Furuno MD-7918 Modulator — Field Reference & Measurement Plan

For: RSB-0034 scanner (MODEL 1830), MD-7918 modulator, IF-7758 IF amp Symptom: receiver noise present, no main bang, no echoes, empty tuning bar, after a soft grounding

0. Source of this data and how much to trust it

There is no public MODEL 1830 service manual. Everything below is taken from the Furuno MODEL 1832/1932/1942 service manual (97 pages, includes full schematics), which documents MD-7918A (MODEL 1932) and MD-7918E (MODEL 1942) — the same board family, same designators (TP801/802/803, VR801/802, JP801), same FET-switching modulator topology as your MD-7918.

Confidence key used throughout:

[CONFIRMED] — read cleanly out of the manual text.
[OCR] — reconstructed from a rotated table that OCR'd badly. Directionally right, treat the exact digits as approximate.
[INFERRED] — my reasoning from the topology, not stated in the manual.

⚠️ Your 1830 is an earlier model than the 1832/1932/1942. Expect the topology and designators to match and the exact numbers to be close but not guaranteed. Verify against the board silkscreen and your own schematic where you can.

1. First, the mystery part: FCX73 is your circulator

From the 1832 service manual, scanner unit parts location (p. 3-10):

Inside of Scanner Unit — IF AMP Board (IF-9214), Magnetron V801 (E3571), MIC (RU-9360), Circulator HY801 (FCX73) [CONFIRMED]

So FCX73 = the ferrite circulator, reference designator HY801. That explains exactly what you described:

Three ports, numbered 1-2-3 on the body.
The arrow shows the direction of circulation (1 → 2 → 3 → 1).

Typical port assignment [INFERRED]: port 1 = magnetron, port 2 = antenna/radiator, port 3 = MIC (receiver front end).

Why this matters for you: a circulator is a permanent-magnet + ferrite device, and it is one of the few RF parts in the radome that a mechanical shock can genuinely damage — cracked ferrite, cracked or shifted magnet, disturbed alignment. A degraded circulator kills transmit into the antenna and receive out of it, while leaving the IF amplifier happily generating noise. That fits your screen exactly.

It is not, however, my first suspect — see §6.

2. How the MD-7918 works

From p. 2-16, "Principle of FET switching modulator" [CONFIRMED]:

High voltage is charged into C through R while the magnetron is inactive. When the trigger is applied to the power MOS-FET, the FET turns on and the high voltage appears at the primary winding of the pulse transformer.

Chain, end to end:

Display (SPU board)
   │  TRIGGER  (positive pulse, ~8–12 V)
   │  PL/A, PL/B  (2-bit pulse-length select)
   │  TUNE  (0–12 V analogue)
   │  ANT +12 V / ANT -12 V
   ▼
MD-7918 ──► DC-DC converter ──► HV rail ~300–370 Vdc ──► charge C through R
   │                                                          │
   │  trigger ──► gate driver ──► power MOSFET ───────────────┘
   │                                    │
   │                            pulse transformer primary (~300 V pulse)
   │                                    │
   │                            secondary (several kV negative)
   │                                    ▼
   │                            MAGNETRON cathode  (V801)
   │  heater supply 7.4–7.6 Vdc ──────► magnetron heater
   ▼
Magnetron ──► CIRCULATOR HY801 (FCX73) port 1
                  ├─ port 2 ──► radiator (antenna)
                  └─ port 3 ──► limiter ──► MIC ──► IF-7758 ──► VIDEO ──► display

3. MD-7918 test points

Board features: TP801, TP802, TP803, VR801, VR802, JP801, connectors J811, J812 (and J701 on the bearing/signal board next to it). [CONFIRMED]

Point	Function	Expected	Confidence
TP803	GND — reference for everything else	0 V	[CONFIRMED]
TP801	TRIGGER input, same signal as the SPU/display trigger output	positive pulse, ~8–12 V, at the PRF	[CONFIRMED function], [OCR amplitude]
TP802	Magnetron current monitor	ST-BY: ~0 V. TX: roughly 0.2 – 1.2 Vdc, varying with pulse length (short / medium / long)	[OCR]
VR801	Magnetron heater voltage adjust	set heater to 7.4–7.6 Vdc at ST-BY	[CONFIRMED]
VR802	Second adjustment, almost certainly magnetron current / TX HV	—	[INFERRED]
JP801	Configuration jumper (model / pulse-length option)	—	[INFERRED]

The manual also lists a 6-way connector on the board [OCR]:

Pin	Signal
1	TX(-)
2	GND
3	+14 V
4	Mag. (heater)
5	Mag. (heater)
6	GND

This is very likely the unpopulated test header you found. Pins 4 and 5 are where you measure heater voltage.

4. Manufacturer's target values

TX high voltage, measured at ST-BY [CONFIRMED]

Model	Spec
1832	300 – 370 Vdc
1932	300 – 370 Vdc
1942	350 – 390 Vdc

Out of range → faulty modulator PCB (MD-9208 on 1832, MD-7918A on 1932, MD-7918E on 1942).

Your 1830 is a 4 kW radome like the 1832/1932, so expect ~300–370 Vdc. [INFERRED]

Magnetron heater voltage, measured at ST-BY [CONFIRMED]

Model	Spec	Adjust with
1832	7.4 – 7.6 Vdc	R106 on PTU-9335
1932	7.4 – 7.6 Vdc	VR801 on MD-7918
1942	7.5 – 7.7 Vdc	VR801 on MD-7918

Measure at ST-BY, not TX — the heater voltage is deliberately reduced during transmit.

Signals from the display's DJ-1 connector [OCR — pin numbers cross-checked against flow charts]

All measured positive lead on the listed pin, negative lead on DJ-1 #20 (GND).

Signal	DJ-1 pin	Expected
TRIGGER	#2	Positive polarity, 8 – 12 V
TUNE	#6	0 → 13 V on 1832 class; sweeps when switching ST-BY → TX during auto-tune search
PL/A	#8	Logic level: 0–1.0 V = L, 7–12 V = H
PL/B	#7	Logic level: 0–1.0 V = L, 8–12 V = H

PL/A and PL/B together form the 2-bit pulse-length code (short / medium / long).

PRF reads as roughly 2000 – 2300 Hz on short pulse [OCR]. The trigger pulse width figure did not OCR reliably — don't chase a number, just confirm a clean pulse train at the right rate.

Line voltages in the display [OCR, approximate]: +12 V rail ≈ 12.1–13.5 V, +5 V ≈ 4.9 V, −12 V ≈ −11.6 to −12.8 V, ANT ±12 V ≈ 12.0 V.

⚠️ DJ-1 pin numbering is from the 1832-generation display. Your 1830 display may differ — confirm against the 1830 installation/interconnection diagram before trusting the pin numbers. The signal names and levels will be right regardless.

5. Safety before you probe

The HV rail is 300–370 V and sits on a reservoir capacitor. Discharge it and verify 0 V before touching the board.
The pulse transformer secondary reaches several kilovolts negative. Do not probe it with a standard 10:1 probe.
Do not stand in front of the radiator while testing for transmission.
Do not run the set in TX with the magnetron removed.
The magnetron contains beryllium oxide ceramic — do not break, grind, or file it.
Remove your watch and use a non-magnetic screwdriver near the magnetron (strong magnetic field).

6. Measurement plan, in order

Do these in sequence and stop at the first failure.

Step 0 — repair the bad negative first

Then verify it under load: with the set transmitting, measure between the display's negative terminal and the radome chassis. > 0.2 V means you still have a return-path problem and every other measurement is suspect.

Step 1 — supply current, STBY vs TX

You haven't done this yet and it's a 30-second answer.

Note DC input current in ST-BY, then in TX.
Same current in both = the modulator is not pulsing. That immediately points at trigger / FET / HV.
Higher in TX = something is firing; move on to find out what.

Step 2 — ±12 V arriving at the radome, under load

Measure ANT +12 V and ANT −12 V at the radome connector, in TX, while the antenna is turning. A sag here (especially with your ground history) explains a dead DC-DC converter output.

Step 3 — TRIGGER, with the scope

Setup: DC coupled, 5 V/div, 200 µs/div, trigger on rising edge, ground clip to TP803.

First at the display end, DJ-1 #2 to #20. Expect positive pulses, 8–12 V, at ~2 kHz on a short-pulse range.
Then at TP801 on the MD-7918.

Present at the display but absent or degraded at TP801 = broken trigger conductor in the multi-signal cable. This is the cheapest possible fix and the cable is the one thing you have not substituted — both of your display swaps ran through it.

Flex the cable at both connectors and at the strain relief while watching the scope.

Step 4 — magnetron heater, at ST-BY

Across the two "Mag." pins (4 and 5 on that test header). Expect 7.4 – 7.6 Vdc.

Absent or low → MD-7918 fault. Do not simply wind VR801 up to compensate; find out why it's low first.
Note that heater voltage is normal in many "no TX" faults, so passing this step proves little on its own.

Step 5 — TX high voltage, at ST-BY

Expect ~300 – 370 Vdc.

0 V or very low: the DC-DC converter isn't running, the charging resistor R is open, or the modulator FET is shorted and clamping the rail. A shorted FET commonly takes the charging resistor with it — check that resistor for an open circuit, it's a classic.
Present and stable: the supply side is fine, the fault is downstream.

A shorted modulator FET is a very plausible consequence of your floating-ground condition. Check it with the board unpowered and discharged: diode-test drain–source and gate–source, and look for gate–source short (0 Ω) or a drain–source dead short.

Step 6 — TP802, magnetron current, in TX

This is the decisive measurement. DMM on DC volts, TP802 to TP803, radar in TX.

~0 V in TX = the magnetron is drawing no current = you are definitively not transmitting.
~0.2–1.2 V depending on pulse length = the magnetron IS firing, and your problem is downstream of the magnetron — i.e. the circulator (FCX73/HY801), the limiter, the MIC, or the radiator feed.

That single reading splits your remaining fault tree in half. Do it before you buy anything.

Step 7 — pulse transformer primary

Only if steps 3–5 pass and step 6 reads zero. With a 100:1 or high-voltage probe (not a 10:1), look at the FET drain / transformer primary: you should see a clean ~300 V pulse at the PRF, of the selected pulse width. No pulse with good trigger + good HV = dead FET or dead gate driver. Ringing, a slow collapse, or a much-reduced amplitude = suspect the pulse transformer — a shock-cracked ferrite core is a known post-grounding failure and it looks perfect to the eye and to a DMM.

Step 8 — if TP802 shows current but there is still no echo

Then the RF path is the fault, and Furuno's flow chart lands on "faulty magnetron or MIC." You've already replaced the magnetron, so:

Inspect HY801 (FCX73) closely: cracked ferrite, cracked or displaced magnet, deformed housing, disturbed mounting, corrosion at the port interfaces.
Inspect the limiter diode between circulator port 3 and the MIC — a shorted limiter kills both main bang and echoes while leaving IF noise intact.
Check the small coax and the connector at the IF-7758.
Check the radiator feed / slot array for physical damage or water ingress.

7. Post-shock mechanical inspection checklist

Independent of the electrical tests, go over the radome for shock damage:

Cracked solder joints on heavy parts: pulse transformer, chokes, large electrolytics, magnetron terminals.
Pulse transformer ferrite core — inspect under magnification, this is the sneaky one.
Circulator HY801 ferrite/magnet.
Magnetron mounting and its seating into the launcher.
Every connector: J811, J812, J701, the video coax, the MIC coax.
Fuses on the MD-7918.
Any sign of water ingress if the radome seal was disturbed.

8. Getting the real documentation

Highest-value action: the Furuno MODEL 1832/1932/1942 service manual is available on ManualsLib as a 97-page PDF, and pages 79–97 are full schematic diagrams including the MD-7918A and MD-7918E. That is by far the closest thing to your board's schematic that exists in public: https://www.manualslib.com/manual/3545387/Furuno-1832.html

Also worth doing:

furunousaforum.com — Furuno's own technicians answer there and have handled MD-7918 questions specifically.
A Furuno dealer or Furuno service department can reference the actual MODEL 1830 service manual.

9. Summary — ranked suspects

#	Suspect	Why	Test
1	Broken trigger conductor in multi-signal cable	Only untested common element; both display swaps ran through it; matches symptoms exactly; the cable is where your ground fault was	Step 3
2	MD-7918 modulator: FET shorted / charging resistor open / DC-DC dead	Floating ground is a credible kill mechanism; Furuno's own chart lands here on noise-but-no-echo	Steps 5, 7
3	Pulse transformer, cracked core	Classic shock failure, invisible to DMM	Step 7
4	Circulator HY801 (FCX73) or limiter/MIC	Shock-sensitive ferrite/magnet assembly; where Furuno's chart ends after magnetron is excluded	Steps 6, 8

Do Step 1 and Step 6 first. Between them they tell you whether the magnetron is firing at all, which halves the problem for the cost of two meter readings.