swr meter design 2.4 ghz
Hi Friends!
There is my translation of this article. Sorry friends, maybe there are a lot of grammatical errors in translation, but I will explain any question about this construction. All language and terminology corrections are welcome.
----------------------------------------
Coaxial Directional Coupler.
Y. Kurinyi (UA9ACZ), V. Pilskii.
There are a lot of directional coupler circuits, but all of them have big disadvantage in relatively narrow frequency band. It compels to use adjustment of sensivety, for example in SWR meters.
This SWR meter allows to measure SWR in wide frequency band (1 to 500 MHz), to measure power in cable (without dependency from SWR level), to explore HF modules for unmatches in connectors and coaxials (with precision up to fractions of percent), to find with big precision positions of shortage or break in cable and other elements of HF lines, to use in half-duplex systems, and so on.
This Directional Coupler (DC) consists of current and voltage sensors and summator. The simplified circuit of the DC is shown on Fig.1, where "+-Ic" means current in cable (sign depends of wave direction), "p" - impedance of a cable, R1 R2 - resistors at voltage sensor, R3 - resistor in current sensor. If R1>>R2=p>>R3 , then mathematical description of this circuit is greatly simplified. In result, we have V(tap)=(V(cable)+-V(cable)/(2*K), where K=p/R3=R1/p - coefficient of attenuation to tap output. Thus, V(tap)=V(cable)/K for direct wave, and V(tap)=0 for reflected wave.
Wide working frequency band of this DC is achived by original construction. Current sensor is realized as one-turn current transformer with central wire of cable as primary coil and special cavity on shield together with cable sleeving as secondary coil.
...
The secondary coil have load made of resistors which are uniformly located on perimeter of sleeving gap. They are named as R3 on simplified circuit diagram. The parasitic inductance of this shunt is much lower then inductance of a cableon this construction. We need to have R3=<(w*Ls) on lowest working frequency, where w - lowest frequency, Ls - inductance of secondary coil. We use a number of ferrite toroids weared to cable to increase this inductance.
Voltage sensor consists of resistive divider R1R2, it is connected between central wire of a cable and output of current sensor. Divider R1R2 is connected in parallel with a cable, current sensor R3 in series with a cable.
...
The practical construction of SWR and Power meter is made as two back-to-back connected couplers (look at color picture #2). Shield, that is used as a box at the same time, is made of textolite. The sizes of this shield/box is free. This construction was designed for 50 Ohm cable with maximum power near 200 Watts with SWR<=4. The maximum allowed power is increased for little SWR and decreased for big SWR. The coupler attenuation is chosen as K=100.
R1 consists of 4 resistors in serial two 1.2K and two 1.3K (all are on 1W). R3 is made of 15 resistors 7.5 Ohm (0.5 W), they are located uniformly on gap of cable sleeve (look at color picture #4). R2 consists of two 100 Ohm resistors (0.25 W) in parallel, they are located above current sensor resistors on distance 1.5 - 2 mm between these two resistors. All resistors are needed to be without parasitic inductance in their construction. Distance from R1 to shield needs to be more than 15-20 mm. Torroids are made of low frequency ferrite core. You need to fix tightly cable sleeve to dielectric on whole lenght, so use wire or thread winded above it.
This construction becomes simple if you use cable connectors, but you can use panel connectors. Look at b/w picture #2, (a) is for 30 MHz upper frequency, (b) and (c) for 500 MHz. Cone in third variant is made with cable sleeve or from tin. The biggest circle of this cone has no critical diameter and it depends of used connector size. You need to solder it accuratly on all perimeter.
If you take all resistors with 1% precision then you have no need to adjust it. In other way you need to correct one of current resistors for minimal signal of reflected wave. (Anyway all this adjustment is needed only for SWR precise measuring less than 1.05).
Our first construction was made without component selection or any adjustment with these parameters:
Frequency Band = 0.3 - 500 MHz,
K = 100 +- 5
There was worsening 2% of directional coefficient on 30 MHz, and 5% on 500 MHz.
There were taken 30 ferrite toroids with 20x10x6 size. As amateur band has lower border on 1.8 MHz, the number of toroids could be 6-7.
You can use double-beam oscilloscope to find inhomogeneity in cable, with sending square pulse you can find time delay between direct and reflected pulse, as wave speed in cable is near 2*10^8, you can calculate distanse to inhomogeneity.
If you are planning to use this SWR meter only on VHF (where powers are not so big), then you can extend frequency band to high. For this, take K=10-20, R3 (1/8 W) in one line, change R1 to one resistor (1/2W) and make some frequency compensation to avoid parasitic capacitence of R1. In this case you can extend upper frequency to 1 - 1.5 GHz.
-------------------------------------------
Best Wishes! klug.