The Nozzle Area Position Indicator is fabricated by VDO and is used in the Panavia Tornado fighter jet. The partnumber is ST443-4 and is marked with nato number 6620-12-178-4916. The description on the type plate is 'Nozzle Area Ind.' The Nozzle Position Indicator shows the aperture of the engine outlet-nozzle. In idle the nozzles are open. When applying maximum power, the nozzles opening is reduced. When the afterburner is activated, the nozzles are opened.
I spent some time disassembling the unit. It's a puzzle... After some fiddeling around I found the 'magic trick' to disassemble the module.  I loosened the four screws of the rear panel. This releves the pressure on the rubber gasket. The screws don't have to be removed, but it's possible. Removing the screws makes it possible to remove tha gasket, but beware not damaging the kapton cable.  I removed the four screws from the rear part of the housing.  The knob must be removed, this is probably the hardest part. The knob isn't just a knob, there's are reductor gears built into the knob. (Nice engineering!) The know don't have to be disassembled as I did. Near the axel behind the knob is a c-shaped ring. Remove the clamp with a screwdriver in the housing direction. The clamp is placed in a groove, so a little pressure is needed. Afte removing the clamp, the locking pin is visible. There's a pin pushed trough the knob and axel. Push the the pin through the knob until the pin falls out of the knob. Now the knob cna be pulled of the axel.  Now the frontpanel can be removed by removing the four screws from the front panel.  The assembly can be pulled out of the housing via de rear opening.
The internal construction can be devided in the electronics (rear) part and the mechanical (front) part. Both modules are connected by four screws and a electrical connector. The electrical part has the powersupply at the rear end. There are also a couple of printed circuit boards with probably some servo driver circuit.
The mechanical module contains a DC motor (left top) which drives the dial indicator via some gears. There's also a stacked/double variable resistor (right top) for position encoding. Since it seems that there are two three-wire parts stacked, two variable resistors are likely but a position resolver can also be possible. This has to be find out later on... There's also a manual indicator for setting a threshold level. By rotating the dial, the marker points out the setpoint. The position of the marker is also monitorered by a 2K2 variable resustor (left bottom).
loose DC motor Avionic parts sold on the internet is like gambling... It's always a guess what the condition is of the instrument. Instruments are discarded/sold when the are broken or worn out. If you're lucky, parts are sold if a whole aircraft is scrapped and the instruments are in working condition. But it's always a guess when buying stuff via eBay. In this case the device was rattling when is was moved or rotated. So it's clearly that there is a mechanical problem. After opening the instrument I saw that the DC motor was loose. There are two clamps which should hold the motor, but apparently the screws came loose. This is odd since these screws are usually locked with liquid thread locker. Due to vibrations or a production fault (...) the motor came loose after a while. The good thing is that the repair is very simple. The motor is installed an the two screws are tightened. I didn't use thread locker since the device is probably never to be used in the air again. I wonder what happened in the air when the pilot encountered the problem. The indicator would 'freeze' so the servo loop was never locked. 'Smart' instrument have a warning system built in if the lock time is too long, an error signal was sent via the wire harness to some warming handling device. I don't know (yet) if this warming system is also built in here...
The DC motor was loose in the housing. Luckily ther was no other damage.
The DC motor and the DC motor position are clearly visible.
wrong counter indication There's a white arrow indicating the set threshold level. There's also a 000...999 counter which should correspond with the manual set threshold level. But the indication is way off. I guess I have to reset the counter by removing a gear, resetting the counter and replacing the gear. Than the counter should match the set point. This is a future project...
There are three parallel connected small lightbulbs positioned at the left top of the dial. After testing the voltage should be around 5 VDC. The current draw of the three bulbs is approximately 210 mA. This makes that the three bulbs use approximately 1 Watt of electrical power. There are three bulbs to get redundancy. If one bulb fails, there are two other bulbs to keep the dial illuminated. Since hot filaments are susceptible to failure due to shocks and vibration I suspect three bulbs are used as a backup for the backup. Pins M and N are used to power the lightbulbs.
Inspired by the videos video #5 and video #6 of Michel I got inspired 'reverse engineering' the wiring. The pin numbers are found out by Michel, but the pin connections were unfortunately unknown. Therefore I de-soldered all the internal boards to find out the interconnections between the boards, the electro mechanical parts and the rear connector. The instrument has no 'traditional' wire loom, bus has a double sided flexible kapton ribbon 'board'. The drawing of the reverse engineering is shown below. By clicking the image, the larger image will appear in a new window. Note: the image will be larger, but my handwriting doesn't get better. ;-)
Michel did the hardest part reverse engineering the most of the boards of the device. But the filter board DZ851 wasn't documented. The power input filter board is the least complex board and to get a 'complete picture' of the device, I documented the board below.
After finding out the relation of the rear connector to the boards, the instrument is connected to a test setup. The video of the first test is shown below.
I applied a 400 Hz power supply signal to pins 'J' and 'K'. I used a (Rigol DG1022Z) signal generator to generate a 400 Hz sine wave of 1 Vpp. The 400 Hz 'tone' is amplified by a '100 V line amplifier' to obtain a power supply voltage. Normally 115 VAC is needed, but it turned out a lower voltage works also. It must be mentioned that the power supply voltage affects the deflection of the needle. The lower the supply voltage, the more the needle deflects at the same sensor input voltage. When the device starts without an input signal, the indicator indicates 50%. I applied a test signal to pins 'E' and 'F' to simulate the nozzle area sensor. If an 400 Hz signal is applied, the meter deflects corresponding to the voltage. 0 V is 50% at the scale. A negative voltage 'sends' the indicator to 0%. A positive voltage 'sends' the needle to 100%. But if the voltage is stable, the phase change related to the supply voltage, the needle deflects also! Since the deflection is different between a sine wave, square wave and ramp wave it's not likely the phase is the right way to control the instrument. I assume that a 400 Hz signal without a phase shift related to the power supply is the base situation and that the positive or negative input signal controls the deflection between 0 and 100%.