift, like the ANC4 and others, is they seriously change amplitude as phase shift is large. They also have limited range. Making it a pi helps so long as reactance of the capacitors is low compared to load and source impedances, as does making it a T with higher reactance, but it really becomes a matching network and so as components are adjusted there will be areas where amplitude changes pretty fast with phase. This makes it more complex than necessary to null something, but again you suffer limited phase adjustment range.

The MFJ and DXE schemes are far better than the ANC4 and other L/C circuit systems. The DXE unit is exceptionally gain-flat as phase is changed, and has by far the widest rotation.

73 Tom

Logged
G8JNJ
Member

Posts: 734


WWW

Ignore
« Reply #13 on: June 27, 2011, 09:21:11 AM »

Hi Tom,

Thanks for your comments. They pretty much concur with my observations.

The units using RC networks actually do provide some degree of variable phase shift. But it's quite unpredictable and well away from what would probably be considered to be the intended design frequency. I can achieve a 30dB null with one of the relatively simple circuits, over a wide range of differing phase relationships. However it takes forever to adjust due to the bad interaction between controls.

Vector modulation seems like the way to go. As long as the phase shift networks can be designed to track each other throughout the required operating frequency range, this should work OK.

Do you happen to know the principle used to obtain the phase shift in the DXE NCC-1 ?

Regards,

Martin - G8JNJ


www.g8jnj.webs.com
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #14 on: June 27, 2011, 10:29:11 AM »

The units using RC networks actually do provide some degree of variable phase shift.

I designed the DXE unit so I'm limited on what I can say. I'm sure you will find a solution, but I would not look towards L C circuits.
Logged
Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!