3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

General discussions
glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

Post by glovisol » Tue Apr 09, 2019 11:59 am

Do you have two receivers but just one antenna? Do you want to compare two RSP's in exactly the same conditions and with the same antenna, but with the minimum possible signal loss? Just read on.....

To the next post for amendment purposes.

Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: 3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

Post by glovisol » Tue Apr 09, 2019 12:34 pm

A 3 dB hybrid coupler is a balanced symmetrical transformer, the equivalent of the old hybrids used in telephony, but at much higher frequencies. The main property of the hybrid is directionality: it lets signal travel in the forward way, but not in the reverse way. In our case the hybrid has one 50 Ohm input (antenna) two equal 50 Ohm outputs (receivers) and a balanced load resistor.

If we call the input A and the two outputs B and C, power in A will be equally shared into B and C (apart from a small insertion loss) BUT, if we connect power to B, most of this power will be dissipated in the balanced load and very little will reach A. The same the other way around. This is what we call isolation between ports.

Any receiver will always radiate a little of its local oscillator from the antenna input. If we connect two receivers to the same antenna without isolation, apart from mismatch loss, chances are that each receiver's local oscillator will interfere with the other. This effect is minimised by using a hybrid coupler, because the local oscillator coming out of B will be dissipated in the balanced load and will not reach A and vice-versa.

The 3 dB hybrid coupler presented here has a minimal power imbalance: e.g. power at A splits equally into B and C minus a very low insertion loss of 0.5 dB for each output. This means that if we have -50 dBm at A, we shall get a power at each output B and C equal to: Po = 50-3-0.5= 46.5 dBm. A negligible loss for any antenna. Furthermore isolation between B and C is between 18 and 20 dB. This means that if the local oscillator of receiver connected at B has a level of -100 dBm it will reach receiver connected to A with a level of -118 to -120 dbm.

More detail in next post.
Attachments
3 dB Hybrid 4.JPG
Figure 1 - 50 Ohm 3 dB hybrid coupler
3 dB Hybrid 4.JPG (169.41 KiB) Viewed 31885 times

Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: 3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

Post by glovisol » Wed Apr 10, 2019 6:47 am

THE 50 OHM SYMMETRICAL 3 dB COUPLER

The balanced transformer is built on a ferrite binocular core selected for this frequency range. The binocular core used is model BN-73-202, where 73 is the ferrite material and 202 is the size: B(length)=14.35 mm / A(width)=13.3 mm / C(thickness)=7.5 mm. With material type 73, mu=2500. LC meter and Q meter measurements gave the uniform result that N=7 turns produce L=580 uH and a wideband Q response of Q=3.5 @1 MHz. From this data:

Al = ((100^2)*L)/(N^2) = 118*10^3 mH/1000 t or Al = 11,800 uH/100t.
This Al is now our parameter for wideband transformer calculations.

=======

For our wideband LF 3 dB hybrid transformer the 50 Ohm primary winding should have a reactance Xl = 50*4 > 200 Ohm @ 100 KHz, hence:

L= (200*10^6)/(2*PI*10^5) > 320 uH
N > 100 * SQRT(320/118,000) = 6 turns, we shall use 7.5 turns.

In a 3 dB/50 Ohm hybrid the turns ratio between primary and each secondary is:
Np/Ns = 1.43
Ns = 8/1.43 = 5.5 turns

Secondaries are made with a 0.2 mm dia. wireline. The wireline is prepared by cutting two 1.5 ft lengths of 0.2 mm dia. enamelled copper wire. The two wire ends on one side are stopped inside a wise and the remainding ends are tightened inside the mandrel of an electrical or hand drill. The drill is slowly rotated in order to get the two wires tightly wound together, obtaining a wireline with 4 terminals.

The transformer is wound with the wireline first and 5.5 turns mean that the wireline ends will come out from the two holes at the same side. Strip the ends, use an ohmeter to sort out which end belongs to which winding. Now wind the primary with 7.5 turns of 0.25 enamelled copper wire, making sure the wire ends come out from the other side of the core.

Prepare the secondary windings by soldering the beginning of one winding to the end of the other winding: this center tap is the terminal for the balancing 27 Ohm resistor. Assemble the transformer as shown in the schematic below.

Assembly procedure, options and measured performance data to come next.

REFERENCES:
- C. L. RUTHROFF, MEMBER, IRE: Some Broad-Band Transformers.
- ANDREI GREBENNIKOV, Infineon/DICE: Power Combiners, Impedance Transformers and Directional Couplers.
Attachments
3 dB Hybrid data.jpg
50 OHm 3 dB Symmetrical Coupler data
3 dB Hybrid data.jpg (62.23 KiB) Viewed 31834 times

Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: 3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

Post by glovisol » Thu Apr 11, 2019 8:43 am

THE 50 OHM SYMMETRICAL 3 dB COUPLER

Looking at the schematic in the previous post, one has the choice of keeping terminals H and K connected and soldered to the common copper surface of a small PC board (as shown in the pic) or of keeping then isolated from one another. In this second case one would use a printed circuit board with separate islands and H would be connected to the antenna ground, while K would be connected to the RSP's ground, thereby providing ground ISOLATION and doing away with an additional isolator.

Uploaded below is the Spectrum Analyser plot of the hybrid performance. Top trace is really two traces one on top of the other. White trace illustrates signal picked up at B while noise generator is connected at A. Yellow trace shows signal picked up at C coming at the same time from A. It can be seen that levels at B and C are IDENTICAL.

Blue trace shows signal delivered at B and picked up at C. Most of the signal is dissipated in the 27 (+7.4% off) Ohm resistor, so the signal at B is attenuated by 18 to 20 dB, demonstrating the balance of the device. Balance and hence isolation, depends on symmetry and precision. With carefully constructed and placed windings and using the right resistor value, which of course should be 25 Ohm exactly, one could reach 30 dB isolation with a little patience. For instance adding a 270 Ohm resistor in parallel with the 27 Ohm, thus bringing the balance resistance value down to 24.3 (-0.03% off) Ohm, increased the isolation to a minimum of 22 dB.

The slope in the plot is due partially to the frequency response of the noise generator and partially to the frequency response of the ferrite core.
Attachments
3 db Hybrid.jpg
Balance & isolation with 27 Ohm resistor
3 db Hybrid.jpg (307.53 KiB) Viewed 31769 times
Hybrid balance 22 dB.jpg
Isolation with 24.3 Ohm resistor
Hybrid balance 22 dB.jpg (178.83 KiB) Viewed 31769 times

Reason: No reason

ON5HB
Posts: 140
Joined: Sat Dec 29, 2018 1:07 pm

Re: 3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

Post by ON5HB » Thu Apr 11, 2019 3:11 pm

What happens when you simply split with T-connectors?

With my websdr I did such, the signal levels did drop a bit, but that is a non issue as the s/n-level-range is enough to compensate.

I have ordered an 8-way splitter but not installed yet.

My question is this: Why do we need insulation between 2 or more RSP receivers?

For a normal receiver I can understand as the local oscillator can be quite strong, but the websdr doesn't have this, at least not in a way that it works the same, nor is it mixed somewhere as far as I know.

As you have a setup, what happens when just splitted with T's, is it bad or nothing happens? How does it compare to a splitter?

Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: 3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

Post by glovisol » Thu Apr 11, 2019 4:45 pm

Hi Bas,

Of course I am only concerned with the SDRplay receivers, the only ones I have worked with up to now. These receivers do have a local oscillator and can be operated in ZERO IF (ZIF) or in LOW IF (ZIF) mode. Some time ago I did a test by using a "T" after the antenna isolation transformer and I could clearly see the interference.

At the end of the day these receivers have very good sensitivity and at HF they can successfully receive SSB signals in the order of -115 / -120 dBm, depending on atmospheric/local noise conditions, so one must not be surprised by Local Oscillator leaks and one self built isolator, costing a few cents worth of components, pushes the limit by another 20 dB with casual construction. If you really have a go at building it properly, a two way, 3 dB hybrid coupler will give you 30 dB isolation.

I am quite skeptical about splitters more than n=2 in HF, unless you use cascaded hybrids like the one I have described and can accept a loss (above the power split) of 0.5 dB per run. For instance a 4 way splitter has 2 runs and each port loses 1 dB above the power split, so the signal level at every of the 4 receivers will be -7dB as compared to the antenna output.

Finally do not forget this thing must be shielded in a proper box, even if I did not show the final completed unit as these die cast boxes take ages to arrive....

Cheers,

glovisol

Reason: No reason

ON5HB
Posts: 140
Joined: Sat Dec 29, 2018 1:07 pm

Re: 3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

Post by ON5HB » Thu Apr 11, 2019 5:14 pm

SDR receivers like the SDRplay do not have a local-oscillator.
They do have an oscillator but it has no mixing functions or other stuff.
It's just a clock to make the chips tick, there is no RF to mix it with or such.

The oscillator has 2 purposes, that is a reference and a steady clock for the chips.
It's impossible that it goes back into the input and hinders other boxes as you state, unless the areal receives it, but then a splitter would receive the same.

Have a look....it's almost impossible to get back into the input.

See the attached shematics.
Attachments
SDRplay_RSP_Schematics_r2p2.pdf.zip
(75.24 KiB) Downloaded 1215 times

Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: 3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

Post by glovisol » Thu Apr 11, 2019 6:25 pm


Reason: No reason

sdrom33
Posts: 128
Joined: Sun Dec 09, 2018 4:38 pm

Re: 3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

Post by sdrom33 » Thu Apr 11, 2019 6:48 pm

ON5HB in the schematic you posted I see 3 mixers the circles with the cross in, look at U5.

Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: 3 dB HYBRID COUPLER/ISOLATOR FOR ANTENNA SPLITTING 100 KHz -10 MHz

Post by glovisol » Tue Apr 16, 2019 7:41 am

THE HIGH IMPEDANCE ASYMMETRICAL 3 dB HYBRID COUPLER

The prevously discussed 3 dB hybrid, that works with equal 50 Ohm coaxial impedances on all ports, is useful for joining a 50 Ohm coaxial antenna lead to two 50 Ohm coaxial receiver inputs. Long wire or Beverage antennas with termination impedances in the order of 560 Ohm would require a 560 to 50 Ohm transformer. Additionally, the very useful balanced input provided by some RSP receivers (e.g. RSPduo or RSP2) would require another extra transformer 50 Ohm to 1 KOhm.

By judicious choice of hybrid impedances and by addition of transformers as integral parts of the coupler, it is possible to accommodate any input and any output impedance at the three ports, keeping total losses in the 1 to 2 dB range, which are negligible, at LF, MW and low HF bands.

The example below shows a 3 dB coupler with a 560 Ohm balanced input for accommodating a long wire antenna, one 50 Ohm coaxial output and one 1000 Ohm balanced output + common, ideal for the HI Z input of the so equipped RSP receivers. Such a coupler is useful for comparing two inputs of the same receiver, or for comparing two different receivers with different input impedances, while at the same time eliminating the danger of crossed receiver radiation altering the measurements. Note that amplitude variations with frequency, shown in the screens below, depend only on the Noise Generator output characteristics, as demonstrated by the top trace, obtained by connecting directly the attenuated Noise Generator output to the RSP-1A input.

Additional information in the next posts.
Attachments
HI Z Hybrid 1.jpg
HI Z Hybrid 1.jpg (321.81 KiB) Viewed 31436 times
HI Z Hybrid 2.jpg
HI Z Hybrid 2.jpg (352.94 KiB) Viewed 31436 times
HI Z Hybrid Pict..JPG
HI Z Hybrid Pict..JPG (178.94 KiB) Viewed 31436 times

Reason: No reason

Post Reply