Does the mesh from the two halves of hydrogen line parabolic dish need to be in good electrical contact, answer from ChatGPT and responses to the ChatGPT response by SARA members.

Does the mesh from the two halves of hydrogen line parabolic dish need to be in good electrical contact, answer from ChatGPT and responses to the ChatGPT response by SARA members.

It is important to read to end of this article.

ChatGPT RESPONSE:

Yes, the mesh from the two halves of your parabolic dish must be in good electrical contact—especially for work at higher frequencies.

Here is why and what it means in practice:

---

🔧 Why Electrical Contact Matters:

1. RF Reflectivity

At radio frequencies, a mesh reflector behaves like a conducting surface if the mesh spacing is much smaller than the wavelength. However, discontinuities (like a gap or poor electrical contact between the two halves) can:

• Introduce scattering
• Cause diffraction
• Reduce gain and increase sidelobes
• Create standing waves and unpredictable reflections

2. Higher Frequencies Are Less Forgiving

At 1420 MHz (21 cm), small gaps may be tolerable if the mesh size and gaps are much smaller than λ/10 (~2 cm).
But if you later use this dish at e.g., 5.8 GHz (5.2 cm λ) or 10 GHz (3 cm λ), then:

• The mesh must be finer (apertures < λ/10 → <5 mm)
• Any non-contact between the two halves becomes a serious issue

---

🛠️ How to Ensure Good Contact

⚙️ Mechanical Solutions

• Use multiple conductive straps or braided wire jumpers between the two halves—placed around the perimeter and near the focus—secured with bolts or spring clamps
• Interleaved metal tabs or overlapping flanges, bolted together with stainless steel hardware
• Conductive metal lugs with star washers to bite through oxide layers

🧪 Material Tips

• Clean contact surfaces to remove oxidation (especially if using aluminium)
• Use conductive anti-corrosion paste (like Noalox or Ox-Gard) on joints
• Avoid relying solely on physical pressure without conductive bonding

---

✅ Summary

Requirement	Recommendation
Electrical continuity	Yes, essential—especially at higher frequencies
Connection type	Bolted/spring-loaded with multiple low-resistance paths
Mesh gap between halves	Avoid gaps >λ/10 (e.g., >2 cm at 1420 MHz; >5 mm at 6 GHz)
Long-term durability	Consider oxidation, moisture, and thermal expansion

COMMENTS FROM SARA MEMBERS TO CHATGPT RESPONSE:

• It is true that at very high frequencies the discontinuities in the surface will represent more than 1/10 lambda. None of the commercial segmented dishes I have ever used have had bonding between segments. Our own 12.8m dish has gaps between segments and no bonding straps. Reflection requires that the surface be conductive, but no requirement for good conductivity between segments. 
• As usual with the more simple-minded free LLMs that seem to use Reddit or Facebook rants as source material, it is a bit off-target. My 3m dish has 8 petals, with a metallic mesh embedded in GRP. Same with my 2.4m 4-petal dish. All the panels are insulated from each other. there is no measurable impact on side lobes and no discernible pattern distortion, at least below the rated frequencies of the dishes. In theory, there will be small spikes and dips aligned with the gaps in the dish caused by diffraction effects because of the phase difference on either side of the gaps. The biggest impact is that usually when you are creating a panelised dish, it is very difficult to maintain the surface figure accurately and to ensure that the joints are perfectly aligned. I have seen a mesh dish in two sections which had a shaped aluminium plate either side of the seam line, with shoulder bolts to attach the two halves with very precise alignment. Another example of how the performance is Little affected is booked one of the extreme distance contacts from the south coast of England to I think New Zealand via the moon at 10 GHz had a cut up metal dish at one end of the contact, so it could be fitted in aircraft luggage. Making an earth moon earth contact with a small dish is already hard enough so the additional detraction loss and pattern distortion that might have been caused by chopping up the dish obviously did not have much of an impact. Certainly, mesh size, surface figure and smoothness matter a lot, and it helps if the mesh is welded unless the polarisation is along the axis of the mesh wires. The VK3UM EME calculator has mesh size simulation to determine the leakage. 0.1 wavelength mesh or smaller is best, as ChatGPT rightly says. interestingly, you can get excellent performance from what would appear to be a terrible mash material. Welded square galvanised mild steel mesh or stainless-steel mesh wire work well. Not as good as 20 µm of silver plating on a solid substrate, but with far less wind resistance, weight, and bank account impact. It might be instructive to compare the ChatGPT output from Google Gemini, as it quotes its sources and shows its reasoning rather than just wildly making stuff up that seems to be from an electrical engineering or vehicle electrics source rather than practical for microwave use. OpenAI o3 is better still at avoiding hallucination and slop generation. Gemini will write you scientific report based on its findings. On the downside it usually sounds like something written by a high school student rather than a rigorously argued scientific journal submission. Conductivity matters at local scales of a few wavelengths, but the difference between a dish made of petals versus a solid at microwave frequencies is insignificant. A big HOWEVER though… If the mesh is insulated, it is a good plan to ensure DC grounding to avoid static buildup that might zap an LNA. I find it very instructive to run EM solver models of dishes and horns in OpenEMS. Free, limited only by computer power and very flexible. It really helps get an intuitive understanding of which parameters are important and which are not
• Really pleased with this track. Thanks for correcting some of my misconceptions. Curious about the info re horns and dishes as to the need for well-bonded joints
• For lightning protection, I guess good bonding is important in some places. LLMs may be conflating this idea. 
• That I do know. Peak currents as high as 30000amps means that a very small resistance will result in large currents, and ground loops will be important. Sort of the opposite end of the rf spectrum :-). There are other conditions where bonding is very important, especially at dissimilar metal joints. There have been some discussions on these in the past.