EEVblog® Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: D Straney on June 30, 2024, 11:01:56 pm
-
The last couple years I've been keeping an eye out for interesting electronics to take apart (https://hackaday.io/DCS) in gold-scrap auctions or from industrial surplus sellers. I've been lucky enough to find some space electronics, and thought other people might be interesting in seeing these too.
Index:
- Hughes mystery RF box: (see below)
- Space Micro microwave transmitter (https://www.eevblog.com/forum/projects/avionics-reverse-engineering-spacecraft-equipment/msg5565207/#msg5565207) & bare dies 1 (https://www.eevblog.com/forum/projects/avionics-reverse-engineering-spacecraft-equipment/msg5568059/#msg5568059), 2 (https://www.eevblog.com/forum/projects/avionics-reverse-engineering-spacecraft-equipment/msg5569371/#msg5569371), 3 (https://www.eevblog.com/forum/projects/avionics-reverse-engineering-spacecraft-equipment/msg5569373/#msg5569373)
- Broad Reach I/O board (https://www.eevblog.com/forum/projects/avionics-reverse-engineering-spacecraft-equipment/msg5577507/#msg5577507)
- Hughes Travelling-Wave-Tube HV power supply (https://www.eevblog.com/forum/projects/avionics-reverse-engineering-spacecraft-equipment/msg5678733/#msg5678733)
Hughes mystery RF box
(https://live.staticflickr.com/65535/53759085730_531e4a921b_z.jpg) (https://flic.kr/p/2pUvnfs)(https://live.staticflickr.com/65535/53759004804_d04375000b_z.jpg) (https://flic.kr/p/2pUuXcb)
The seller listed this as coming from Hughes Aircraft, which means it was likely from their Space & Communications Group (https://en.wikipedia.org/wiki/Hughes_Aircraft_Company#Hughes_Space_and_Communications_Group), which is now Boeing (https://en.wikipedia.org/wiki/Boeing_Satellite_Development_Center). From this and other sources (a 1992 promotional photo showing the Hughes Space & Comms group's spacecraft) (https://www.ebay.com/itm/364832045034), their main business seemed to be communications satellites.
So, we can expect this to be most likely communications-related. This particular module has a couple DE-15 connectors for power & control, and generates an RF output of some kind, as well as having another coax that according to the "T.P." label serves as an internal testpoint(??):
(https://live.staticflickr.com/65535/53759085720_4875b19e91_z.jpg) (https://flic.kr/p/2pUvnfh)(https://live.staticflickr.com/65535/53758668786_97f600fc5c_z.jpg) (https://flic.kr/p/2pUteiL)
DC control/power half
After removing many screws and taking off the top lid, you can see the low-frequency half:
(https://live.staticflickr.com/65535/53758668591_6b0e6da64b_z.jpg) (https://flic.kr/p/2pUtefp)
The outer section contains relays, diodes, and filtering components, which probably switch & filter bias power to the RF section, via the many feed-throughs (https://www.engineering.com/the-engineers-guide-to-feedthrough-capacitors/) visible in holes in the board.
The inner section contains an op-amp (DIP package, JM38510/13503 = OP27A) and possibly a second op-amp (metal can labeled "OP37034J", might be an OP37 according to one obscure reference). This inner section, judging from the feed-through connections, appears to provide the final amplification and/or complementary bipolar drive voltages for biasing a varactor in the RF portion (we'll get to that in a bit).
(https://live.staticflickr.com/65535/53757765772_db8ca19bd2_z.jpg) (https://flic.kr/p/2pUoASy)
You can see the thick layer of conformal coating, and the very nicely-bundled-and-epoxied wires:
(https://live.staticflickr.com/65535/53759004799_bcec1e3304_z.jpg) (https://flic.kr/p/2pUuXc6)(https://live.staticflickr.com/65535/53757765692_bbcb32352d_z.jpg) (https://flic.kr/p/2pUoARb)(https://live.staticflickr.com/65535/53758875423_bf9b36331e_z.jpg) (https://flic.kr/p/2pUuhJt)
RF half
The other lid, with even more screws for all the separate compartments, holds a wonderful variety of microwave magic:
(https://live.staticflickr.com/65535/53759085670_2598b3d969_z.jpg) (https://flic.kr/p/2pUvneq)
There's a lot of additional internal shields, but after removing those, the circuitry becomes more clear:
(https://live.staticflickr.com/65535/53759085615_d3ef60885d_z.jpg) (https://flic.kr/p/2pUvndt)
Starting in the top-right corner...
1. Oscillator #1 & amplification, filtering
(https://live.staticflickr.com/65535/53757765777_a49e31a640_z.jpg) (https://flic.kr/p/2pUoASD)
The metal can seems to be an oscillator: it's labeled "122.542338 Mhz". This drives a series of amplifier stages, in the form of RF transistors in interesting glass-lid packages. It's possible, depending on the biasing, that some of these are being driven into saturation to create harmonics for frequency multiplication, but I have no way of knowing.
(https://live.staticflickr.com/65535/53757765802_c9d9c1b8e9_z.jpg) (https://flic.kr/p/2pUoAT5)
A filter follows, which has an interesting combination of distributed elements (the rectangular traces) and lumped elements (the spring-looking inductors):
(https://live.staticflickr.com/65535/53758668616_040804dcb8_z.jpg) (https://flic.kr/p/2pUtefQ)
A couple more amplifier stages follow the filter:
(https://live.staticflickr.com/65535/53757765842_db6b46f68c_z.jpg) (https://flic.kr/p/2pUoATL)
2. SRD comb generator & filtering
After this, some frequency multiplication almost definitely happens: the signal goes through a lumped pi filter, to a diode connected in a shunt configuration to ground. See the middle of this photo:
(https://live.staticflickr.com/65535/53759085555_031e80582a_z.jpg) (https://flic.kr/p/2pUvncr)
This is probably a Step Recovery Diode (SRD) (https://en.wikipedia.org/wiki/Step_recovery_diode) or something similar. SRDs have a reverse-recovery current when reverse-biased which stops very suddenly; if you excite it correctly with an AC current you can create a train of very short pulses, which due to their sharpness contain many harmonics of the input frequency. For frequency multiplication, you can then apply a bandpass filter to pick out the specific harmonic that you want, while suppressing the others. In this case, you can see the non-uniform parallel-coupled lines filter (https://en.wikipedia.org/wiki/Distributed-element_filter#Parallel-coupled_lines_filter) directly following the SRD.
I don't know the dielectric constant of the ceramic substrate, but the length of the filter sections (~3 mm for 1/2 wavelength) suggests somewhere in the 20-40 Ghz range.
Oscillator #2 & amplification
Let's back up a bit now before shit gets weird, and look at the cavity at the top-center: this holds an oscillator tuned by a dielectric resonator, or a DRO (Dielectric Resonator Oscillator). This type of oscillator is common in satellite dish up/downconverters, and uses a high-permittivity ceramic material as an E-field resonant cavity due to the cylindrical resonant modes it can support. The resonant frequencies depend only on the geometry of the resonator & its permittivity: go here (https://www.sciencedirect.com/topics/physics-and-astronomy/cavity-resonator) and click on "Read More" on the first article (https://doi.org/10.1016/j.pnmrs.2014.09.003 (https://doi.org/10.1016/j.pnmrs.2014.09.003) by Andrew Webb) for an explanation of resonant cavities. There's also some pretty good diagrams from that article here (https://ars.els-cdn.com/content/image/1-s2.0-S0079656514000624-gr7.jpg) and here (https://ars.els-cdn.com/content/image/1-s2.0-S0079656514000624-gr13.jpg) which help visualize what's going on.
(https://live.staticflickr.com/65535/53759085565_e188322bb7_z.jpg) (https://flic.kr/p/2pUvncB)
In our case, that yellow disc is the dielectric resonator, the trace running next to it loosely couples the signal to and from the resonator, and the gold part to the left of it is the transistor that adds amplification to turn it into an oscillator. I assume that the transistor's reverse-transfer(/Miller) capacitance is what adds the feedback needed to create the oscillation.
Also coupled to the resonator (better visible in photo below) is a separate trace with a white varactor, which is used to slightly detune the resonator, and therefore adjust the tuning of the oscillator:
(https://live.staticflickr.com/65535/53759004984_bd62a30077_z.jpg) (https://flic.kr/p/2pUuXfh)
You can see that the anode and the cathode of the varactor both go to semicircular RF-blocks, and then to feedthroughs: the op-amp(s) in the DC half seem to drive the voltage across this varactor, to set the fine-tuning control of oscillator #2.
Splitting & mixing
The oscillator #2 output signal goes through a couple stages of amplification, and then enters a branchline coupler (https://www.microwaves101.com/encyclopedias/branchline-couplers) which splits the signal along two paths.
(https://live.staticflickr.com/65535/53759004949_6f13e65e4c_z.jpg) (https://flic.kr/p/2pUuXeF)
One path goes to the output coax connector, through a chain of amplifiers (along the left side in photo):
(https://live.staticflickr.com/65535/53759004894_4032fd9fb5_z.jpg) (https://flic.kr/p/2pUuXdJ)
The other path goes through a circulator to a mystery module. This mystery module gets its other input from the filtered SRD-multiplied reference frequency from oscillator #1:
(https://live.staticflickr.com/65535/53758875578_9cef2f956e_z.jpg) (https://flic.kr/p/2pUuhM9)
Unscrewing and flipping over the mystery module reveals it's made by Watkins-Johnson:
(https://live.staticflickr.com/65535/53757765947_42791e0ae4_z.jpg) (https://flic.kr/p/2pUoAVz)
It has a 3rd connection, which goes straight through a feedthrough to the DC side above, where it connects to the board with the op-amps.
What's actually happening here?
The only thing that makes sense to me, with this set of connections, is this:
Oscillator #1 is a stable reference frequency source (generating 20-40 Ghz via the SRD), while oscillator #2 is an inaccurate but tunable oscillator, which is also used as the output of this whole box. The mystery Watkins-Johnson module is a mixer, which is used to compare the two oscillator outputs. The op-amps in the control section here adjust the DC tuning voltage on oscillator #2's varactor until the two oscillators match frequency - this keeps oscillator #2 (and therefore the final RF output) frequency-stable in the long-term. However, the varactor-driver op-amp also sums the feedback control with an external modulation signal from one of the DE-15 connectors, which allows short-term changes to the tuning to produce frequency and/or phase modulation on the final RF output. Something like this would normally be used as the RF signal source for a transmitter.
There's other ways to produce a stable-but-modulated frequency for communications: you can...
- split your frequency reference into I and Q signals (in-phase, and 90-degrees shifted), amplitude-modulate each one separately, and then re-combine to produce phase and amplitude modulation
- mix directly with your frequency-modulation signal, if you have a double-balanced mixer to suppress the LO in your output
- do multi-stage upconversion like the "exciter" from a SINCGARS military radio I looked at before (https://www.eevblog.com/forum/projects/reverse-engineering-some-old-sincgars-military-radio-boards/msg5231139/#msg5231139): this is similar to the previous item
- etc.
However, the I/Q scheme doesn't allow for frequency modulation, and working in the 10s-of-Ghz range severely limits your options for active devices; you'll have a very hard time implementing a good balanced mixer (especially considering the 1994 date code on one of the op-amps), so I can see why you'd go for this dual-locked-oscillators scheme.
Hope this was interesting, let me know if you've got any info or insights I missed.
-
Meant to mention, the external "T.P." coax connector is connected to the filtered, amplified, pre-SRD output of oscillator #1 (the metal-can frequency reference).
Also here's a better photo of the SRD: you can see the package standing on end, with one terminal attached to the case (ground) and the other terminal with "pass-through" foil bonds to the circuit sections before and after it. There's a small gap just to the left of the SRD, on the output side. This acts as a very small-value capacitor, which serves as a high-pass filter to block all the unwanted low frequencies (such as the input frequency) that come out of the frequency multiplication.[attachimg=1]
-
Great post with technical explanations. I hope your hands did not become tired after dealing with so many screws. By the way, were screws lock-tited (beside dab of gray glue on side)?
Do I understand correctly that this metal box with cavities is milled from single block? Are RF ceramic assemblies and DC biasing circuits bolted to the same central metal plate, just from 2 sides? Probably provides heatsinking function as well.
Implementation of dielectric resonator looks very interesting, yet simple at first glance. Small hole is visible on top resonator surface, was it originally there?
There is so much effort in interference suppression (multiple circulators, RF absorbers, metal cans in metal cans, etc.). I wonder how long it takes to assemble and adjust as there are so many adjusting gold strips added.
Those glass transistor packages are pre-tested wire-bonded and encapsulated bare die transistors/amplifiers. I guess they could be added to big ceramic plates directly without glass encapsulation, but wire-bonds and extra small components were deemed too unreliable at this scale. There are no wire bonds visible on big ceramic plates, only goldstrips. Can you do a close-up photo of glass encapsulated transistors?
-
Are those boards definitely ceramic? They kinda look like teflon.
-
Thanks! Yes, removing the screws from the top and bottom was absolutely an exercise in patience.
Screw locking: I think it was mostly just the dab of glue on the side for external and internal screws.
Enclosure construction: That's right, I'm about 99% sure it's made from one single solid block, with the middle plate used for mounting from both sides as you describe. Kind of a subtle addition to the "no expense spared" vibes of the gold and the many specialty substrates, to take a big (potentially-special-alloy) metal block and mill away most of it.
Dielectric resonator: I'm not sure about the chip on the corner of the resonator; I vaguely remember (this was sometime last year) having to put in extra effort to remove that particular lid, so it's possible that in the process of levering up the lid with one of my beater "mini-crowbar / too-worn-for-normal-use" screwdrivers that I hit the resonator. (Wouldn't be the first time) I do wonder how much that would affect the resonance though, and whether it's well within the tuning range of the varactor or not (might affect the Q more than the frequency?)
Gold adjustment strips: Yes, good eye! I'd meant to call attention to those too, they definitely stood out looking up-close in person; can only imagine some poor technician spending a week per box staring at a VNA while repeatedly bonding bits of foil.
PCB material: I see what you're getting at, with the color, it sure does look like Teflon in the photos. In person though it's very hard (under tweezers) and shiny.
Transistors & bonding: Yeah I was curious about those bonded foil strips between components & board sections: I guess wirebonding-style thermal/ultrasonic methods get used for the same size ballpark when assembling large-die power modules, so it could be the same. Getting an up-close view of the transistors is a good idea - my normal camera doesn't get in that far, but I regularly visit the "open electronics night" at my local makerspace for their high-zoom microscope, and was just wondering what to put under it next time...
-
PCB material: I see what you're getting at, with the color, it sure does look like Teflon in the photos. In person though it's very hard (under tweezers) and shiny.
It was the shininess that made me think it was teflon. Technical ceramics usually give falrly diffuse reflections. A lot of defence RF stuff uses pure teflon PCBs for its high frequency performance. They are a PITA, as teflon flows so well. It need lots of support to keep it dimensionally stable over time, especially in things like missiles, where vibration is so intense it can liquefy all sorts of things. I think dealing with liquefaction is probably the key skill that makes someone a rocket scientist.
-
Enclosure construction: That's right, I'm about 99% sure it's made from one single solid block, with the middle plate used for mounting from both sides as you describe. Kind of a subtle addition to the "no expense spared" vibes of the gold and the many specialty substrates, to take a big (potentially-special-alloy) metal block and mill away most of it.
HDD cases are roughly similar to this RF box - cases are cast first, then last few % milled. That saves few $ when done in volumes. Here we see "no expenses spared" solution aimed at ultra high performance and reliability, thus milling straight from the block.
Note that case surfaces are coated with conductive yellow metal (gold solution?) and most walls/surfaces have RF absorbers.
-
I`d say this is machined from a block and not cast. Setting up casting is quite a process, makes sense If you want to produce millions of units. For a one-of-one item, machining is way cheaper and faster. Still expensive... Plus you might be able to use better suited materials.
-
I meant that this RF box is milled from one block and not cast. HDD cases are cast, then processed.
-
I meant that this RF box is milled from one block and not cast. HDD cases are cast, then processed.
Military grade RF boxes are normally milled from a block. That was a PITA 50 years ago, but CNC has made is easy, if still not exactly cheap. Casting leaves too many voids for their purposes, and doesn't offer the same strength.
-
great post, really enjoyed it
-
It was the shininess that made me think it was teflon. Technical ceramics usually give falrly diffuse reflections. A lot of defence RF stuff uses pure teflon PCBs for its high frequency performance. They are a PITA, as teflon flows so well. It need lots of support to keep it dimensionally stable over time, especially in things like missiles, where vibration is so intense it can liquefy all sorts of things. I think dealing with liquefaction is probably the key skill that makes someone a rocket scientist.
Huh that's interesting, makes sense that the creep would start causing all sorts of problems with tuning. The teflon I've seen has been either raw stock for machining (where it's got a pretty dull surface) or the Rogers-type microwave substrates where there's other stuff in there too to explain the color/texture, wouldn't have expected it to be shiny.
-
Interesting post. Thanks for the explanations. I would have guessed microwave RF and that’s as far as I would have gone. Did you weigh the unit? With the metal case, looks hefty. Do you think an identical device could have been part of a satellite or more likely a ground-based or airplane RF unit?
Mike
-
Glad you liked it - I haven't weighed the box and I'm terrible at guessing weights, but it is pretty heavy for its size, with a lot of solid metal! Ground equipment for space installations is usually rack-mount stuff that looks much more "normal", as it doesn't have to put up with unusual environmental conditions; everything about the packaging on this one, based on lots of photos of other space hardware, says "I'm made for spaceflight". Can't be 100% sure of course, but my best guess is that other copies of this are on a communications satellite somewhere (or that this was a prototype for a design which ended up on a satellite).
Ok, here's #2.
Space Micro microwave transmitter
These few boards came in a set (for gold scrap), and all appear from the labeling to come from the same unit: a "µXTx"-model X-band (~8-12 Ghz) microwave transmitter, which takes in a digital data stream and outputs modulated & amplified RF directly to an antenna. Here's a datasheet for a very similar model (https://www.spacemicro.com/products/communication-systems/%CE%BCXTX-200%E2%84%A2%20X-BAND%20TRANSMITTER.pdf), except these boards date from about 2008-2011. The manufacturer is Space Micro, a commercial supplier of electronics & optics for satellites. These are not COTS parts meant for low-cost cubesats where the lifetimes are measured in weeks or months though; the Space Micro products seem to be mostly radiation-hardened high-reliability stuff meant to work for a decade+ (that translates to more-interesting parts for us to look at!).
Digital Board
(https://live.staticflickr.com/65535/53837664306_e9de9f463f_z.jpg) (https://flic.kr/p/2q2s6Vu)(https://live.staticflickr.com/65535/53837664296_70636ec554_z.jpg) (https://flic.kr/p/2q2s6Vj)(https://live.staticflickr.com/65535/53838106305_54094dd9c6_z.jpg) (https://flic.kr/p/2q2unja)
To see where the data starts its journey, let's look at the digital board first. The digital data stream most likely enters from outside the box into the connector at the bottom-left. This looks a lot like a specific connector series made by Samtec that I've used before, which has impedance-controlled contact geometry and keeps a ground plane present at all times, with a flat plate on the male side and a bunch of mating contacts on the female side - this is designed specifically for, and works great for, carrying high-speed digital signals. Here's a link to the Samtec Q-strip series, where you can see some images of what I'm talking about (https://www.samtec.com/high-speed-board-to-board/dual-row-mezzanine-strips/qstrip/).
This data makes its way to the Actel ProASIC3E FPGA that does almost all the digital processing on this board:
(https://live.staticflickr.com/65535/53838106325_ca0c52a863_z.jpg) (https://flic.kr/p/2q2unjv)
There's also a mystery ASIC to help it out - the "LDPC" in the name might stand for "Low-Density Parity Check" coding (https://en.wikipedia.org/wiki/Low-density_parity-check_code). I don't understand exactly how they work, but they seem to find a lot of use in communications, which supports the idea of this ASIC being a hardware accelerator. You can also read an article here on its implementation in an ASIC for multi-Gbps data (https://link.springer.com/article/10.1007/s11265-021-01680-0), which again seems pretty applicable here.
(https://live.staticflickr.com/65535/53838106365_41f3f43597_z.jpg) (https://flic.kr/p/2q2unkc)
The FPGA also has access to a 128K x 8 EEPROM (5962-3826716). I'm not sure exactly what this is storing; can't imagine it's the FPGA configuration but I don't see a dedicated EEPROM for that, or it could also be program code for a soft processor used in the FPGA.
(https://live.staticflickr.com/65535/53838017369_227cda0678_z.jpg) (https://flic.kr/p/2q2tUSM)
This system is an SDR (Software-Defined Radio): the modulation is all done digitally on the FPGA, which then drives a DAC to produce the modulated IF signal. Flexibility is very important in space applications (as you can't just pop the cover and swap a board once it's in orbit...if you're not a high-priority space telescope, that is) and the datasheet for the uXTx transmitter mentions many aspects that are configurable in the field, such as modulation scheme - this is how that's done.
(https://live.staticflickr.com/65535/53837664266_c5b2020165_z.jpg) (https://flic.kr/p/2q2s6UN)
The DAC itself is a Texas Instruments DAC5675A (https://www.ti.com/product/DAC5675A-SP), which does an impressive 14 bits @ 400 Msps. This is fast enough, if run at full speed, to produce a modulated IF output in the 10-100 Mhz frequency range.
You can also see an interesting pattern here, which is repeated throughout these boards: notice that the footprint on the board is meant for the much wider ceramic & gold package, of the rad-hard space-grade part you can view here (https://www.ti.com/content/dam/ticom/images/products/package/h/hfg0052a.png:large). But they've used a little daughterboard which adapts the this footprint to the infinitely-cheaper commercial plastic-packaged version of the DAC5675A - this strongly suggests that these boards were an early-stage prototype/bench-evaluation copy, definitely not a model meant for environmental testing or flight. When you might blow up some chips and just want to test the basic design functionality in a climate-controlled lab, why waste thousands of dollars per IC?
The DAC gets its clock from the modulator board (we'll look at that next), which provides a reference clock so that all the clock sources are synchronized. This clock enters through one of the two coax connectors, gets converted to LVDS by a DS90LV031A, and sent directly to the DAC. A second copy of the incoming clock is also converted to LVDS and sent to the FPGA, but first it goes through the winding delay line at the left here. You can see the multiple groups of three resistor footprints: at each one, the path can be tuned to be slightly longer or shorter (therefore tuning the delay time) by choosing whether the two "longer-path" or the one "shorter bypass path" resistor is populated.
(https://live.staticflickr.com/65535/53838106295_fd34ed85d0_z.jpg) (https://flic.kr/p/2q2uniZ)
The FPGA can't change change the output data to the DAC right at the DAC's clock edge(s): it needs to change the data somewhere in the middle of the clock cycle, to meet setup and hold time requirements. I believe that timing constraint is being enforced here by this hard-wired delay, to provide the FPGA with a clock that can be used to synchronize the output data to the DAC.
I did a similar thing about 12 years ago, but with an ADC sending a few lanes of DDR serial data to a Spartan 6 FPGA. The reference implementation there from TI/Xilinx used a tunable delay element built into the FPGA's I/O block, which was automatically tuned at startup by setting a test pattern, finding the delay values that would align the data with both clock edges, and then setting a final delay value halfway in between those two for maximum headroom. That's much harder to pull off with outgoing rather than incoming data, though, plus I don't know if this rad-hard FPGA even has a tuneable delay; it also adds extra complexity and room for things to go wrong, if the desired clock delay is a known quantity and can be determined ahead of time.
Anyways, the DAC's output now drives a coax output via a transformer that does the differential-to-single-ended conversion (top-right in the photo below). You can also see the linear regulator (MSK 5900RH) which likely provides a low-noise supply for the DAC.
(https://live.staticflickr.com/65535/53838106300_54094dd9c6_z.jpg) (https://flic.kr/p/2q2unj5)Digital board (https://flic.kr/p/2q2unj5) by D Straney (https://www.flickr.com/photos/147639706@N02/), on Flickr
Finally, there's an LM139 quad comparator. I don't know what this does: maybe it monitors RFPA temperature for the FPGA? (We'll see that later) From the connections, and lack of resistors on the bottom side of the board below the comparator, it looks like only one of the comparators is actually used.
(https://live.staticflickr.com/65535/53837664286_f50dd008f9_z.jpg) (https://flic.kr/p/2q2s6V9)
Modulator Board
This board generates the reference clock for the digital board, and up-converts the digitally-generated IF signal to the X-band output frequency.
(https://live.staticflickr.com/65535/53837923998_f97ed0d3ca_z.jpg) (https://flic.kr/p/2q2tr7W)(https://live.staticflickr.com/65535/53837923988_b6dd8821cc_z.jpg) (https://flic.kr/p/2q2tr7L)
It's a sign of how far miniaturization has come, that most of the complex RF functionality (VCOs, mixers, etc.) is stuffed into a few MMICs in a row along the bottom...
(https://live.staticflickr.com/65535/53838021164_2287f254c7_z.jpg) (https://flic.kr/p/2q2tW1d)
(https://live.staticflickr.com/65535/53838111105_d5d9974d3b_z.jpg) (https://flic.kr/p/2q2uoJV)
...while the largest component by far is just a passive quadrature coupler, which splits a signal into in-phase and 90-degree-shifted components:
(https://live.staticflickr.com/65535/53836776337_5cd169bf9f_z.jpg) (https://flic.kr/p/2q2nxXF)
I really have no idea why this particular part is used; the MITEQ catalog lists it as supporting 50 - 110 Mhz @ 250W average, but it never sees even anywhere near a single watt on this board. With such a low frequency range I can't imagine the coupler uses distributed elements within, which would make it physically large no matter the power level...but it's also very flat, so who knows.
Anyways, things begin at an empty footprint, which I'm 99% sure used to hold a reference oscillator module like the one from the Rocketdyne CPU board (https://www.eevblog.com/forum/projects/avionics-teardown-looking-at-some-mystery-rocketdyne-boards/msg5508781/#msg5508781):
(https://live.staticflickr.com/65535/53838109835_6dd6bed14f_z.jpg) (https://flic.kr/p/2q2uon2)
(https://live.staticflickr.com/65535/53733015721_b6b4183b86_z.jpg) (https://flic.kr/p/2pScKxx)
The chip in charge of generating the LO signal for upconversion is this Peregrine PE9763, which contains all the functionality of a frequency synthesizer, minus the VCO itself:
(https://live.staticflickr.com/65535/53838109745_02663f4b18_z.jpg) (https://flic.kr/p/2q2uokt)
Its parallel control inputs probably get set by the FPGA.
If you look at how everything connects, it all makes sense:
[attachimg=1]
The PE9673 frequency synthesizer maxes out at 3.2 Ghz, but the design here takes advantage of the fact that the VCO has a "divided-by-2" output to synthesize a higher LO frequency. It still doesn't entirely make sense, as the VCO (Hittite/Analog Devices HMC510) has an ~8.5-9.5 Ghz range, which is still >3.2 Ghz when divided by 2. Maybe I missed something when tracing the connections. Either way, the DAC gets a divided-by-16 copy of the VCO as its clock. The upconverted RF output goes out to the power amplifier on a different coax connector.
The long metal-boxed filter down the side of the board seems to be anti-aliasing for the DAC. The input would be at the port marked "output" on the filter, but as it's a passive device with the same impedance at input and output ports, the input and output are going to be interchangeable anyways - the actual filter schematic inside is almost definitely perfectly symmetric.
The one uncertain part is the mystery device with a custom-looking part number here, which I haven't been able to find any references to. It has DC biasing resistors connected to all the RF signals that go through it, and given its place in the circuit, the only thing that makes sense in context is if it's a dual amplifier (see schematic above).
(https://live.staticflickr.com/65535/53838021169_1256fbcdfa_z.jpg) (https://flic.kr/p/2q2tW1i)
I have no idea why it would be necessary to make a custom package with two RF amplifiers here, as this doesn't seem like a particularly unusual need, or space-constrained in any way. Let me know if you've got better ideas. Maybe I'll pop the lid on this device and have a look inside to confirm.
RF Power Amplifier
The final board in the signal chain here is the power amplifier, which boosts the modulated RF to a few watts for transmission through the external antenna.
(https://live.staticflickr.com/65535/53836783297_1187c4afdb_z.jpg) (https://flic.kr/p/2q2nA2F)(https://live.staticflickr.com/65535/53838027864_cedfe86887_z.jpg) (https://flic.kr/p/2q2tXZJ)
You can see that it's meant to dissipate some real power, as the base plate is a solid metal block, which the power transistors are connected to directly.
Things start at the input connector, where the modulated RF signal goes through a couple stages of preliminary amplification (HMC346 & HMC441 amplifiers)...
(https://live.staticflickr.com/65535/53838119650_2e6f4f133f_z.jpg) (https://flic.kr/p/2q2urhf)
...before driving the first dedicated power transistor:
(https://live.staticflickr.com/65535/53837933513_f831d82d23_z.jpg) (https://flic.kr/p/2q2ttWZ)
The output stage consists of two paralleled transistors sharing the output power: the signal gets split in half through a branchline coupler, fed to the two larger power transistors, and then re-combined through another branchline coupler.
(https://live.staticflickr.com/65535/53838119665_c6a986dfda_z.jpg) (https://flic.kr/p/2q2urhv)
Other features of the output stage, visible in this photo:
- A directional coupler (the upside-down "U" at the left) picks off a small portion of the output signal. The forward-power signal from the coupler goes through a resistive attenuator, and then into a mystery part (large gold box at bottom-left). This mystery box also connects (from its right side) to one of the discrete-wire connections that bring DC power to the RFPA. Because the connections to this second signal are not set up at all for high frequencies, it must be DC of some kind - my best guess is that the "mystery part" is a power detector, used to monitor the output power level.
- A circulator (silver box at left, likely with resistive termination inside) to keep out reflected power from the filter (discussed next) & antenna.
- Output filtering, just before the coax connector (top-left corner): the standout feature here is all the little semicircular wideband quarter-wave stubs. These are normally used as an RF short at their quarter-wavelength frequency to block RF on DC control traces - you can see the ones used for this purpose at the bottom-right & top-right of the photo, on the DC bias lines for the power transistors. However, here, their in-line placement with the RF signal, and the small size compared to the branchline coupler, makes me think they're supposed to suppress a specific harmonic frequency in the output. They look roughly 1/3 to 1/4 the size of the DC bias ones at the bottom right, so they're probably supposed to block the 3rd or 4th harmonic of the output frequency. The branchline couplers are much larger, but the lines in these don't have to be 1/4 wavelength: they can be made any odd multiple of 1/4 wavelength, whatever makes the layout easiest (with some bandwidth penalties I think).
A temperature sensor lives next to one of the output-stage transistors, to monitor their temperature. I'm not sure exactly where on the other boards this is monitored.
(https://live.staticflickr.com/65535/53838119705_4996a4a2f8_z.jpg) (https://flic.kr/p/2q2uric)
One thing I didn't have any luck with, was finding out how the power transistor biasing was done. Most of the time you can't just put a constant DC voltage or current on the gate or base, as the threshold voltage or beta will change wildly with temperature and send the DC operating point into all kinds of undesirable (too-low & too-high) regions. The input bias connections on this board just went straight through filters to off-board wires, and none of the other boards had circuitry that could be doing collector-current-based bias control or anything similar. Then again, a lot of these types of amplifiers I think are full-on ICs these days, rather than just a plain transistor, and so could easily have their own bias-control circuitry on-chip (this is not my field at all so please correct me if I'm wrong).
Power Board
The last board here is the under-appreciated power supply, which creates all the low supply voltages (+10V, +/- 5V, +3.3V, +2.5V, +1.5V) for the other boards, generated from an external 28VDC input bus.
(https://live.staticflickr.com/65535/53837929018_2136496278_z.jpg) (https://flic.kr/p/2q2tsBu)(https://live.staticflickr.com/65535/53837928968_9500148e88_z.jpg) (https://flic.kr/p/2q2tsAC)
Most of this board is off-the-shelf DC-DC converter and EMI filter modules; we can look inside those later.
(https://live.staticflickr.com/65535/53837673241_62f00a283c_z.jpg) (https://flic.kr/p/2q2s9zx)
There's also some linear regulators, which create the lower voltages:
(https://live.staticflickr.com/65535/53837929028_4673658072_z.jpg) (https://flic.kr/p/2q2tsBE)
Here's what's going on circuitry-wise on this board, with details omitted where they don't add anything:
[attachimg=2]
The interesting bits on this board are...
1. At the output of the +10V supply: really a +/-5V supply, with the -5V referred to ground. I'm guessing the +10V is used for the RFPA.
(https://live.staticflickr.com/65535/53838026334_36ae44e261_z.jpg) (https://flic.kr/p/2q2tXxm)
These components are in the "input power present indication" box on the schematic. A TL431 and optocoupler produce a 2.5V digital output which is high when the input power is above a certain voltage threshold: this is probably sent to the FPGA, and used to tell when the input voltage is starting to sag for some reason.
2. Behind the linear regulators:
(https://live.staticflickr.com/65535/53837673316_549943e060_z.jpg) (https://flic.kr/p/2q2s9AQ)
The two ICs here are a quad op-amp, and a quad AND gate. An op-amp and two AND gates are used, as described on the schematic, to wait for the +3.3V supply to be present and also a "high" control signal from the FPGA, to activate the +10V supply. This further supports the idea that the +10V supply powers the RFPA.
The other 3 op-amps are used in an interesting way: each subtracts one voltage rail from another, and the results travel off-board via a set of discrete wires. My best guess here is that these wires go to an external connector on the box, and are used during testing. During environmental testing of a piece of space equipment, you'd want to be able to make sure that all the internal power supplies are working correctly even when the box is being baked or shaken, and so monitoring the internal supply voltages would be important. Looking at linear combinations of the supply voltages lets you verify this with fewer wires, though, than if you just ran every internal supply voltage to the connector: the (+2.5V) - (+1.5V) output, for example, (TP27) will always read about +1V if both supplies are working correctly. If the measurement is significantly different than +1V, it doesn't tell you whether the 2.5V or the 1.5V supply is failing, but you know that something is wrong.
With the SOIC packages above, you can see the same type of "dual prototyping & flight" component footprints that I mentioned before with the DAC, on the digital board. Notice how there are standard plastic SOICs populated, the same as I'd buy from Digi-Key - but the pads extend much further, with a little bit of soldermask dividing the two sections of each pad. This is likely so that rad-hard flight-qualified ceramic flatpack versions of these same ICs, which are much larger, can be populated on non-prototype versions of the boards. If you go back and look at the two optoisolators on the power board, you can see a similar thing: next to each plastic optoisolator is an unused slightly strange-shaped footprint, meant for a space-grade optoisolator in a package like this one (https://www.digikey.com/en/products/detail/tt-electronics-optek-technology/4N24U/2103264).
Anyways, that was a whole lot of text - hope that was an interesting look inside some real satellite hardware.
-
A few extra things about the Space Micro boards:
Coax connectors: There were some extra little spring-loaded adapters on the Modulator board coax connectors, that had a lot of "play"; I don't know if this is the equivalent of a helical shaft coupling (https://en.wikipedia.org/wiki/Coupling#/media/File:Beam_Coupling_-_Helical_Standard.jpg) but for coax, to keep vibration and shock from being transmitted to the soldered connector? I can imagine on a launch vehicle, even a short length of well-secured cable could pick up a good deal of inertia.
(https://live.staticflickr.com/65535/53837667936_9b28cdb0d1_z.jpg) (https://flic.kr/p/2q2s815)
Power amp input stages: First IC in the RF PA, the HMC346, is actually a controllable attenuator. This is probably used to set the transmit power.
Missing board: I feel like there's probably a missing board, because the RF PA leaves a lot of loose ends. Besides the biasing circuitry I mentioned earlier which it may or may not need, with all the "introspection" & engineering telemetry normally done on spacecraft equipment I'd expect there to be current-draw measurements for the RF PA's power power supply. There's also no low-speed ADC for measuring things like the temperature sensor on the RF PA, or the output of the possible-power-detector in its output stage. It's possible that the temperature sensor is a (solid-state) switch with an "ok vs. too hot" threshold output, and that the power detector output connects to the comparator on the digital board to give a simple "power vs. no power" indication...but it seems less likely. There's also no low-speed analog output to set the HMC346 attenuator on the RF PA.
DC-DC and EMI filter modules: Let's look inside those modules on the power board...
(https://live.staticflickr.com/65535/53840307090_73457796db_z.jpg) (https://flic.kr/p/2q2FDwJ)
The EMI filters are straightforward: each one has, in order...
- A common-mode choke
- Capacitance
- A strictly-differential-mode series inductor
- More capacitance
- Capacitance with a series resistor, to damp the resonance with the series inductor
- Capacitance from each output to the metal enclosure
(https://live.staticflickr.com/65535/53838973782_a261e33c62_z.jpg) (https://flic.kr/p/2q2yPbE)
(https://live.staticflickr.com/65535/53839867271_bf21681f5c_z.jpg) (https://flic.kr/p/2q2DoMD)
There's a large DC-DC (which generates the +10V rail), and a smaller DC-DC (which generates the +/- 5V used for everything else). The large DC-DC is almost identical in overall topology and control scheme to the other aerospace DC-DC module I opened already; go look at that one for details on the circuitry (https://www.eevblog.com/forum/renewable-energy/looking-inside-an-aerospace-dc-dc-converter-module/). The only high-level difference is just that these have bipolar outputs, and so use the split secondary winding differently with an extra pair of diodes - except for that, they've got the same single-transistor forward converter topology, the same inductively-isolated control section which is likely sending a peak current setpoint from output to input side, and the same current transformer on the input.
This is the large DC-DC:
(https://live.staticflickr.com/65535/53840215999_7e1d716b47_z.jpg) (https://flic.kr/p/2q2Fbsc)
(https://live.staticflickr.com/65535/53839864421_2a68eef7d9_z.jpg) (https://flic.kr/p/2q2DnWv)
(https://live.staticflickr.com/65535/53840124353_cab727ddbe_z.jpg) (https://flic.kr/p/2q2EHd6)
...and this is the small DC-DC: one unique part about it is the output common-mode choke at the bottom-right. It's still very similar to the other DC-DCs, although with the lower power level, the lack of output buck inductor, and only a single power diode per output, I think it's a flyback rather than a forward converter. (Could be doing something weird like operating discontinuous and using some intentionally-high transformer leakage inductance as the buck inductor, but that feels a whole lot less likely)
(https://live.staticflickr.com/65535/53840309265_ac36a3989c_z.jpg) (https://flic.kr/p/2q2FEbe)
(https://live.staticflickr.com/65535/53838972722_0f6e625a8f_z.jpg) (https://flic.kr/p/2q2yNSo)
(https://live.staticflickr.com/65535/53839866251_110578e825_z.jpg) (https://flic.kr/p/2q2Dou4)
(https://live.staticflickr.com/65535/53840310555_f36aabe4ab_z.jpg) (https://flic.kr/p/2q2FEyt)
-
Very interesting again! It looks like (part of) the bias circuitry was in the lower left corner of the PA board, since both inputs of the final stage and the temperature sensor seems to connect there. However, what ever lived there has been brutally removed, even taking some traces off the board.
-
Very interesting again! It looks like (part of) the bias circuitry was in the lower left corner of the PA board, since both inputs of the final stage and the temperature sensor seems to connect there. However, what ever lived there has been brutally removed, even taking some traces off the board.
Good spotting: it's hard to see from these photos, but that area (besides the couple diodes & caps) has mostly square pads with solder on them, which look very much like they had discrete wire connections off-board. As you say, looks like someone pulled hard on those temperature sensor wires rather than snipping them. So that's why I'm guessing there was some biasing happening on an additional missing board, where these wires connected. I looked at the discrete-wire connection points on both the digital board and the power board, but neither of them had any connections that would make sense for the temperature sensor.
-
I'm not sure exactly what this is storing; can't imagine it's the FPGA configuration but I don't see a dedicated EEPROM for that,
The Actel FPGAs are Flash-based (non-volatile configuration), so they do not require any external storage or initialization.
-
Note that case surfaces are coated with conductive yellow metal (gold solution?) and most walls/surfaces have RF absorbers.
That's likely a hexavalent chromium conversion coating (aka 'alodine') - makes the aluminum conductive. Otherwise, it passivates by forming an oxide layer that is insulating.
Coax connectors: There were some extra little spring-loaded adapters on the Modulator board coax connectors, that had a lot of "play"; I don't know if this is the equivalent of a helical shaft coupling (https://en.wikipedia.org/wiki/Coupling#/media/File:Beam_Coupling_-_Helical_Standard.jpg) but for coax, to keep vibration and shock from being transmitted to the soldered connector? I can imagine on a launch vehicle, even a short length of well-secured cable could pick up a good deal of inertia.
(https://live.staticflickr.com/65535/53837667936_9b28cdb0d1_z.jpg) (https://flic.kr/p/2q2s815)
These look like SMP or SMPM connectors with spring loading. The play is deliberate, to allow misalignment between the module and whatever it connects to. You can find similar connectors on Digikey for backplane connections.
https://www.digikey.com/en/products/detail/amphenol-sv-microwave/1132-4114/16400285 (https://www.digikey.com/en/products/detail/amphenol-sv-microwave/1132-4114/16400285)
Thank you for the teardown pics, this is really cool stuff!
-
Plus the BeO marking on the linear regulators, must be nice hybrid package there is they used that, and worth opening as well. Likely all discrete IC's there, with some nice bonding wires joining them all together.
-
Plus the BeO marking on the linear regulators, must be nice hybrid package there is they used that, and worth opening as well. Likely all discrete IC's there, with some nice bonding wires joining them all together.
I think I recognize the part used (MSK5900). That is more than likely a handful of passives with two die; first would be the PNP output transistor and the other is the RH version of the LT1573. MS Kennedy did a lot of work with LTC devices in packaging up RH parts we put out.
-
I think I recognize the part used (MSK5900). That is more than likely a handful of passives with two die; first would be the PNP output transistor and the other is the RH version of the LT1573. MS Kennedy did a lot of work with LTC devices in packaging up RH parts we put out.
Yep that's right! Haven't gotten the die under a microscope yet so can't confirm 100% yet that it's LTC, but you sure can see the big-ass pass transistor.
(https://live.staticflickr.com/65535/53847031665_fd0f46d4d4_z.jpg) (https://flic.kr/p/2q3h7vF)
Similarly, took the lid off the "possibly two amplifiers/buffers" mystery device, and can see that it's a single die with a strange grid pattern. Looking forward to getting this one under the microscope too to see what's going on.
(https://live.staticflickr.com/65535/53846958464_90515f57a1_z.jpg) (https://flic.kr/p/2q3gJKA)
-
I think I recognize the part used (MSK5900). That is more than likely a handful of passives with two die; first would be the PNP output transistor and the other is the RH version of the LT1573. MS Kennedy did a lot of work with LTC devices in packaging up RH parts we put out.
Yep that's right! Haven't gotten the die under a microscope yet so can't confirm 100% yet that it's LTC, but you sure can see the big-ass pass transistor.
Yes, I recognize the die after having worked with it over the years. If you go to the SMD document at https://www.analog.com/media/en/technical-documentation/controlled-drawings/05-08-5223.pdf (https://www.analog.com/media/en/technical-documentation/controlled-drawings/05-08-5223.pdf) you can see the pad locations of the die are a simple 180 rotation of how it is mounted in the package. It looks like they did a laser trimming of the resistor in series with the base of the PNP to help get a consistent current limit as well.
-
Very cool! Yes I see exactly what you mean.
Tonight was the Night of the Free Microscope Time, so die shots for everyone. I got a nice up-close shot of that particular controller:
(https://live.staticflickr.com/65535/53848994121_ab290a044d_z.jpg) (https://flic.kr/p/2q3saT8)
Pass transistor in the linear regulator is less exciting, but I still like seeing that familiar winding pattern:
(https://live.staticflickr.com/65535/53849353639_aec3a0ea4b_z.jpg) (https://flic.kr/p/2q3u1KH)
Also check out the mystery device from the modulator board: it fits my circuit-functionality expectations by being a transistor array:
(https://live.staticflickr.com/65535/53848108577_c0ed8e4eb3_z.jpg) (https://flic.kr/p/2q3nCD8)
I wonder if the metal layer was customized for each application (like with many digital ASICs). First thought was that only the working transistors were connected, but that doesn't make sense as probing each one with an automatic tester, without any kind of bond pads, seems wildly unrealistic.
Edit: This is actually the HFA3127 "ultra-high frequency" transistor array (https://docs.rs-online.com/8edb/0900766b8002512d.pdf). The many unused transistors makes me wonder if all 4 products in that datasheet share the same die, so that all 20+ transistors are always on the die, but only the metallization changes based on the part number.
Can you do a close-up photo of glass encapsulated transistors?
Done - here's one of each type.
(https://live.staticflickr.com/65535/53849261848_e3d1b0c69e_z.jpg) (https://flic.kr/p/2q3txt7)
(https://live.staticflickr.com/65535/53849261843_fb93efd528_z.jpg) (https://flic.kr/p/2q3txt2)
(https://live.staticflickr.com/65535/53849439475_e00a6e484f_z.jpg) (https://flic.kr/p/2q3usgD)
(https://live.staticflickr.com/65535/53849000056_84b9b0cb9b_z.jpg) (https://flic.kr/p/2q3scDs)
(https://live.staticflickr.com/65535/53849000031_0fbe2d43c6_z.jpg) (https://flic.kr/p/2q3scD2)
(https://live.staticflickr.com/65535/53849359594_992d526118_z.jpg) (https://flic.kr/p/2q3u3wo)
The one-off transistor boosting the DRO output is especially nice to look at:
(https://live.staticflickr.com/65535/53849262083_08024c5a03_z.jpg) (https://flic.kr/p/2q3txxa)
(https://live.staticflickr.com/65535/53849000141_32e36f5cae_z.jpg) (https://flic.kr/p/2q3scEV)
-
Also, here's a few bare dies inside the large DC-DC module from the power supply board.
(https://live.staticflickr.com/65535/53849343319_cd66d3b7f7_z.jpg) (https://flic.kr/p/2q3tXFM)
(https://live.staticflickr.com/65535/53849343309_aa99cd296e_z.jpg) (https://flic.kr/p/2q3tXFB)
Mystery symmetrical STMicro part (rotate the right-hand side 180 degrees in your head to see it):
(https://live.staticflickr.com/65535/53848098522_c5ba8744d8_z.jpg) (https://flic.kr/p/2q3nzDL)
(https://live.staticflickr.com/65535/53849343289_974aa039ab_z.jpg) (https://flic.kr/p/2q3tXFg)
Texas Instruments LT10009 2.5V shunt voltage reference (https://www.ti.com/product/LT1009):
(https://live.staticflickr.com/65535/53849423270_a0d5719375_z.jpg) (https://flic.kr/p/2q3unsf)
Mystery mostly-symmetrical STMicro part that may or may not be a dual op-amp:
(https://live.staticflickr.com/65535/53849343269_152fa8bd4c_z.jpg) (https://flic.kr/p/2q3tXEV)
Another Texas Instruments LT10009 2.5V shunt voltage reference (https://www.ti.com/product/LT1009): probably one for input side, one for output side.
(https://live.staticflickr.com/65535/53848983886_d261ae4637_z.jpg) (https://flic.kr/p/2q3s7QE)
Gate driver transistors:
(https://live.staticflickr.com/65535/53848983796_76fb255639_z.jpg) (https://flic.kr/p/2q3s7P7)
(https://live.staticflickr.com/65535/53849423215_065a2316e4_z.jpg) (https://flic.kr/p/2q3unri)
One of the two paralleled primary-side power MOSFETs - when you get close enough, you can see the hexagonal cells (it's essentially a few hundred tiny MOSFETs in parallel). This is where the (IR) HEXFET series name comes from, if I remember correctly:
(https://live.staticflickr.com/65535/53849423225_bc6020d5d2_z.jpg) (https://flic.kr/p/2q3unrt)
(https://live.staticflickr.com/65535/53849423220_702c64f8e8_z.jpg) (https://flic.kr/p/2q3unro)
-
That is a laser trimmed resistor.
-
Broad Reach I/O Board
I don't know much about this board's function besides that it's made by Broad Reach, and the latest date codes I can see are from 2009 (with some going back to 2002). Broad Reach, now part of Moog (https://www.moog.com/markets/space/spacecraft-avionics.html) (no, not the synthesizer company), seems to make general-purpose spacecraft & satellite electronics.
(https://live.staticflickr.com/65535/53865760894_b445e14141_z.jpg) (https://flic.kr/p/2q4W74q)
(https://live.staticflickr.com/65535/53865821085_a767201611_z.jpg) (https://flic.kr/p/2q4WpXc)
This looks like an I/O board, judging by what's on it - the blue connector at one end looks like it goes to a common backplane, while the small D-sub connector at the other end seems like external I/O.
The board isn't conformal coated, but with a lot of tightly-packed components and most of the traces being internal, there's a limit to how much I was willing to figure out with continuity checks alone. So sticking to a sanity-retaining number of strategically-placed continuity checking and applying a lot of educated guessing, this is my best guess at how things are connected.
[attachimg=1]
Unlabeled ICs are all Linear Technology RH1014 (https://www.analog.com/en/products/rh1014m.html) quad op-amps, except the gold package which is an Intersil OP470 (https://www.renesas.com/us/en/document/dst/hs-op470arh-hs-op470aeh-datasheet) quad op-amp.
1. Power Supply: An isolated power supply, based on the classic UCx845 PWM controller series, creates the logic & analog supply voltages. Input is probably from a 28VDC bus. The metal box on the bottom side is the primary-side power MOSFET. The vertical cylinder is the bobbin & core of a flyback inductor. An ST RHFL4913 (https://www.st.com/en/space-products/rhfl4913a.html) LDO on the bottom side probably provides a low FPGA core voltage or something like that.
2. Analog Outputs: 2x identical channels, each with an NPN Darlington & PNP Darlington pair for some class-B analog-output action; all in metal cans. NPNs are 2N5667, PNPs are 2N5416. Diodes on the bottom offset the NPN & PNP bases (with a substantial voltage gap for no idle current), and the 5th metal-can transistor per channel creates a current source for biasing. Because of the inherent inaccuracy of a straight class-B emitter-follower, esp. with the large NPN/PNP conduction gap, I assume some of the op-amps wrap feedback loops around these. There's current limiting and common biasing involving some smaller SMT transistors.
3. Push-Pull Digital Outputs: 3x identical channels, each with an IRHNM57110 N-MOSFET & IRHNM597110 P-MOSFET with their drains connected to each other, and to an output pin. Outputs are clamped to ground & supply by diodes on the bottom and top. Gate drive & level-shifting seems to be done by small transistors on top side. The last un-paired P-MOSFET at one end gates the positive supply voltage to all the channels' P-FETs, as a global power enable.
4. Open-Drain Digital Outputs: 2x identical channels, each with an IRHNM57110 N-MOSFET on the bottom side (with its drain connected to an I/O pin) and gate drive circuitry on the top side.
5. Current Sense: 6x separate current sense channels, with Kelvin-connected current sense resistors & op-amps to differential-amplify the current sense voltages; these seem to measure the current draw on various supply voltages, probably ones that get used for driving outputs to monitor total output current. These connect to analog mux inputs (discussed next) and so these current draw readings are probably read by the ADCs along with all the other analog input values.
6 & 7. Mystery: I have no idea what these do. #6 has 2 NPN power transistors on the bottom, with 7 smaller transistors on the top. #7 has a single NPN power transistor & 2 optoisolators, the Mii 66183 (https://www.datasheetcatalog.com/datasheets_pdf/6/6/1/8/66183.shtml). The top-side 2N5339 only has connections to the input-power bus and has a zener to its base, so may be for something like fast discharge on power-down, or crowbarring the power supply's input voltage.
The FPGA, an Actel part, likely provides an interface between a processor bus from a controller card elsewhere, and all the peripherals here.
(https://live.staticflickr.com/65535/53865823680_1b566b8524_z.jpg) (https://flic.kr/p/2q4WqHW)
Besides the digital outputs and various controls, the main peripherals here are 3 DACs (the 12-bit Analog Devices AD667S (https://www.analog.com/en/products/ad667s.html)) and 3 ADCs (the 16-bit Maxwell, now DDC, 7809LP (https://www.alldatasheet.com/datasheet-pdf/pdf/116799/MAXWELL/7809LPRPFK.html)).
DACs: 2 of these likely create the setpoints for the analog output drivers (#2 in the list above). The 3rd may drive a general-purpose low-power analog output.
ADCs: 3x Analog Devices MUX-16 (https://www.analog.com/en/products/mux16s.html) 16:1 muxes seem to feed up to 48 analog channels to the ADCs, such as this one on the bottom side with a thermistor attached to it:
(https://live.staticflickr.com/65535/53865413596_22345e606b_z.jpg) (https://flic.kr/p/2q4UjPw)
Some of the muxed ADC channels are definitely used for power-supply current sensing (#5 in the list above), and the rest are probably for general-purpose analog inputs from the outside world.
Reference voltage: The "pattern-breaking" metal can near the analog outputs is a Linear Technology LT1021 (https://www.analog.com/en/products/lt1021.html) 5V voltage reference; I'm guessing this creates a common reference voltage for the ADCs & DACs.
The ADCs themselves are the ones in the interesting extra-thick packages:
(https://live.staticflickr.com/65535/53865409971_d3d4fde482_z.jpg) (https://flic.kr/p/2q4UiK2)
From reading the 7809LP datasheet, it looks like the radiation-induced latchup protection is implemented by using a commercial ADC die, with a rad-hard power control circuit wrapped around it to detect fault conditions and cycle its power. I considered opening one of the packages to see this hybrid construction inside, but decided to read up on the Rad-Pak packaging first, covered by US Patent # 6455864 (https://patents.google.com/patent/US6455864B1/en), and realized it would probably be full of goop and therefore not worth it. The core of the Rad-Pak shielding method, if I'm reading the patent correctly, is a conformal coating consisting of tungsten particles held together by an adhesive binder. The high density of tungsten helps block ionizing radiation, similar to lead shielding used with nuclear reactors, medical X-rays, etc.
Here's some better close-up views of...
The power supply:
(https://live.staticflickr.com/65535/53865827090_2b4bec2f1c_z.jpg) (https://flic.kr/p/2q4WrJJ)
(https://live.staticflickr.com/65535/53865660843_744d6a2ed2_z.jpg) (https://flic.kr/p/2q4VAjp)
...including the stack of ceramic caps, given a lead frame to trade vertical space for horizontal space, and to avoid some cracking issues created by mismatched thermal expansion between the ceramic & the PCB material. The ADCs have some pretty large SMT ceramic caps too, but not quite as large (horizontally) as these.
The external-I/O connector:
(https://live.staticflickr.com/65535/53865824565_dc2962b98b_z.jpg) (https://flic.kr/p/2q4WqZc)
(https://live.staticflickr.com/65535/53864504992_d71169d667_z.jpg) (https://flic.kr/p/2q4PEHW)
...with a metal-can LM117 adjustable regulator next to it (I haven't figured out where the LM117's output voltage goes to: not to the I/O connector)
Current-sense circuitry, where you can see the Kelvin-sense terminals on the series resistors:
(https://live.staticflickr.com/65535/53865762854_bc6b61624b_z.jpg) (https://flic.kr/p/2q4W7Dd)
(https://live.staticflickr.com/65535/53864503457_6d7b63a689_z.jpg) (https://flic.kr/p/2q4PEgt)
The LDO:
(https://live.staticflickr.com/65535/53865655173_43334e53a6_z.jpg) (https://flic.kr/p/2q4VyCD)
Misc. things on the bottom side:
(https://live.staticflickr.com/65535/53865406651_1ef5f600f9_z.jpg) (https://flic.kr/p/2q4UhKM)
(https://live.staticflickr.com/65535/53865829880_f37f1dd81a_z.jpg) (https://flic.kr/p/2q4WsyQ)
(https://live.staticflickr.com/65535/53865663573_0af873f4e4_z.jpg) (https://flic.kr/p/2q4VB8t)
(https://live.staticflickr.com/65535/53864510387_716a4a8203_z.jpg) (https://flic.kr/p/2q4PGjX)
(https://live.staticflickr.com/65535/53865829875_e69f7d5e14_z.jpg) (https://flic.kr/p/2q4WsyK)
You can see the impressive amount and quality of reworks applied, in some of these photos, which makes me wonder if this was a prototype of some kind. There was also an impressive amount of polyimide tape applied to every single wire and "floating" component, to hold them all securely in place - I removed most of the tape though to see the board (& reworks) better.
Anyways, hope this was interesting - wish I could provide a schematic for this one.
-
Mystery symmetrical STMicro part (rotate the right-hand side 180 degrees in your head to see it):
(https://live.staticflickr.com/65535/53848098522_c5ba8744d8_z.jpg) (https://flic.kr/p/2q3nzDL)
Wonder if LM319. Anyone got a comparison pic?
Tim
-
Oh shit. Good call! (https://zeptobars.com/en/read/LM319M-comparator)
-
Meant to mention, in the context of the ADCs:
Ionizing radiation, which there's a lot of in space in the form of cosmic rays etc., causes havoc with electronics and ICs in particular by ionizing things that shouldn't be ionized, creating carriers in what's supposed to be an insulator, and therefore creating leakage currents where they're not supposed to be. This means that sometimes you'll get unexpected spikes of current - but also has a secondary effect called "latchup", which essentially crowbars the IC's power supply when this leakage happens to turn on a parasitic transistor that's inherent to how the transistors are constructed on a shared conductive silicon substrate. The high current draw resulting from a supply-to-ground short then generates a ton of heat and destroys the IC, if you're not careful. Radiation damage also accumulates over time, permanently adding leakage, but that's a separate issue.
The ways to deal with all radiation-induced effects are:
Suck it up: With a cheap cubesat that's only expected to work for a couple days, this is a common option. Use normal commercial parts, it's fine, hope you don't get unlucky.
Shielding: Add dense materials, such as steel, lead, or tungsten to block/attenuate the ionizing radiation. Nobody enjoys doing this on space equipment though, considering how expensive per pound it is to put something in space.
The normal way to deal with the transient current-spike events (especially in digital circuits, memory in particular) known as "single-event upsets" is:
Circuit design: Add redundancy (to logic circuits) & error-check bits (to memory; really just another form of redundancy). The cross section of a cosmic ray is pretty small and so it's not going to affect half your IC at once - the unaffected majority of the circuitry can correct for the affected minority.
The normal way to deal with latchup in particular is:
IC process: Get rid of the shared silicon substrate that enables latchup in the first place, and replace it with an insulator, in a Silicon-on-Insulator (SOI) process (https://en.wikipedia.org/wiki/Silicon_on_insulator). Sapphire is a popular option; this is expensive.
I thought it was interesting though that the rad-hard ADC here explicitly takes a "tolerate and reset" approach to latchup, quickly shutting off the power before the high supply current can do any damage, and resetting the chip. It makes a lot of sense for an ADC - this approach wouldn't work for something like a processor, where there's a whole lot of "state" associated with it (the contents of all the registers, progress of the instruction currently being executed, etc.) that can't be recovered if it's randomly reset. With an ADC though, you're likely taking continuous readings anyways, and so the worst that can happen is that you miss one measurement due to a latchup-induced reset, and have to wait until the next reading cycle to get that measurement. If you can plan your circuit & system design around that, then that's an easy constraint to deal with! This means it's easier if all the ADC settings are done with voltages on external pins, rather than having a serial interface (SPI or I2C) to set up internal settings registers, but that's a small price to pay for not having to use an expensive special-purpose SOI part.
-
FYI, SoS used to be used, but oxygen ion implantation is the common method today AFAIK.
Might well still be using it for space applications, heh, I have no idea.
Tim
-
There is a whole section of the military electronics business concerned with ensuring that the in flight energy in all the inductors and capacitors in a system is incapable of damaging any silicon that goes into heavy conduction due to EMP. If it isn't actually damaged, it might do unfortunate things in the moment, but Microsoft knows how to recover the situation - turn it off and turn it on again.
-
Interesting, bet that drives a lot of power supply design decisions! I'd always thought of the (now-long-passed) move from voltage-mode to current-mode control of power converters as a less directly visible thing from the system level, but can see how that would be a huge help for that criteria: go with minimal energy storage directly on the power rails, have fast-responding peak-current-controlled buck(s), and lump all the bulk capacitance on the buck's input where it's safely behind that cycle-by-cycle current limit.
FYI, SoS used to be used, but oxygen ion implantation is the common method today AFAIK.
Huh cool, my knowledge is not particularly up-to-date - for anyone else who doesn't know oxygen ion implantation (SIMOX) already, looks like they implant oxygen ions in the surface of a wafer to create an insulating SiO2 layer (sand, essentially) and then grow an epitaxial silicon layer on top of that to use for the transistors.
-
These modules are also wrapped in several layers of shielding (the craft, perhaps; outer module, PCB, inner module), so we can quite safely assume any surge is conducted in on device pins alone.
Current mode also minimizes component size by decoupling the LC resonance; DCM is perfectly reasonable, and the control pole can be above 1/sqrt(LC). Quite handy here, as well as the inherent current limiting, which can be rolled into a limit detect and fault signal, or automatic (hiccup mode, etc.), further avoiding damage.
SIMOX is even cleverer than that: the trick with implantation is, ions are deposited at a range of depths, depending on ion mass and energy. Si can be cleaved into very thin wafers by implanting H+ (annealing --> coalesces into a boundary layer of Si-H bonds and free gas --> tape and peel), or basically the same thing but bonded by SiO2 still (this way). The top layer is crystal-aligned with the substrate and subsequent epitaxy, diffusion, etc. will remain coherent with it. Thus very small transistors (in terms of depth as well as width and height) can be made, without any yield loss due to random crystal orientation.
Tim
-
Meant to mention, in the context of the ADCs:
Ionizing radiation, which there's a lot of in space in the form of cosmic rays etc., causes havoc with electronics and ICs in particular by ionizing things that shouldn't be ionized, creating carriers in what's supposed to be an insulator, and therefore creating leakage currents where they're not supposed to be. This means that sometimes you'll get unexpected spikes of current - but also has a secondary effect called "latchup", which essentially crowbars the IC's power supply when this leakage happens to turn on a parasitic transistor that's inherent to how the transistors are constructed on a shared conductive silicon substrate. The high current draw resulting from a supply-to-ground short then generates a ton of heat and destroys the IC, if you're not careful. Radiation damage also accumulates over time, permanently adding leakage, but that's a separate issue.
The ways to deal with all radiation-induced effects are:
Suck it up: With a cheap cubesat that's only expected to work for a couple days, this is a common option. Use normal commercial parts, it's fine, hope you don't get unlucky.
Shielding: Add dense materials, such as steel, lead, or tungsten to block/attenuate the ionizing radiation. Nobody enjoys doing this on space equipment though, considering how expensive per pound it is to put something in space.
The normal way to deal with the transient current-spike events (especially in digital circuits, memory in particular) known as "single-event upsets" is:
Circuit design: Add redundancy (to logic circuits) & error-check bits (to memory; really just another form of redundancy). The cross section of a cosmic ray is pretty small and so it's not going to affect half your IC at once - the unaffected majority of the circuitry can correct for the affected minority.
The normal way to deal with latchup in particular is:
IC process: Get rid of the shared silicon substrate that enables latchup in the first place, and replace it with an insulator, in a Silicon-on-Insulator (SOI) process (https://en.wikipedia.org/wiki/Silicon_on_insulator). Sapphire is a popular option; this is expensive.
I thought it was interesting though that the rad-hard ADC here explicitly takes a "tolerate and reset" approach to latchup, quickly shutting off the power before the high supply current can do any damage, and resetting the chip. It makes a lot of sense for an ADC - this approach wouldn't work for something like a processor, where there's a whole lot of "state" associated with it (the contents of all the registers, progress of the instruction currently being executed, etc.) that can't be recovered if it's randomly reset. With an ADC though, you're likely taking continuous readings anyways, and so the worst that can happen is that you miss one measurement due to a latchup-induced reset, and have to wait until the next reading cycle to get that measurement. If you can plan your circuit & system design around that, then that's an easy constraint to deal with! This means it's easier if all the ADC settings are done with voltages on external pins, rather than having a serial interface (SPI or I2C) to set up internal settings registers, but that's a small price to pay for not having to use an expensive special-purpose SOI part.
Fortunately when designing chips for spacecraft a lot of these effects are known. Ionizing radiation tends to shift MOSFET thresholds lower (charge buildup in the oxide) meaning NMOS devices are easier to turn on and PMOS devices take more voltage to turn on. It also can do crystal damage over time, causing leakage currents that can change biasing or other operating conditions in an IC.
There's a whole slew of effects that occur from a number of different things happening. Ionizing radiation is actually one of the easier things to deal with, whereas Single Event Effects (SEE) items are one of the tougher things to design for. Redundancy is great since, as you pointed out, you're not likely to get multiple heavy ion/cosmic rays hitting your chip at once. It's easy to do in digital, not so much in analog. Process of course can help with a number of things as well, such as avoiding gate rupture on large MOS devices.
There's a lot of things to think about when designing for these environments, and very few IC designers that understand them enough to develop devices for them.
-
SIMOX is even cleverer than that: the trick with implantation is, ions are deposited at a range of depths, depending on ion mass and energy. Si can be cleaved into very thin wafers by implanting H+ (annealing --> coalesces into a boundary layer of Si-H bonds and free gas --> tape and peel), or basically the same thing but bonded by SiO2 still (this way). The top layer is crystal-aligned with the substrate and subsequent epitaxy, diffusion, etc. will remain coherent with it. Thus very small transistors (in terms of depth as well as width and height) can be made, without any yield loss due to random crystal orientation.
Now that's a good trick!
And yes with the voltage-mode control, I distinctly remember doing my first voltage-mode buck compensation (as an exercise, learning how to use a TI DSP-MCU dev kit), seeing the giant Lbuck & Cout resonant peak, and getting a sinking feeling as I realized I was going to have to give my loop a super-slow crossover just to squash that - one of those "how did people get anything done before this?" moments.
-
Here's a much older piece of spacecraft equipment:
Hughes Traveling Wave Tube & HV Power Supply
If you're not familiar with Traveling Wave Tubes (TWTs) (https://en.wikipedia.org/wiki/Traveling-wave_tube), these are vaccuum-tube RF amplifiers which work by modulating a beam of electrons across a multi-wavelength distance (similar to a klystron). Before microwave-frequency transistors were available and reliable (esp. in 10W+ or 100W+ power levels), these were the main way to build a multi-Ghz RF amplifier. Accelerating the electron beam requires a high (multi-kV) DC voltage, and generating the free electrons in the vacuum requires a low-voltage filament - generating both these voltages, as well as health monitoring etc. is the job of the attached power supply (called the "Electronic Power Conditioner" or EPC in the TWT world).
(https://live.staticflickr.com/65535/54069714500_24ab674135_z.jpg) (https://flic.kr/p/2qnXqmU)
According to the seller (who has another of these, and was kind enough to give me a good discount) (https://www.ebay.com/itm/383924819373), this particular unit came from Intelsat (https://en.wikipedia.org/wiki/List_of_Intelsat_satellites), and was originally part of a life test rack - normally this is where production electronics are exercised continuously, sometimes under elevated temperatures or other rough conditions, to check for early failures. The Intelsat V family of communications satellites were built by Ford Aerospace: it seems strange for a car company to have a satellite business, but at some point in the early 60's they acquired wide-ranging electronics company Philco (https://en.wikipedia.org/wiki/Philco) and aerospace company Aeronutronic (https://en.wikipedia.org/wiki/Aeronutronic), to form their military & aerospace business. This aerospace division produced several families of communications satellites and various military projects, but was later sold off to become "Space Systems/Loral" and then Lockheed Martin, all part of the big, dumb game of "military-industrial-complex trading cards".
The TWT & power supply were built by Hughes Aircraft, though, who had their own thriving communications satellite business (the first entry in this thread came from them as well).
There's not much to see with the TWT itself, esp. as the label warns about the beryllium oxide inside. Combining high thermal loads (esp. at the "collector" in a TWT which dissipates the leftover energy in the electron beam after coupling to the RF output) with high voltages means that this is a prime place for very thermally-conductive but electrically-insulating ceramics.
There's also a little RF module which the input(?) RF signal passes through, before reaching the TWT. No idea what this is, as I wasn't able to get it open, but it's 2-port & passive.
(https://live.staticflickr.com/65535/54069714860_02d340b836.jpg) (https://flic.kr/p/2qnXqt7)(https://live.staticflickr.com/65535/54069511938_e8727822f2.jpg) (https://flic.kr/p/2qnWo9s)
So, let's look inside the power supply.
Power Supply Board
(https://live.staticflickr.com/65535/54069257746_0c33bf0a8b_c.jpg) (https://flic.kr/p/2qnV5zQ)
(https://live.staticflickr.com/65535/54069511773_dfec4e48ef_c.jpg) (https://flic.kr/p/2qnWo6B)
The 10-20W RF output of the TWT means the power supply is probably rated for well under 100W, so there's nothing crazily high-power here. The box is a thin metal shell with the PCB attached at both top & bottom with angle brackets, for rigidity. After taking out a lot of screws, snipping the output wires to the TWT, and un-sticking a bit of adhesive, I was able to pull the PCB out of the box to get a better view:
(https://live.staticflickr.com/65535/54069512908_6fe66a18ae_c.jpg) (https://flic.kr/p/2qnWorb)
(https://live.staticflickr.com/65535/54068383582_56259835a1_c.jpg) (https://flic.kr/p/2qnQAJ3)
They really did not skimp on the conformal coating here! There's a very thick layer of yellow coating over both the bottom and the top sides of the board, which gives the components an interesting "underwater" look because of the refraction. The orange and yellow of the board nicely matches the fall colors of the leaves where I took the photos outside my apartment though.
At the right-hand side, there's a 15-pin D-sub connector for interfacing to the outside world, plus 3 power transistors in interesting packages (with Texas Instruments logos).
(https://live.staticflickr.com/65535/54069518208_31d2514a33_z.jpg) (https://flic.kr/p/2qnWq1y)
(https://live.staticflickr.com/65535/54068388907_4f26d57b1f_z.jpg) (https://flic.kr/p/2qnQCiR)
(https://live.staticflickr.com/65535/54069519613_1c95541c3a_z.jpg) (https://flic.kr/p/2qnWqqM)
I can't connect the part numbers on these unusual power transistors to anything publicly accessible, but the '72 TI power device catalog (https://bitsavers.org/components/ti/_dataBooks/1972_TI_Power_Semiconductor_Data_Book_First_Edition.pdf) shows this as their standard "QQ" package (p.93). The 2N3263-3266 (p.192) seems like a good candidate for what parts these might be, as they're meant for "high-speed switching applications", which fits the switch-mode power supply application here. Rated power dissipation is 67W even at 100°C case temperature, which shows that the Rth-jc is nice and low at ~1.5 K/W, and that they can dissipate a lot of power if properly heatsinked. The only other place I've seen this type of transistor package is in the both fascinating-and-horrifying Minuteman III guidance system (https://airandspace.si.edu/collection-objects/guidance-system-minuteman-iii/nasm_A19770995000).
On the bottom, underneath this control and switching section, are three ICs:
(https://live.staticflickr.com/65535/54069264731_af43c5ced9_z.jpg) (https://flic.kr/p/2qnV7Eg)
(https://live.staticflickr.com/65535/54069721165_5b3837c129_z.jpg) (https://flic.kr/p/2qnXskP)
2 have the AMD logo and are labeled LH2823; I can't find any references to this part number, but from the pinout & connections (discussed later) these seem like op-amps, likely one of the AM102/AM112/AM216 series. The 1 remaining IC is also from AMD and has some kind of custom part number, but may be digital logic(?).
The most noticeable feature of the board is the big orange potted block in the middle. This likely has all the high-voltage stuff inside. You can see below...
- (top-right) a large toroidal transformer, probably for voltage step-up
- (top-middle) a dark-colored mystery block, which may contain high-voltage diodes and/or a voltage multiplier on the transformer output
- (top-left) a yellow-colored block of film capacitors, which probably provide the DC filtering for the high-voltage output(s)
- (bottom-right) a mystery cordwood module with some transistors
- (bottom-left) a small toroidal transformer
(https://live.staticflickr.com/65535/54069264826_d9aeee6b53_z.jpg) (https://flic.kr/p/2qnV7FU)
Here's a better view of the cordwood module and 2nd transformer:
(https://live.staticflickr.com/65535/54069519683_a5891908c9_z.jpg) (https://flic.kr/p/2qnWqrZ)
The first 3 items in the list seem pretty clear as to function, but I'm not sure what the cordwood module and 2nd transformer are for: these might be over-current protection in series with the HV output, or maybe regulation for the filament voltage.
At the other side of the block, you can see some wires exiting: these are 4 of the 6 wires that go to the TWT.
(https://live.staticflickr.com/65535/54069722800_97398f7ccf_z.jpg) (https://flic.kr/p/2qnXsQ1)
I'd like to get a better view of what's inside this block to figure out what the extra parts are doing, but unfortunately the orange material is too hard to cut or peel away easily without damaging the components inside. I'd also settle for being able to get a smoother finish on the top face, so that it's optically clearer like the side face shown before, but it turns out this material is also too soft to use sandpaper effectively.
Finally, at the left-hand/output side of the board, there's some additional circuitry that includes:
- another LH2823 op-amp on the bottom, for some kind of secondary-side control
- a mystery Caddock resistor network up against the orange block, possibly used for sensing one of the output voltages?
- a high-voltage film cap on the bottom
(https://live.staticflickr.com/65535/54068388097_335b865cd4_z.jpg) (https://flic.kr/p/2qnQC4T)
(https://live.staticflickr.com/65535/54068389587_289cf9a683_z.jpg) (https://flic.kr/p/2qnQCvz)
Connections
Now let's talk about how it works! At a minimum, the TWT needs a low filament voltage, a cathode voltage (usually shared with one side of the filament), and an anode/collector voltage - go read the Wikipedia article linked earlier to see what these do inside the TWT. The seller's description also mentions that this TWT has additional collectors (which each have their own different HV potential, if I remember correctly) for higher efficiency.
The power supply has 4 wires coming out of the potted HV block (2x yellow, brown, & red) and 2 wires coming off the PCB (blue & black). On the TWT side, there's 3.4Ω from brown to yellow (filament), and both yellow wires are connected together: this suggests the yellows are a combined filament & cathode common connection. Black connects to the TWT housing. This leaves the blue & red wires for collector voltages.
Since the collectors & anode have to be positive relative to the cathode & filament, and the majority of the power dissipation is in the collectors, if I were designing a TWT I would want to make the collectors as close to enclosure-ground potential as possible. This would give me an easier time conducting heat away from the collectors, with less insulation (and therefore thermal resistance) needed between the collectors and the enclosure. So my best guess is that the cathode & filament are at a high negative voltage, while the various collector voltages get closer and closer to 0V (relative to enclosure ground); the black wire may even be the final collector, at 0V. This would also explain why the filament wire (brown) comes directly out of the HV block: the voltage across the filament is low, but it has to float at a very high common-mode voltage.
Circuit Analysis
Unfortunately, I wasn't able to do as thorough of a job of reverse-engineering this one as I'd like. The unfriendly HV-block potting material didn't help, and there's 5 different pieces of magnetics on the board with many windings each: I made educated guesses on the schematic below at how the windings are arranged internally, but not being able to do continuity tests through the thick conformal coating, I can't be sure and so some of the transformer connections are still a mystery.
Let's look at the auxiliary power supplies first though: these are the low voltages to power the op-amps and other control circuitry.
[attachimg=1]
The DC power (likely 28V, by transistor & capacitor ratings) enters through an EMI filter, and charges the primary-side auxiliary supply ("+VauxPri") through a 220K resistor for startup. This auxiliary supply takes its input from the main power converter through a couple diodes, once that's running. There's also a nested switching arrangement, where aux. power to some devices is switched on and off separately ("+VauxPriSw"). One of the op-amps and a control-transformer winding are supplied through an additional switch ("+VauxPriSwSw").
One of the small transformers is part of a push-pull (https://en.wikipedia.org/wiki/Push%E2%80%93pull_converter) blocking oscillator (https://en.wikipedia.org/wiki/Blocking_oscillator), which uses two transistors in a self-oscillating configuration to generate an isolated bipolar power supply for the secondary-side/output-side op-amp. An additional winding seems to be tapped off and goes to a mystery location on the HV module, possibly to be used as a power supply for the cordwood module judging by location. One of the collector waveforms also gets filtered to power one of the op-amps.
The powertrain consists of another much larger push-pull converter, with those unusual-looking transistors, feeding the primary side of the HV transformer inside the potted HV module.
[attachimg=2]
This push-pull is current-fed, with a series inductor on the input, which gives it a boost-like characteristic, useful for stepping up the voltage on the primary side (and reducing the HV transformer's winding ratio). There's also an additional power supply that involves two of the metal-can transistors, powered from an additional winding on that boost inductor; I'm not sure what exactly they would feed inside the HV module: maybe a separate isolation transformer for the filament supply?
The control scheme is confusing, as it involves 4 separate transformers, none of which I can continuity-check as discussed before. The "small transformer (right-hand)" has its own self-oscillating converter to generate the auxiliary supplies, but the roles of the "small transformer (left-hand)" and the two series emitter transformers ("CT 1", "CT 2") are less clear. CT 1 & 2 probably provide some positive feedback to help switch the main power transistors on or off faster, but I'm not sure if they actually control the switching, because "small transformer (left-hand)" also has a charge pump which puts a negative voltage on the power transistor bases to switch them off. The "small transformer (left-hand)" also either is driven by or drives a separate connection to CT 1 & 2 (see the 2 series diodes), and these are also both connected to the self-oscillating aux. supply collectors via "Aux_C1" and "Aux_C2" - I don't know if the aux. supply actually generates the switching frequency for the whole system, or if this is just a startup mechanism and the main power supply self-oscillates via some combination of CT 1 & 2, and "small transformer (left-hand)".
One of the op-amps generates a control voltage that connects to CT 1 & 2, and likely somehow controls the duty cycle or frequency of switching. I don't see any reflected-transformer-voltage primary-side sensing of the output, so either the power supply runs open-loop (and the op-amp control voltage is for soft-start and/or protection), or there's some feedback path inside the HV module, maybe through the small toroidal transformer.
The secondary-side output circuitry was too thickly coated for me to be able to follow the connections, so all that I was able to tell is that the blue wire is some kind of high-voltage output from the HV module, fed through a 100KΩ+ series resistance and filtered with the 1kV(?) film cap on the bottom side of the board. This could be one of the intermediate TWT-collector voltages. I was completely unable to figure out how the black wire & the Caddock resistor network connect to anything, unfortunately.
Wrap-Up
Anyways, I hope this was interesting! Please let me know if you have any insights into how the surprisingly-complicated main powertrain's controls work.