Making scaler ring for cantenna on dish – conversation with ChatGPT 2/5/2025
I came across pictures of scaler rings for feedhorns and had a conversation with ChatGPT on whether such a ring would improve the performance of my 150cm dish made from solar cooker reflector and cantenna.
Andy
Can you design for me a scaler ring to put on end of cantenna I am currently using as feed for my 150cm dish for 1420MHz? My cantenna is 15cm diameter and cylindrical made from large coffee can.
Yes, I can help you design a scaler ring for your 15 cm diameter cantenna at 1420 MHz (21 cm wavelength). A scaler ring improves impedance matching, reduces diffraction effects at the waveguide opening, and enhances illumination of your 150 cm dish.
Key Design Considerations
- Cantenna diameter: 15 cm (0.71λ at 1420 MHz)
- Dish f/D ratio: Please confirm, but assuming it’s about 0.4–0.45 for a typical 150 cm dish.
- Wavelength: λ = 21.1 cm
- Number of rings: 4–6 is typical
- Ring spacing and depth: Usually about 0.25λ (≈5.3 cm)
Scaler Ring Design (Basic Circular Groove Model)
1. Ring Parameters
- Outer diameter of ring array: ~30–35 cm (1.5–1.7λ) to match illumination pattern to dish edge
- Number of rings: 5
- Ring spacing (center-to-center): 5.3 cm
- Ring depth (groove depth): 5.3 cm
- Ring width (groove width): 0.5–1.0 cm
- Material: Aluminum or copper is ideal
Visual Layout
- The rings are radial circular grooves cut concentrically into a metal plate (or rings soldered/welded in).
- The cantenna is mounted flush to the center, protruding slightly if needed.
- This structure sits at the dish focus, with rings facing the dish.
Here’s a set of radii from the center for five scaler rings spaced at 0.25λ intervals (i.e., 5.3 cm), starting just outside the edge of your 15 cm diameter (7.5 cm radius) cantenna:
Scaler Ring Radii (from center):
Assuming the first ring starts just beyond the can edge:
- Ring 1: 8.0 cm
- Ring 2: 13.3 cm
- Ring 3: 18.6 cm
- Ring 4: 23.9 cm
- Ring 5: 29.2 cm
Each ring is spaced ~5.3 cm from the previous, matching 0.25λ at 1420 MHz.
These radii define the centerline of each groove if you’re cutting circular grooves into a flat metal plate. If instead you’re forming physical metal rings, you can make each ring with a small width (e.g., 1 cm) centered on these radii.
Would you like a diagram or CAD-style sketch of this layout?
What vertical height do the rings need to have?
Great question. The vertical height (or depth, if you’re cutting grooves) of the scaler rings primarily affects their ability to suppress diffraction and match the waveguide to free space. For 1420 MHz (λ = 21.1 cm), the ideal height is typically:
Scaler Ring Height:
- Recommended height: ~λ/4 = 5.3 cm
This height gives the rings a quarter-wave depth, which is effective at creating the right impedance transition and suppressing sidelobes.
Summary for Your Scaler Ring:
- Number of rings: 5
- Radial positions from center: 8.0 cm, 13.3 cm, 18.6 cm, 23.9 cm, 29.2 cm
- Vertical height of each ring: 5.3 cm
- Ring width: ~0.5–1.0 cm (not critical, just mechanically sound)
This configuration works well with your 15 cm cantenna and 150 cm dish.
Would you like a printable drawing or 3D model format for fabrication?
What difference to my reception of 1420MHz signal will adding a scaler ring to my cantenna make?
Adding a scaler ring to your 1420 MHz cantenna feed can make a significant difference to your reception by improving how well the feed illuminates your 150 cm parabolic dish and reducing unwanted signal artifacts. Here’s how:
1. Better Illumination of the Dish
Without a scaler ring, your cantenna tends to over-illuminate the dish (spills signal beyond the dish edge), which:
- Wastes gain
- Increases ground pickup and thermal noise
With scaler rings:
- The feed pattern is narrowed and better matched to your dish’s geometry (especially important for f/D ~0.4–0.45)
- This means more of the dish area contributes usefully, boosting effective gain
2. Reduced Sidelobes and Spillover
The rings act as a choke, suppressing:
- Sidelobes (which pick up interference and sky noise)
- Spillover (which picks up warm ground noise)
This results in:
- Improved signal-to-noise ratio (SNR)
- Better rejection of terrestrial interference
3. Impedance Matching and Mode Control
At the can’s aperture, there’s a sudden transition between waveguide and free space. The scaler ring:
- Smooths this transition, improving impedance match
- Helps confine energy in the dominant TE₁₁ mode of the cylindrical can
- Reduces reflection and standing waves back into the can
4. Net Effect on Performance
You can expect:
- Gain increase: up to 1–2 dB
- System temperature drop: due to reduced ground spillover
- Sharper beam pattern: better directionality and angular resolution
In practice, for a 1420 MHz radio astronomy setup:
- Fainter sources become detectable
- HI line spectra become more stable and less contaminated
- Overall sensitivity and data quality improve
Here’s how you can test and quantify the improvement after adding the scaler ring to your 1420 MHz cantenna feed, using practical and astronomy-friendly methods.
1. Measure System Noise Temperature (Tₛᵧₛ)
Method:
- Point dish at cold sky (zenith, away from Milky Way)
- Then point dish at warm ground or sun
- Record receiver output power for both
What to look for:
- Improved Y-factor (ratio of hot/cold power levels) after adding scaler ring
- Estimate Tₛᵧₛ using:
Tsys=Thot−Y⋅TcoldY−1T_{\text{sys}} = \frac{T_{\text{hot}} – Y \cdot T_{\text{cold}}}{Y – 1}
where Y = P_hot / P_cold
- Expected result: Lower Tₛᵧₛ after adding the ring, due to reduced ground spillover
2. Measure Beam Pattern (optional but valuable)
Method:
- Mount a 1420 MHz signal source (e.g., small Yagi or another cantenna) on a nearby mast or drone
- Sweep the feed pattern across angles and record signal strength
What to look for:
- Main lobe should become narrower
- Sidelobes should be reduced
- Beam should better match dish geometry
3. Observe Astronomical Source (e.g., the Sun or Milky Way plane)
Method:
- Log signal level (in dB or ADC counts) while tracking
- Use consistent integration time and bandwidth
What to look for:
- Increased peak signal level from same source
- Lower baseline noise when pointing at cold sky
- More stable or cleaner HI spectral line profile (less ripple)
4. Spectral Line Observation (HI line test)
Method:
- Record HI spectrum from known target (e.g., galactic plane at known coordinates)
- Integrate over 5–15 minutes
What to look for:
- Cleaner baseline, reduced ripple from reflections
- Possibly stronger line signal due to improved gain
Recommended Tools:
- SDR software with dB output (e.g., SDR# with calibration, or GnuRadio, or custom Python scripts)
- Spectrum averaging and logging
- Simple Python or spreadsheet calculator for Y-factor and Tₛᵧₛ
Python script or spreadsheet template to help calculate Tₛᵧₛ and compare before/after results:
Here is a simple Python script to calculate system temperature (Tₛᵧₛ) from hot and cold measurements using the Y-factor method. You can run it in a Python environment like Jupyter, or adapt it for a spreadsheet.
Python Script: System Temperature Calculator
def calculate_tsys(p_hot_db, p_cold_db, t_hot=300, t_cold=10):
“””
Calculates system noise temperature (Tsys) from Y-factor.
Parameters:
– p_hot_db: power when pointing at hot source (dB)
– p_cold_db: power when pointing at cold sky (dB)
– t_hot: estimated hot load temperature in K (default = 300 K for ground)
– t_cold: estimated cold load temperature in K (default = 10 K for sky at 1.4 GHz)
Returns:
– Y factor
– T_sys in Kelvin
“””
import math
# Convert dB to linear power
p_hot = 10 ** (p_hot_db / 10)
p_cold = 10 ** (p_cold_db / 10)
# Y-factor
Y = p_hot / p_cold
# T_sys calculation
T_sys = (t_hot – Y * t_cold) / (Y – 1)
return Y, T_sys
# Example usage:
# Measured power levels in dB from SDR software or spectrum analyzer
p_hot_db = -30.0 # Example: pointing at ground
p_cold_db = -35.0 # Example: pointing at zenith sky
Y, T_sys = calculate_tsys(p_hot_db, p_cold_db)
print(f”Y-Factor: {Y:.2f}”)
print(f”Estimated System Temperature: {T_sys:.1f} K”)
Spreadsheet Template (Manual)
Parameter | Value |
Hot Power (dB) | -30.0 |
Cold Power (dB) | -35.0 |
T_hot (K) | 300 |
T_cold (K) | 10 |
Y = P_hot / P_cold | =10^((A2-A3)/10) |
T_sys = (T_hot – Y*T_cold) / (Y – 1) | Use cell formula |
Click on link below to download Excel spreadsheet to provide a plot of Tₛᵧₛ vs Y-factor created by ChatGPT 2/5/2025:
Andrew,
I know this has been posted many times but
https://www.w1ghz.org/antbook/chap6-3.pdf
This has a lot of good stuff!!
I run kumar choke feeds on my 3m dishes and they do lower difference for ground noise pickup.
So you will see multiple positions for the extension of the choke ring.
1) flush
2) cantenna extended
3) cantenna retracted
I run number 3, cantenna retracted. When i would go out and adjust the declination , the antenna would see me walk back of the dish. So I brought the choke ring out a couple inches towards the dish and that has helped a lot. i am sure there will be argument
The cantenna is 6 HVAC pipe and the choke ring 4″ deep and 14″ in diameter (so 4″ on a side)
73
mike w9ys
The question for smaller dishes is whether the aperture blockage implied by a choke ring is “worth it” in terms of overall
performance improvement. I think 1.5m diameter was mentioned. Not huge.
Marcus