Most of us will have a hazy idea of how radar works to detect aircraft by listening for reflected radio waves. And we’ll probably also know that while radar can detect aircraft, it’s not the most efficient or useful tool in the hands of an air traffic controller. Aircraft carry transponders so that those on the ground can have a clearer picture of the skies, as each one reports its identity, altitude, and position. [Yeo Kheng Meng] was lucky enough to secure a non-functioning aircraft transponder and do a teardown, and his write-up makes for interesting reading as he explains their operation before diving into the hardware.
The 1978 and 1979 date codes on the various integrated circuits and transistors identify it as having been made in 1979, so not having a CPU is not entirely unsurprising given its age. Instead this is a straightforward device that responds to pulse lengths of different timings with sequential bursts of data.
[Yeo Kheng] is mystified by the RF strip and associated components, which look to us like a typical crystal oscillator and frequency multiplier strip from that era, along with some screened boxes that probably contain cavity filters and given that there is also a high voltage power supply present, a tube RF power amplifier. GHz-capable semiconductors were quite exotic in the 1970s, while high-frequency tubes had by then a long history.
It’s evident that the tech behind aircraft transponders has moved on since this unit was built, but one thing’s certain. Hackers in 1978 would have had to go to a lot of work to listen to them and interpret the results, while here in the 21st century it’s something we do routinely.
Funny enough, transponder information, dialed number and altitude is simple enough.
Its history goes back to WW2 of the first use of military IFF, Identification Friend or Foe.
so that friendly targets were not shot at.
Note the relative lack of internal connectors. Lots of vibration in aircraft, and I have always been told that connectors are a weak point for reliability.
Connectors can cause problems, but they’re also useful for modularity especially with todays electronics. Internal connectors are designed differently and include retention screws to hold them in place. This doesn’t prevent companies from doing stupid things. I once saw a piece of equipment that had standard aircraft connectors on the outsides…. and RJ45 on the inside. As soon as we understood why it wasn’t working, it went out the door.
I am really curious to hear more about how the RJ45 connectors fail in this kind of application.
When I first encountered the “RJ” registered jack connectors I was blown away. In the early 80s they seemed like some exotic artifact belonging to some serial-data-over-twisted-pair Utopian future, which is what they turned out to be in the end.
Intermittent disconnections due to vibrations, or the entire jack coming loose out of the socket due to a big jolt. Any connector that can become disconnected without direct human manual intervention is not a good thing to put on an aircraft. I’ve had a network cable on my computer come loose simply because the monitor cable was resting on the RJ45 retention tab.
That, and there’s no need for interchangeable modules. The transponder has a fixed function, doesn’t need to be “upgraded” or have options. It’s cheaper to make as a single board. It’s a fixed frequency receiver and a fixed frequency transmitter, which transmits pulses.
The only option, so to speak, is whether there’s an altitude encoder attached, and that comes in through the back panel connector.
For a moment I was about to suggest they might want to swap out the transmitter for a different frequency version (e.g. EU/US). Then I realised that might cause issues…!
Transponder frequencies and protocols are global standards, and things change very slowly in aviation. That said, there *have* been major changes and improvements to transponders, and they generally build on, rather than replace, the previous technology. While you’re unlikely to swap parts *within* something like a transponder, modern avionics often do support add-on modules, firmware updates, and the like.
Actually, they do need to be upgraded. First, there is the I/O. Altitude via Gilham encoding is an ancient relic as are a number of air data inputs/outputs. Second, if you are building avionics without modularity, you will suffer greatly when it comes to sustainability. Third, most aircraft are now required to be ADS-B compliant which means a Mode S Transponder (s for selectivity, to deal with congested airspace where squawk codes could run out) or they need to have a UAT. Without upgrading to be compliant, aircraft cannot expect access to certain airspace(s).
Lol no
I worked on transponders in the Navy. They are modular in order to be easier to troubleshoot. Swap half of the cards with known good cards and cut it in half.
If its through the RF chain you can even modulate your signal lol. Thats a fun failure.
Yet they used an RCA plug on the transmitter. Well, they’re cheap.
The article mis-states the reason for having altitude information encoded in the transponder response. Radar equipment is usually capable of determining an altitude, but the problems with anomalous propagation bending the signal over long distances mean that that value is very often wrong. Military aircraft also carry altitude encoding transponders for this reason.
Air route surveillance radar can’t distinguish height information. It’s 2D. At long range, they wouldn’t be able to accurately measure height anyway. The altimeter that feeds the transponder signal is pretty accurate, good enough to space aircraft.
*a big fancy phased array radar can very precisely determine the 3D position in mere fractions of a second even if the aircraft is not cooperating, trouble is those are over an order of magnitude more expensive then simple 2D radars with normal antennas, so generally only the military has the budgets to pay for them.
Hopefully as tech keeps advancing, they might slowly become cheap enough and creep into civilian ATC…
Actually, primary radar is being phased out in favor of cooperative methods like ADS-B. Big parts of Canada aren’t covered with primary radar, and air traffic control relies on ADS-B position information to separate aircraft. Basically, it’s the aircraft themselves telling ATC where they are. Apart from multilateration, the ground part of the system has very few options to verify wheter the planes are telling to truth or not…
+1
Bob Pease would approve of this title.
+1 Loved Bob’s articles.
He’s certainly an influence of mine. :)
The tubular thing supplied by the HV power supply could be a pcarcinotron, permaktron or some other kind of travelling-wave tube. It could be used as amplifier or mixer. To say more, I would have to hold the thing in my hands…
*carcinotron
I’d guess a lighthouse tube, a simple triode amplifier. Needs to be driven by a coded pulse.
Well, I see it this way – there is heating on the side marked as “K”. A coaxial cable goes in at this point as well. So, it is probably the input – I see no point in having cathode follower used as amplifier.
Then there is another coaxial cable coming out of the tube form the middle. This is probably the output – it has cinch connector, probably with matched impedance. The output then goes into the filters and to the antenna.
The last coaxial cable goes into the anode and probably goes into the HV power supply. Since on this part of PCB there is nothing RF-related by the looks, the anode is probably only fed with HV and shielded so that it does not have unwanted emissions into the rest of the device.
Nearby the suspected output on the tube, there is a screw, probably for tuning.
I think it isn’t an usual vaccuum tube like triode because it has separate anode and RF output, also it has a tuning screw and it is completely shielded by metal (yeah, I know that military used all-metal tubes, I have several black all-metal Russian triodes). Never have I seen a triode with tuning screw. But I can be wrong.
Theoretically, it could be a reflex klystron but it doesn’t have the usual resonance cavities. The size of the whole tube could be lambda/4 at this frequency, but I don’t know how reflex klystrons without cavities work. It could be a similar device (look into the summary at the end of my comment).
A magnetron would require magnets around it. Magnets were heavy and bulky in 1979 and there are none in the pictures. So probably not a magnetron.
A carcinotron quite matches the shape and terminals. However, carcinotron requires magnets as well.
Klystron also quite matches the temrinals, but also needs magnets.
Here, it is called a “GE resonance cavity” and it needs 1400 V for its anode (usual cavities without amplification do not require power supply and also don’t have anode): http://www.aeroelectric.com/Installation_Data/Jerry_Gordon%27s_TecnNote.pdf
It also mentions modulator connected to the 1400 V line – it could be modulating the frequency using anode voltage variation… It also says that it can output cca 250 W of RF power (probably only pulsed power) which was probably impossible using solid-state technologies in 1979. In 1976 in our country (eastern block, so don’t take this as representative of the tehcnology at the time), the best transistor on market was KF630S which was made only for VXW100 transceivers (available only to army) which could output 1 W at 146 MHz. Nowhere near 250 W at 1030 MHz.
A manual can be found, but has to be bought. It exists both in printed and digital form, costs about $40 for the digital manual.
Summary: I don’t know what it is. It probably isn’t a usual triode due to separated anode and output and tuning screw. It doesn’t seem to be anything else I have ever had in my hands because it lacks magnets to be a carcinotron, klystron or magnetron. It also lacks the usual resonance cavities found in reflex klystrons. However, it could be a combination of reflex klystron with a usual tube – i have seen this used once in an amateur and improvised way here http://ok1ike.c-a-v.com/soubory/klystron.htm . The whole tube could be designed as a resonator (coarsely tuned by the screw and finely by its working point using a PLL), then it could work as reflex klystron even without the usual cavities.
@MS-BOSS said: “The tubular thing supplied by the HV power supply could be a pcarcinotron [sic], permaktron or some other kind of travelling-wave [sic] tube.”
“Permaktron” – haven’t heard that one for awhile – same as a Traveling Wave Tube (TWT), but in the old Eastern Bloc(?) It could very well act as the transmit amplifier in the transponder.
“Carcinotron” – yeah that’s a strange one too. In the U.S. we call this one a Backward Wave Oscillator (BWO). A BWO is not a TWT, but it does have a slow wave structure. This could be used as a modulated oscillator/transmitter in the transponder application, but probably not. Output power is rather limited unless the BWO gets large and heavy and BWO’s are not known for frequency stability and low phase noise.
[Yeo Kheng Meng] is asking himself whi the box is full of NAND gates.
Well, NAND gates are one of the most versatile logic gates. Every digital logic circuit (including modern CPUs) can be described by and constructed from a bunch of NAND gates.
Plus i can cery well imagine a circuit where the bit clock shifts through a number of flip-flops (comprised solely of NAND gates) and being compared with the dialed-in sqawk to emit the corresponding digital signal throug the RF transmitter.
For somebody who did design TTL-graves in the 80s it is actually a standard circuit job. Reversing the logic shouldn’t be too difficult at this size and scarcity of components.
I thought NANDs were easier/more reliable as well?
I’ve had to reverse engineer and tune up the stationary version of this equipment, used by primary radar sites. Besides a higher power transmitter and a large antenna (on top of the primary radar), much of the RF chain is similar. The information code is kind of weird, it’s backwards in time with an 8-4-2-1 code. But I piped it into a channel on my PPI and was able to decode it by eye by looking at the pulses and adding them up to get the squawk.
The tweakers in the circuit adjust drive levels for the RF chain. The shielded circuit at the top is probably the 60 MHz IF strip. The thing in the silver tube with many wires looks like the transmitter tube (note the heat sink). It operates at a very low duty cycle, and transmits the coded pulse only when triggered by an interrogator pulse. It sends back a 4 digit squawk, height, and a couple mode bits. You can tell the radar that your radio is out or you’re getting hijacked, for instance.
I just noticed the “K” terminal on the transmit tube can, probably means cathode. It’s made by GE so maybe a special cavity/tube socket. The big square silver plated can could be the diplexer (1030-1090 filters).
If this unit is dead I’d check the tube first, then the HVPS. It would be cheaper to retrofit with a transistor amp than to find a tube. You could check the receiver to see if it picked up interrogator signals, then see if the driver stages put out a response. Worth fooling around with given the cost of new units.
Hard to do a Big Clive teardown from a picture.
I tried adding a tweaker to an old 20 channel CB rig once and it just smoked all the crystals.
As a pilot, this is how they work for me:
TOWER: “N123AB, transponder appears inoperative.”
Me: “Ok, we’ll re-cycle, 3AB.”
*Switches the transponder from STDBY to ALT*