Chapter 246: RADAR modèle 37 "Galène" - III
Seven days passed.
Gone were the initial sketches, loose theories, and speculative notes.
In their place sat actual components vacuum tubes, wire spools, metal chassis mounts, power resistors wrapped in brown wax paper, and spools of military telephone cable.
The table had been moved to the center of the room to accommodate a long, cloth-covered crate.
Moreau entered at 08:02, removed his gloves, and took his place without a word.
A single nod was all he gave.
The meeting began.
Langevin spoke first. "We've prototyped the antenna's lobe pattern using both four and five elements. The four-element setup gives a 20-degree main lobe, but side-lobe suppression is weak. You'll get spurious detections on the periphery. I recommend we switch to five. That narrows the beam to fifteen degrees and cuts side lobes below minus 20 decibels. Better discrimination, cleaner echoes."
Moreau asked, "Any change in gain?"
"Gain increases by about 2 dB. Estimated total now 11.5, perhaps 12. Field testing will confirm."
"Mounting effects?"
"None. We can retrofit it onto the mast design without structural changes."
Chrétien added, "We'll need to recalibrate guy wire tension to offset the added wind load. Higher gain comes at the cost of stability."
"Do it."
Ponte stepped forward next and unrolled oscilloscope paper from a protective tube. "Power modulation held at 120 kilowatts peak for just under nine minutes. At ten, the triode's core temp began to exceed safe operating margins. I adjusted pulse width from 10 microseconds to 9.2 under load. Power output dropped slightly, but system stability improved. If we maintain this profile, tube burnout risk falls by 30 percent."
Moreau leaned forward. "Cooling solution?"
"Closed-loop water cooling is working. Auxiliary radiator performs to spec. But we'll need forced airflow over the pump valves. I'm adding two low-RPM fans to the assembly."
Moreau nodded. "Is the radiator field-replaceable?"
"Yes. Bolted housing, no soldered joints."
"Add a spare to the Dieppe shipment."
Auger, who had remained quiet until then, stepped in front of the schematic pinned to the board. "We tested PRF staggering using a two-tier timing oscillator. Staggered PRF breaks range ambiguity. It works. But... it introduces timing drift. At 0.3 percent over 60 seconds, we lose sync unless the master oscillator is crystal-controlled."
Moreau asked, "Can it be tied to the transmitter directly?"
"Already done. Clock reference from the transmitter is now fed directly into the pulse-forming network. Drift reduced to under 0.2 percent. Reliable enough for manual tracking."
"And side lobe ghosting?"
"We haven't filtered it yet. If the side lobe return is strong enough, it still appears on the A-scope. We can taper the gain on faint echoes, but we'll need to teach operators how to discriminate."
"You're saying the human eye has to decide what's real."
"For now, yes."
Langevin interjected. "It's manageable. The scope traces fade faster on side-lobe echoes. If operators compare sequential sweeps and see consistent location and strength, it's a real target."
"Add that to the operator training manual," Moreau said.
Perrin stood last. "I have full commitment from CSF. Two engineers assigned under military contract. Their names are on the sheet sealed. Both have prior experience with radio valve design and industrial calibration. They'll be redirected to Dieppe under a cover story."
"What cover?"
"Coastal radio-telephony interference measurement. CNRS has submitted a dummy survey to support it."
"Security clearance?"
"Processed. They've signed internal secrecy protocols. If breached, prosecution under Article 410."
"Good."
Moreau looked around the room. "We're close. What's still missing?"
Langevin glanced at Auger.
Then spoke. "Elevation. We still can't resolve vertical angle. Azimuth and range we've handled. But we can't determine height."
Auger stepped forward. "Two solutions. First, we mount a second Yagi, fixed at a steeper upward angle say 25 degrees. If both beams get returns, compare echo strength. Crude height approximation."
Chrétien added, "Or we install vertically stacked elements on the mast and mechanically switch them via a timed relay. Two positions horizontal and elevated. Alternating pulses. Stronger return suggests elevation."
Ponte frowned. "But switching will interrupt PRF. We'll halve detection rate."
"Better than guessing altitude," Moreau said. "Proceed with the fixed dual-beam solution first. Evaluate results before adding switching logic."
They nodded.
Moreau asked, "What's our minimum detectable range?"
Auger replied, "Around 2 kilometers. We've built in a pulse blanking circuit suppresses receiver input during transmit phase. Avoids self-interference. But targets inside 1.5 kilometers remain unreliable."
"Acceptable for an early warning device."
Chrétien shifted a crate off the table and set a smaller chassis down.
It had a seven-inch cathode ray tube embedded in the front, shielded by a darkened hood.
"This is the revised A-scope. We've added a polarizing filter and contrast dial. Scope readability under daylight improves by 40 percent. Also includes a range cursor dial and headphone jack for audio cueing."
Moreau examined it closely. "Will these be field-serviceable?"
"Tube sockets are modular. Any soldier with basic electronics training can replace them in minutes. We'll provide spares in the equipment pack."
Perrin opened another folder. "Site prep is under way at Dieppe. Antenna field is cleared. Concrete mixing starts Wednesday. Control cabin will be finished in fourteen days. Generator shed by end of month."
Moreau asked, "Power routing?"
"Buried cable, armoured. Generator to control cabin: fifty meters, triple-phase, with voltage stabilizer. Surge protection inline. Backup petrol generator under review."
"Comm relay?"
"Dedicated landline to local command post. Military-grade telegraph switchboard. Wireless backup will be Morse-only."
"Add secure call signs."
"Already done."
Langevin interjected. "Antenna mast construction team has requested final elevation specs."
"28 meters. Fixed. Secondary beam mounted at 25 degrees, three meters below primary. Both duralumin booms, copper feedlines. Rotation via belt-driven gearbox. Variable speed allowed, default to 6 rpm."
Chrétien confirmed. "Guy wire anchors are reinforced concrete. Should hold under Beaufort scale 6 wind."
Moreau turned to Auger. "Tube consumption?"
"We'll go through one power tube every 20–25 hours of continuous duty. Order triple redundancy. I suggest we keep six per station in reserve."
"Procurement source?"
"Philips or Thomson. Fallback, Italian contractors via neutral re-exporters."
"Document it."
Ponte raised a final concern. "Operator training. Scope reading under stress is still an unknown. During tests, some of the signal blips were misread or ignored."
"We'll design a curriculum," Langevin offered. "One week theoretical. One week practical with test target flyovers. Simulated echoes included."
"Also," Chrétien added, "blip width calibration must be tied to pulse width. If they change PRF or pulse width mid-shift, it throws off the distance scale."
"Lock it," Moreau said. "Standardize pulse width. No adjustments allowed without officer authorization."
Perrin checked his sheet. "Are we sticking to the July 25th test date?"
"Yes. Calibration begins August 1st. System must be detection-ready by August 10th."
"Operators?"
"Military radiomen. Eight, rotating shifts. One commanding officer. Two electrical engineers. You'll oversee their instruction."
Langevin looked around. "We are nearly there. But you understand this isn't just a prototype now."
Moreau looked at each man.
"I do."
He paused.
"This is a national capability. Once we emit, others will hear. This frequency range is unencrypted. If Berlin has direction-finding posts along the coast, they will log it. So we activate once. Once only. With purpose. And only when we are ready."
Silence followed.
Perrin broke it. "Our best chance is that they hear it and realize we're not blind."
"That is the plan."