Amen I was far more excited about the SDR approach combined with the Offline, turn it on equipment.

Couldn't agree more. The cheap RTL-SDR got me into the hobby RF world a few years ago, and I have since demodulated NOAA/GOES images with home built antennas, heard transmissions from New Zealand on Shortwave (I live on East Coast of the US), and many, many, many more cool things from around the world. My laptop is a cheap Dell Latitude and it works great with SDR# or any other open source SDR software. Buying a low-cost LNA and cheap FM filters go a long way as well with amplification and filtering, since most SDR's have a Bias Tee software enable option to power the LNA from the SDR itself.

yeah what happened to the sdr stuff??

What do you suggest as an alternative?

@@dh-_1011 Have you seen the thread on EEVBlog about how Keysight has stopped all sales and support to private (non-business) customers? Might be tied up in that.

@@guidoscalise The license is stored on the machine as part of the purchase so at boot the configuration is already fixed and there is no requirement to check. If you purchase a support agreement then checking for latest versions etc would seem sensible so there is always scope for ongoing revenue but this should be at the request of the owner.

@@campbellmorrison8540 That wasn't the case for my MSO-X 3024A scope. I had to manually enable each 'new' feature of my scope manually from documents that came with it. I suspect that the bulk of time that Marc was complaining about was system talking to the mother ship to license each function.

@@gregkrekelberg4632 Yes I agree and I now assume that was because it was a rental. Im pleased to hear you get to fix your licenses when you purchase the scope.

In the upper 5 if not already low 6 digit range for sure. Wanna see what makes it tick!

etesttool wants 91k for a n9020b mxa + i'm sure there are license fees to pay

the pile of analogue boxes they have WAS the equipment they used in the 60s

my E8285A takes ages to boot :(

I was thinking the same. There are still SCA broadcasts out there on the commercial FM band. I can receive a few with an SDR and soundcard configured as an SDR. There used to be an encrypted stock quote service as well, back in the 1980s.

I posed a question in the top thread about the purpose of the pseudorandom signal injection. You seem to have your head wrapped around the encoding schemes, so I wondered if you might be able to help. If you have a minute, would you mind taking a quick look and sharing your thoughts? (No obligation.) It starts with “ _Does anyone have any recommendations of where I could learn more about the pseudorandom signal injection…_ “

What are they trying to prevent you from? You need the hardware to make the software valuable. So strange.

When I feel very clever while listening to a teacher, it just means that the teacher is good 😊.

Not only that. Marc has the detailed knowledge of Apollo transmission technology AND has the talent to explain it to us in clear terms. Wow!

Honestly it's just to display the UI and easily interact with commercial software like LabView. Plenty of high-grade instruments will have second processor that runs some kind of RTOS for time critical stuff plus a lot of hardware computations in ASICs/FPGA and then dump the processed data over through PCI Express to the "user" portion.

There is 'vintage' HP kit which runs Windows -- the Infiniium scopes are the example that immediately springs to mind. From memory they run either Windows 2000 or XP, depending on how recent the unit is. It's really just to satisfy business interests in making the UI easy to develop (Windows application developers and tools to support them are comparatively common), and customer interests in making the UI familiar. Most of the grunt work will be done by the FPGAs and ASICs and sub-processors deep within the instrument.

@@philpem There's plenty of Infiniium scopes that ran Windows 98 but just loaded system and apps to ramdisk on boot so if you turn it off like classic oscilloscope it wouldn't break anything:)

I still prefer the way that we look for knob and buttons on vintage equipment than try to memorize multi level of manuals on the modern one which can get lost from time to time 😅

Was going to comment exactly this; it's good that the data subcarrier isn't an even multiple, but with the frequency spreading from the voice channel it's still overlapping pretty heavily. Wonder if this is how the original signal looked, or if it's a side effect of how the modern Keysight equipment is generating the signal.

Considering the SNR they would get with the power budget they probably had, I tend to believe the original gear would have a filter to reduce the harmonics. Or, maybe, a different power distribution between signals.

All done by the perfect Keysight equipment. This is not AM. That’s what PM is supposed to when you put a tone on it, you get the sine of a sine, which is a series. I suppose that’s also why they put the first sub is at 30kHz and the second at 70kHz, so the harmonics don’t overlap too much.

Vincent, not if it is running labview or matlab underneath the covers. I've used labview on a number of projects and it incredibly shortens development time and cost. It is very robust for screen indicators and controls and easily interfaced with custom hardware.

Running a Windows OS on HP/Agilent has been a thing for around 20 years now.. Marc's gear predates all of that nonsense. I mean... what's wrong with running Linux? :D

Because windows have standard GUI and quite good support in visual studio. Even I, as total beginner wrote some software to analyze XRF spectrum using entry level vb.net and free visual studio. Original XRF software for instrument can not do that job even after 3 years of requests. Instrument is Linux based and it is hard to connect to other lab software. But it is better compared to windows mobile based previous versions.

My guess; money. MS always tries to make deals and such. Full blown Windows is indeed a very weird choice here.

@@VincentGroenewold There's a good chance it is Windows Embedded. That is actually a pretty good choice for a lot of embedded gear. One advantage of Windows is that anyone that makes a hardware interface makes a Windows driver for it, for at least some version of Windows. Remember, this stuff doesn't connect to the Internet as a general rule, so you don't have to worry about nasty diseases floating around in the ether. It also probably has a fair amount of the normal Windows background trash turned off, and not even installed in the OS image. If it works when it is shipped, there is no reason it won't keep working as long as the hardware does.

@@VincentGroenewold more likely ease of drivers and framework and updates rather than anything else I think

It's because it allows to run Matlab, LabView and their comprehensive data demodulation analysis package live. The latter in particular is the big important one for 5G and other complex applications.

They indeed did send the voice and data. The data was via modems, I suppose the voice also via phone lines? I’m not sure. But each site also had their own local voice comms and their own equipment for generating the digital commands if they had to. They could relay the commands via teletype.

I guess it is a modified windows version. I remember some PLC stuff I worked with had modified (hardened) Win95 and WinCE on it. They had also more configuration possibilities, like USB ports read-only and so on. And you didn't install the OS like you do with normal desktop OS. You had a setup tool where you selected what you want and need, and then the windows was built and copied to a compact flash. That was about 20 years ago, I don't know how it works today. But yes, I personally also like an embedded linux more :)

@@ernstoud Linux has been used forever in this type of equipment and there, development tools are great as well (using it myself). It really doesn't make any sense here, as displayed by having to boot it up for 10 minutes. :)

You're not alone. There are other RTOS options employed by other hardware makers that are many orders of magnitude lighter on resources and faster, most likely greater stability too. QNX comes to mind. RTX is another, and one that I noticed was used as the OS in my Icom IC-705 transceiver. And, of course, linux with a custom built kernel only containing the extensions and modules required to support the hardware it's embedded in. I'm sure some suit at microsoft got a bonus for landing the keysight contract, probably after giving another suit at keysight some incentive or gifts.

Spot-on! I couldn't agree more!

@@ernstoud What! have you ever used Linux? What is so great about Virtual (Visual?) Studio. It's an IDE, there are many out there. If you need Visual Studio it is because Microsoft have forced you to work with their closed source languages.

Lol !

same here..... ha

Curious. Why so many use "we"?

If only there were some way to...

I believe the voice uplink was 300 Hz to 3kHz. On the downlink it was 300 Hz to 3000 Hz in the early missions, then they reduced it to 300 Hz to 2500 Hz because they had some issues with the EKG biomed data on the top subcarriers interfering (yes, there are subcarriers within the voice signal itself on the downlink!). In the CSM there was some further preprocessing including a 6 dB/octave pre-emphasis to emphasize the higher frequencies for intelligibility, and 12 dB "clipping", but I am not yet sure what that refers to (removal of noise ?). On the LM there was no pre-processing. Emergency voice had a more aggressive set of filtering and clipping, which the astronauts (Apollo 13) complained about, but I can't find the numbers at the moment. [Edit] Just found it, the emergency voice was 300 Hz to 2300 Hz. The downlink itself was much higher bandwidth, 300 to 13 kHz. The upper part of the spectrum was occupied by the biomed from the CSM during EVAs, or the retransmitted biomed from the LM.

@@CuriousMarc Thank you for the very elaborate answer, Marc! For me it is also interesting the bit shaping for the before they modulate the PSK. Is anywhere manual/schematics available?

@@cvetomircvetkov5670 Not that we have found yet.

See Episode10: kzfaq.info/get/bejne/q9yFpJtiz6_egJs.html. The hardware you mention might not be able to do it, but the entry level professional SDR sure will - with lots of programming.

* shudders * I prefer Linux and open source.

Even those AWGs run Windows. They run Win CE. I own one of those - it's got a Win CE license sticker on the back. Not that you'd ever know - you never see the desktop and you can't exit the AWG program. The scope he had (a 2k or 3k series) also runs Win CE. Most modern test equipment runs some sort of OS. Even Siglent equipment runs Linux.

FM decoding circuits are more complicated than AM for sure, and there are more than one of them. Two simple model examples would be: [Digital] Have 2 AM decoders running in parallel, one at each of the 2 frequencies ('1' and '0') and getting the same incoming signal. Treat the signal as "OOK" on-off keyed for both '1' and '0'. Only accept a '1' or '0' if one of the 2 decoders has actually decoded something, else there is no signal arriving. [Analog] An older decoder uses the "S-curve" method of decoding a signal by having a parallel "tank" LC circuit, tuned at the higher end of the FM signals frequency range. As the modulated FM signal rises in frequency the tank circuit resonates more strongly (higher amplitude across the tank ckt) and as the FM frequency lowers in frequency (to the bottom extreme of the frequency range) the tank circuit resonates much less. Take the waveform across the tank circuit (buffered) and use a diode/filter to get a changing DC voltage that represents the amplitude of the resonating tank circuit. That would be the FM demodulated signal. I recommend watching some older videos from w2aew's youtube channel as well. There re several very well explained FM demodulation techniques there and he explains it much better than I can in this comment. Cheers,

@@EngineeringVignettes Will do! And your response helped a lot, Thank you!

Hug granted!