Noise on VCO voltage ramp

Used U800 hybrids are available, along with a drop in replacement that you can assemble from information available online.

I've now uploaded 3 photos of what I'm seeing. The most obvious noise (apart from being obvious) is the 'comb' across the screen. I'm getting the most spurii at the lowest 'fundamental' (around 5Mhz) but as I sweep up in frequency the number of spurs declines markedly so by

The other form of noise I mentioned in an earlier post is the more attenuated and spread out one visible just to the right of the centre-frequency line. This noise, unlike the 'comb' is constantly flickering and random.

Not to mention ingominous failure! Well, the traces I'm seeing (and have now posted images of to the group) do bear a certain distinct similarity to what you described, Jeff, so if that be the case I'm wondering if one of the oscillators is being over-driven. I'm guessing that might give rise to 3rd harmonic spurii, don't you think?

replace the VCO control voltage with a known clean supply like from AA batteries and a pot so you can tune it.

Are the RF spurs still there or not?

Wide range VCO are exquisitely sensitive to signals on the control line.

Batteries you say? You would not even trust a made-in-USA linear power supply of 40 years of age to be clean?

Well, I have been meaning to do this test. There is provision for an external control voltage on the relevant board, and in view of the above I shall carry this check out next. If there are glitches on the VCO control voltages, I don't possess anything fast enough to see them, since my No.1 scope decided to blow up last week.

The wires will be antennas. The battery you can put inside the unit.

Cheers

Phil Hobbs

Good thinking, Phil. JL was right; this RF malarkey is the work of the Devil! :-D

Online RF seminar, Feb 5th, 2020:

Master the fundamentals of RF design and test, from the comfort of your desk.

Michael Terrell wrote in news: snipped-for-privacy@googlegroups.com:

The only decent post you've made in years.

Ummm... U800 is the horizontal output IC. Methinks he wants the U300 vertical amp hybrid instead (not sure and too lazy to download the manual again and check).

U300 vertical hybrid:

Also see the "replace everything except the hybrids" kit:

Also see: and for older scopes see:

Sorry for the late reply, but I had to waste the last few days dealing with the California Dept of Motor Vehicles to renew my drivers license and obtain a Real-ID endorsement. I passed both but am still recovering and might need a few more days to calm down.

Interesting spectrum analyzer display and description. The most interesting part is how you managed to produce the photo, without bothering to mention the:

1. Network analyzer test Frequency
2. SA (spectrum analyzer) horizontal MHz/div.
3. SA vertical full scale, dBm per division, and dBm full scale.
4. Whether you were misusing a scope probe (again) or feeding it with 50 ohms. Also, any attenuators or couplers in the RF path.
5. Did you remember to turn of the 1, 10, 50Mhz markers on the HP8754A?
6. Any other numbers which might be useful, especially those you previously mention, but which I'm too lazy to dig through your old postings to find.

Had you supplied some numbers, I would have dived into the HP8754A network analyzer service manual, and determined the frequencies of the two oscillators (3.6 - 4.3GHz minus 3.6 - 3.0GHz) used to produce your unspecified output frequency. I would then compare the possible intermodulation mixes of these two frequencies, with the mess on your spectrum analyzer screen, and see if any of the spurs are the result of oscillator feedthrough, intermod mixing, or overdriving the spectrum analyzer and creating the spurs in the SA. However, since your photos were not accompanied by any numbers, I can't do this for you. Incidentally, you might want to ask in one of the HP test equipment forums for what a working HP8754A RF output looks like on a spectrum analyzer. I would not be surprised if the mess on your screen is quite normal.

Incidentally, if you download and read the HP8754A service manual, you'll find a large number of Performance Tests. Performed properly in the manner specified, you should get a better idea if your network analyzer is functioning as expected.

Also, this is your second warning about providing insufficient information and numbers. Provide numbers with your problems or prepare thyself for the wrath of the radio gods. The recent failure of your Tek 2645A scope was a poorly aimed warning shot (I was aiming for your Tek 2232A scope) and a hint of things to come.

Make sure that the negative lead from the battery is connected _directly_ to the base of the varactor or other frequency defining element. Any common ground path between the varactor ground to other grounded circuitry will have some resistance and any dirty ground current will cause a voltage loss that is directly added to the control voltage and hence varies the frequency.

Use a low resistance potentiometer to reduce voltage noise in the control line. Add a capacitor to ground and an inductance to separate the bias supply from the RF (not a noisy resistor).

A 100 MHz/V control voltage sensitivity and a few microvolts of control voltage noise will make the phase noise spectrum quite dirty:-).

Did someone evoke the RF spirits, the Devil, and perhaps their present manifestation (me)? If so, kindly specify which JL you are referencing. Since I have not suggested that RF is "malarkey", it must be the other JL.

Incidentally, methinks you are misusing the term malarkey which means "insincere or foolish talk":

While RF is certainly the work the of the Devil, as no mere mortal could possibly contrive such a confusing tangle of fields, waves, SPICE models, and component pricing, I contend that there is little magic involved. The magic probably originated when an initial pilot run of (for example) ten RF related PCB's resulting in 10 radically different performance measurements. I've had it happen to me. Lacking a suitable explanation or understanding of the effect, the cause was most easily attributable to magic. Similarly, random, variable, and non-reproducible performance measurements were also attributed to magic. Since these are all quite normal phenomenon in RF, everything in RF soon became attributed to magic. Quoting Arthur C. Clarke "Any sufficiently advanced technology is indistinguishable from magic". In other words, to the clueless, uninitiated, and inexperienced, RF might as well be considered magic.

If it is powering anything else as well, the quality of the power supply isn't the only thing to worry about.

Apart from the VCO itself, obviously.

My sympathies. I've heard about the queues in those places and they must be pretty bad indeed to make the international news! Is it true they recruit retired-disabled ex-NYC taxi drivers for their counter staff?

Yeah, I know. I'm sorry but you know about as much as I do! If I'd zoomed out on those shots there's a bezel around the outside of the screen on that SA which is *supposed* to show all that info. Unfortunately, it's bust. This is a partly working unit I picked up at a ham rally and am in the process of re-commissioning. Unfortunately, my No.1 RF SA (the 8566B) is currently out of action with a recent "YIG oscillator unlocked" error message, so there's nothing to fall back on. It's kind of tough to do stuff with all this busted equipment. Even the equipment I picked up in prime working condition is letting me down lately. All I *can* say about those photos is the SA is set on the 0.01-1.8Ghz display range, I'm sniffing the VNA output with a few turns of wire rather than a direct connection, the 40dB attenuation is set (and seems to work!) and the VNA output frequencies in each picture are the filenames for each of those pictures. So the situation is "not ideal" but you have to work with what you've got, I guess.

I just noticed some points you mentioned I forgot to cover in my earlier post...

Shown in picture title for example 10Mhz.jpg means the VNA was set to

Full band mode so ~180Mhz/div.

No reliable info.

Neither - sniffing with few turns of wire.

Nope. Just the 40dB attenuation setting of the SA.

Yep, no markers.

Good suggestion.

I do actually have the original physical hard-copy of the service manual and was working my way through those performance tests when I got side-tracked by something else whilst waiting for the required phase-matched power splitter and some new terminations to arrive. I really MUST get back to that as they have since turned up in the post.

I also need to try out some other poster's suggestion re sweeping the VCO ramp with DC from batteries, too. So I have plenty to get on with whilst you digest my replies, Jeff. :-/

Sorry, Jeff; I was indeed referring to the Larkin. :)

Seems it's one of those words with somewhat different meanings in US and British English:

yeah, I was just kidding of course. "Elusive" might be a better description. :-D

Noted, thanks.

Not sure about the design aspect - a bit more info than necessary for me, but I shall certainly mark this for the testing PoV, thanks, Michael.