#22: Spectrum Analyzer Basics / Tutorial, and the Tektronix 1401A
Vložit
- čas přidán 23. 07. 2024
- I picked up a Tek 1401A Spectrum Analyzer at a hamfest recently, and it inspired me to do a video or two. This one presents a generic block diagram and tutorial of what a conventional swept spectrum analyzer is and how it works. It also shows a quick demo of the Tektronix 1401A Spectrum Analyzer Module from the mid-70's.
Video Notes Page:
www.qsl.net/w2aew/youtube/Basi... - Věda a technologie
I am a "newbie" to the RF world. I found this video to be extremely useful. You really broke down the "components" well and explained them in a very understandable way. Thank you very much for taking the time.
yeah me too. gave me a real radio boner. wide band.
As always, a wonderful, understandable explanation. Thank you, Alan!
Just bought my 1st spectrum analyzer and gained new knowledge from your presentation. Excellent as always! Thank you!
That must have been a pretty awesome hamfest! Thanks for breaking this down with your excellent explanations.
How fortuitous, your summary of superheterodyne receiver basics came up in my CZcams feed first . Three cheers for synchronicity and a thumbs up for yourself , thanks again.
Many thanks Alan for your continued educational efforts to share electronics knowledge! As I write, pondering my own purchase of a spectrum analyser to complement my interest in amateur radio (I have the callsign M0NJX), I return to your golden source of insights and support, time and time again, to address my gaps in electronics understanding . 73 de Matt
Thanks for the video. I watched it from beginning till end. Greetings from a fellow EE in Bulgaria:)
I never realised spectrum analysis could be done without FFT and always thought the analyzers are costly because of the high speed special purpose DSPs inside. This was a revelation and a lesson about how the same problem can have an enitely different type of solution. Thanks for sharing your knowledge with newer generation.
Spectrum Analyzers were developed and existed LONG before DSP circuits were ever invented.
Very nice video. I am going to have to keep my eyes open for one of the 1401A's. Thanks for posting.
I really like the way you explain things.
This is excellent video to help beginner to understand how spectrum analyzer works.
I have watched many of your videos, its truly Highly Informative, Sir. Especially the way you explain these hard concepts is truly amazing. Thanks a lot for making such a great effort. It would be of great help if done a video on S-Parameters. Thanks in Advance.
Excellent tutorial and great find. Thanks for sharing your knowledge.
Thank you very much for the knowledge!
How did I just find your youtube channel? Really great stuff, you are excellent at explaining some complex stuff. Thank you.
Very cool! Takes me back to my Fourier Series course in college.
I tore down a Scientific Atlanta spectrum analyzer from the mid-80s last year, and I still have some of the boards lying around. It was introduced IIRC in 1978 and was one of the first digital spectrum analyzers on the market, made primarily for use by the Navy.
Thanks for the class. I am trying to prepare for a job interview for a position that requires working knowledge of testing equipment, and though I was experienced with this kind of equipment while in my last job as a communications operator in the Marine Corps, I find myself needing a lot of refreshers... Thanks for the help.
Thanks for making this video.. 1401A is pretty cool stuff. I liked it.
Presentation was awesome very easy to understand.
Thank you! Excellent video!
Thanks very much. You're a natural for presentations.
That was about the simplest explanation of a spectrum analyzer as I've ever seen. Good to get back to basics sometimes, of course, like your REAL spectrum analyzer, lots of buttons and knobs to confuse you, kind of like sitting in the cockpit of an airplane.
Way back in the day when repairing TV sets, if one wanted to tune the IF coils to get a specific bandpass, one used a sweep generator and an oscilloscope. The frequency output of the generator went to the input of the IF strip, the output went to the vertical input of the scope. The sweep voltage from the generator was not a sawtooth, as we saw above but actually a sine wave. But then, it was a sine wave that was driving the sweep generator, which connected to the horizontal input of the scope. So you know what frequencies you were viewing on the screen, you had what was called markers, a marker generator created specific frequencies. You had a spectrum analyzer, but we didn't know it was called that at the time.
My sweep generator was made by EICO, Electronic Instrument Company, what we called "bench quality." My oscilloscope had a huge 5MHz bandpass, just good enough for working on TVs, couldn't even be used for Citizen's Band radio. The brand name of my scope was Hallicrafters, a familiar name to ham radio people, actually bought the scope in a kit form from Radio Shack, when it actually sold electronics.
from now, May 6th, you are my heroe!
Very nice video!
yeah man ya right that is a fun wee box. cheers for sharing your wisdom.
i have been looking for one of these.....good find KK4MSA
Awesome Video !
Hi,
Thanks for your video tutorial. I found your diagram and explanation very instructive.
Kind regards,
Sam
@w2aew You taught me something, in my books that a teacher...thx
Would've been interesting to see digital or SSB signals from a pre-amp on your antennas.
Can this thing be used as a SWR meter? Perhaps in combination with a white noise generator, you can measure the reflection of that noise with that analyser.
Very clear explanation, I have 2 defective AVCOM is more easy to understand the failure, thanks for share !!!
great video, thanks so much. Please make some more on specific applications!
I actually picked up the scope today. Now I just want the spectrum analyzer.
Thanks so much for this video. Great stuff. 73s
Very well explained, thanks.
W2AEW de KI4DBK ...Very nice video and very informative for a relatively young ham. I see that you are into Kenwood, as I recently got my hands on my first FB piece of Kenwood gear, a TS-520, which I love. Next to CW, another hobby of mine is repair of vintage HF gear and homebrew electronics, but I have a lot to learn of course. Keep the great videos coming, and in the meantime I will watch all your test equipment tutorials. 73
Very helpful! Thanks.
Great info, I picked up a Tektronix 7633 at a ham auction for $20, no HV and I am digging into it, not sure if I can fix it but from what I understand that scope can be used as a spectrum analyzer with the right plugin.
Yes, you'll have to find a 7L5, 7L12, 7L13, 7L14 or 7L18.
Yeah, problem is that these days, the plugins can fetch a grand or more.
Very informative yes good information al always, I have the TTi version of this and was wondering if i could use it to look at the if of a reciver as a spectrum scope i wonder what your thourghts on this ,would be is this possable
Great video, do you know if it is possible to use this type of spectrum analyzer to listen to radio transmissions (AM, SSB, etc) as well (perhaps with some added circuitry)?
impressive presentation! I have an HP141-T with the 8555A RF module which I got as a gift from a retiree. I've played around with it some but would like to learn more about it. Also have a comb generator that came with it....
Thanks, my HP8565A has an IF output I assume would be the best place to start. Also has a "1st LO output" which is interesting.
Great video thanks
excellent video. i am looking into building my own simple SA. So upon googling, ive seen some designs that have multiple mixing stages such as 2 or 3. Whys is that?
Mainly to provide flexibility in filtering and RBW selections, as well as control over mixing images and harmonic products. Frequency planning (LO frequencies, IF frequencies, etc.) is a complex topic.
Hi Alan I have what i think is the matching tektronix 323 oscilloscope and it is is nice condition working. The case appears to be the same size and has the same handle and same goofy ac plug . It is interesting that the 323 has the sony brand on it as well and i did not see that on your Analyzer. Great video by the way.
+allaird1 Yes, the 323 scope was done in Japan by the joint Sony/Tek venture. The 1401A spectrum analyzer adapter was designed to pair with that scope, but was designed entirely in Oregon - hence no Sony name on it. There was a way to link the handles together to make one "unified" unit. It was popular with the surveillance community due to it's small size.
Thanks Alan. Lovely and clear. Our club has a HP141T spectrum analyser with a RF section module which I acquired. All 4 stones of it! Would you be able to demonstrate such a unit in a video? We're a little wary of it.
73 de M0YDH
+David Holman I would certainly make a video with that analyzer if I had access to one. Unfortunately, I don't have one.
Excellent explaination. Where were you when I was in University? (30 years ago)
I have one (not working) off Ebay. I have not laid my hands on it yet but was wondering if you knew something about spares for the same. I purchased the service manual but is very thin on details about the 690 1590 VHF oscillator. No part numbers for the osc xistor or the coil and varicap are provided. Thanks for any help.
Good stuff! Just right. 👍🏿
thanks mate very helpful
cheers
hello love the channel I was wondering if you could use a nano type analyzer to do cal. on something like a analog scope . Or are they not accurate enough THANK YOU
Personally, I wouldn't.
Awesome as always with great detail! I've been wanting to build one of these ever since I first found out they exist as I've been wondering about if they do. Then randomly shopgoodwill.com had a ITC SA500 that I missed bidding on. Lately I've been thinking I can build with cheap modules... though still thinking will make an older style design. I'll add this make and model to my list of I've found so far, i.e. ITC SA500, Protek_7700_7800 and MIT-3100_3260.
If anyone is aware of any others, please note. Thanks!
Thank you 😊
As always great video !! Would be nice to build a homebrew 1401 :)
You actually can, I once saw a SA built from an electronic TV tuner, which uses a VCO. I saw it at a ham fest one time.
Hi!. We share one interest. Do you have any details on the Tek 1401A? need to know more about the 695-1195 first local oscillator. Nothing in the service manual. Please see above for details. Thanks.
I'm interested to find out more about the characteristics of the VCO used as the local oscillator. Is it sinusoidal or square wave? How does one go about designing one that can sweep are wide range of frequencies? And what parameters make a difference to the overall performance of the analyzer?
These are all good questions - with BIG answers - much more than can be answered in a text reply here. Very briefly: The VCO could be sinusoidal or square wave - it all depends on the design of the mixer and what it needs for LO drive. VCO designs vary quite a bit, although the most common for very wide range tuning often involve varying a current that is used to charge/discharge a capacitor. However, these designs have poor stability and phase noise. Range switched LC oscillators, or harmonic mixing are employed sometimes to improve this. There are LOT of parameters that make a difference to analyzer performance. The most important parameters are dynamic range, linearity, noise floor, etc. Much of this comes down to the performance of the LO, the mixers and the amplifier chain.
w2aew Thank you very much for this information. Am I correct in assuming that the first step in designing an SA would be mixer selection, then IF choice, then LO design? I'd like to see how far I can go with it and cook up a SA that can take DC (or at least the bottom of the audio band) to a few MHz. Main use would be AF distortion analysis, testing filters etc. Given this, I figure we will be mixing up, rather than down - I've heard that this is what is done for such designs. There seem to be a couple choices here: AD8343 looks best, PE4141 has a good-for-nothing datasheet. The former seems content with a square wave - I'm thinking LTC6990 pumped into a PLL. Analog Devices make some great high speed op amps so I doubt IF filtering is a problem. Am I having sane thoughts so far?
There are a lot of interacting factors - frequency/IF planning, preselection filtering, IF filtering, LO design, mixer selection, etc. However, for applications below a few MHz, these days it may be best to simply implement it digitally. There are many very high performance ADCs out there that can digitize the entire band of interest and then you can compute spectral results using FFT or Chirp-Z transform.
is it possible to isolate hum in the audio signal of a pre amplifier
using the spectrum analyzer? If yes, how would the hum be displayed on the analyzer? Also, can you apply the input of the spectrum analyzer to the circuit under test where dc voltage is also present or does the input of the analyzer require only ac signal without any DC voltage?
Many (most) spectrum analyzers don't measure very low frequencies like audio hum. Most cut off at 9kHz. Also, many spectrum analyzers cannot have DC on the input. The label on the front panel near the input connector will indicate if DC is OK or not.
@@w2aew Thank you!
Is a Tektronix 495P a good SA for radio repairs like sweeping and such?
It is a very nice analyzer, but keep in mind that it is approaching 30 years old...
needless to say.. very interesting thanks a lot for the information
I have a question regarding a mixer anomaly that, it seems, would occur during the L.O. sweep. When down converting which, I'm assuming, is what would be done to make narrow resolution bandwidths practical, would there not be two peaks detected since there will be two conditions where the difference frequency will appear in the filter pass band?
For example, consider a case where the input frequency is 100MHz and the I.F. filter is 10MHz with some arbitrary bandwidth. As the L.O. sweeps through 90MHz there will be an output at the filter (100MHz input - 90MHz lo) and thus a peak detected and displayed. But, as the sweep continues, there will also be another peak as the L.O. sweeps through 110MHz. (110MHz lo - 100MHz input). Or am I missing something?
This question arose as a result of my investigating the practicality of some of the simple and inexpensive USB spectrum analyzers from China (eBay). I have seen several reviews that confirm two peaks being displayed instead of the expected single peak. I wracked my brain over why this happened, and the above is one explanation that seemed logical. If true, and if there was a way to resolve the problem, through a reasonably simple modification, then the simple USB analyzer would be acceptable for my needs. Otherwise, not so much.
Yes, this would absolutely happen as you sweep from what is called low-side injection to high-side injection. This is typically dealt with in a "real" spectrum analyzer by employing multiple conversion stages. Each stage employs only high-side or low-side injection, and unwanted images are pre-filtered or post-filtered away. Remember that with high-side injection, the resulting image out of the mixer is spectrally inverted.
@@w2aew Thank you. Kind of confirms what I thought was going on with the simple USB analyzer I was looking at. Published schematics show a rudimentary single conversion scheme, just like the diagram you show in the video, and simple fixed BW filter.
Hi Alan. I want to complement you on your truly excellent set of videos explaining the operation of analog swept spectrum analyzers. I have a physics background and have recently been thrust into the world of spectrum analyzers. I was finding it really frustrating coming up with understandable explanations for RBW, VBW, ref level, and noise floor. Your videos offered me complete clarity for all of these terms.
I have a question for you. I think I finally understand the role of the ramp generator that is shown in your spectrum analyzer block diagram. However, I am not certain my understanding is on firm ground. Here is what I believe is going on. First, the ramp generator applies a time varying voltage to the tunable local oscillator that causes the frequency of the local oscillator to move over a range of frequencies. Am I correct up to this point? If I got the first part of my question correct, I want to move to the second part of my question. As a spectrum analyzer user, I am able to specify a span of frequencies over which I want the spectrum analyzer to scan. Specifically, I am able to select a minimum starting frequency for the scan as well as a maximum ending frequency for the scan. To keep things concrete, let me assume I select 10 MHz as my minimum frequency and 100 MHz as my maximum frequency. This choice of values establishes a span of 90 MHz. Now, I am guessing (and hoping) that my choice of 10 MHz as a minimum frequency maps to a unique lower limit voltage on the ramp generator. Similarly, my choice of 100 MHz as my maximum frequency maps to a unique high limit voltage on the ramp generator. The ramp generator then repeatedly drives the tunable Local Oscillator over this voltage range, which means my 90 MHz span is being cyclically pushed through the Resolution Bandwidth filter, to produce the signal that is displayed on the oscilloscope screen.
Yep, you got it.
Great stuff. Might want to consider cutting down on the "mmm'k" . Ditto for the 'see you later' pun. I used to work on simulators. I had to put up with 'see mu lator' for years. It gave me PTSD. Still in therapy...
Did you ever get a Tek 323 for this? I used to repair and calibrate these at Tektronix. They were designed for the cable TV industry, which was why the narrowest filter was 10 kHz. Not all that great for amateur radio unless you're doing TV transmissions.
No, never got a 323 to use with it.
A clean little photograph of your block diagram in the links would have been good, i'm feeling hard to draw with the camera tilted and moving
As with most of my videos, there is a link to a PDF of the notes and diagrams that I use in the video. The link appears at the bottom of the Video Description.
www.dorkage.com/youtube/Basics_of_Spectrum_Analyzers.pdf
really sorry man, I had a look at it after i put a comment, Thanks for that too man.
Do you have or could you make a video that defines the terms of "linear" , "linearity", "Non-Linearity" . These terms always give me a headache. Anyone that can explain them much appreciated. I have attempted google without much success.
Did you see this:
czcams.com/video/CK0zl34nZJc/video.html
?
Thanks !
Thanks
Can we do something fancy with the detector to get fast sweeping at narrow resolution bandwidths?
Not with a swept architecture. The narrower the filter, the longer the time constant. There's no getting around the physics of this. It has nothing to do with the detector. The sweep has to be slow enough to allow the filter to fully charge. FFT based analyzers have an advantage here. While you still can't get around the time constant, you can "compute" the spectrum over a wide range in one go. So, even though the FFT vector is long, it's only computed once. For example, a traditional swept analyzer will typically take 20-40 minutes to do a 1GHz sweep at 1kHz RBW. The same can be done with a wideband FFT analyzer in less than a few seconds.
@@w2aew is there any electronic analog to a prism, separating light in parallel? Not computational like an fft, but some sort of physical implementation? (kc1nyu here, btw)
@@nickcaruso I believe there are magnetic resonance structures (similar to NMR applications) as well as surface acoustic wave (SAW) devices that can be employed in a similar fashion for RF/microwave.
interesting. thanks! your videos are fantastic! very helpful for this software engineer electronic hobbyist! 😀
Oh, and it also used an oscilloscope for its primary output.
w2aew, what is your background?
@w2aew at some point it would be great if you could discuss the difference between swept-spectrum analyzers vs. fft ones. That stuff can get freaky with pulsed signals.
I'll have to plan on a video like that. The short version is: A swept-spectrum analyzer is like what is described in this video - a swept-tuned receiver with a detector on the output. An FFT based analyzer is like a wideband receiver, where the receiver BW is sufficient to cover the span of interest. The incoming RF is downconverted into this wideband IF and digitized, much like a scope digitizes a signal. The sampled IF waveform is then processed by an FFT to compute the spectral content. Advantage is that you can do a triggered capture so that the sampled IF signal is just the signal (pulse) of interest. An FFT analyzer can give you the spectral content of a short, continuous record (instantaneous spectrum), while the trace from a swept analyzer is drawn over a comparatively long time - not possible to see just the spectrum of a short pulsed signal for example.
@@w2aew Don't some analyzers do an inbetween thing? If the SA can take the whole span in one go and then do FFT on it that would make it a real-time analyzer right? (not shure if there is an commonly accepted definition of what a RTSA is, suppose it could be based on POI). What confuses me is that there are some analyzers, that are not sold as RTSAs, that have an "No FFT" Sweep Mode - that would imply the normal sweep mode is some sort of FFT. My understanding is that those analyzers, because they are not fast enough to do RTSA and can't take the whole span in one go, do a "half-way" thing. They sweep the IF in steps, sample and FFT process it in chunks to get faster overall sweeps. I'm trying to get my head around how this potentially could cause problems. In particular with very short pulses.
@@juliussokolowski4293 All FFT based analyzers have a given span that they can digitize - this is called the instantaneous BW, or acquisition BW. You're right, most times this acq BW is narrower than the desired measurement span, so the spectrum is computed through multiple segments to build up the entire span. If the Span is set to be equal to, or less than the acq BW, then no stepping/sweeping is required. This is how vector signal analyzer work. They capture/record baseband (usually IQ data) over a given acquisition BW for further processing, often modulation analysis, etc. Because FFT based analyzers process a seamless record of time-domain data, they often can be "triggered" to capture data at a specific time - very useful for looking at pulsed signals. When the analyzer has the ability to perform streaming FFT processing on the IF (IQ) samples as they're flowing through the processing chain, then this is what is called realtime spectrum analysis. Some do this just to enable triggering on transient spectral events, and most these days can also display the composite result of these realtime FFT results. These are what are typically called RTSAs. If you ask an RTSA to show you a span that is greater than this realtime BW, then it will often "emulate" a swept analyzer by stitching together several segments (that are acq BW wide), as was described a bit earlier.
@@w2aew That stepping is what I have been seeing on the SpectrumMaster - the trace was built up in "chunks" or "sections". Not the regular sweep I'm used to from my FSEB or my 8594. I was looking a over a pretty wide span I should add. Gotta do some more studying - in particular on this pulsed signal thing - never dealt with it before. In my quest to up my SA game I went looking for some reference litterature on spectrum analysis last night and found out that R&S has a book you can download from their website called "Fundamentals of Spectrum Analysis" by Christoph Rauscher - looks to be a nice primer. Anyways, thanks for your insight. Keep it up - your videos are a great help to my Ham License students. Btw. The round table is great!
@@juliussokolowski4293 Yes, I believe the Anritsu SpectrumMaster has a digital IF, thus the final IF processing is done as an FFT. I don't know if there is a way to sync the segments to your RF pulse with that unit though.
Dear Allan W I did not unsertand how you know my name but i saw your vídeos and i saw your Face, so now i recognized Nice to see you again and have a nice day Regards Jesus Osorio
Do all spectrum analyzers sweep frequencies ? Do I need a sweeping function generator to connect to a spectrum analyzer to see the filter response of some simple breadboard passive filter ?
Most spectrum analyzers sweep. However, there are some that don't - they sample the signal and perform a discrete fourier transform to compute the spectrum. The traditional way to sweep a filter is to use a tracking generator such as in this video:
Measuring low-pass filters in a Ham Radio HF Packer Amp for Amateur HF Bands
However, you can also use a broadband noise source at the input of the filter, then the spectrum analyzer will show you the filter's response, like this:
How to measure an HF Bandpass filter response with the MDO3000
*****
What piece of equipment could I buy that is a broadband noise source ? Like white noise ?
Peter B Many Arbitrary Function Generators can generate a noise waveform. Or, you can build one. Typically a zener diode in reverse breakdown mode is a good noise source - then follow it with an amplifier. Here's one example I found after a quick Google search:
www.maximintegrated.com/en/app-notes/index.mvp/id/3469
and another example:
www.wb0smx.net/?p=572
Is this spectrum analyzer any good ?
www.gwinstek.com/en/product/productdetail.aspx?pid=4&mid=61&id=1339
Will it allow me to see frequency spurs from mixers and oscillators and also use the noise source to see filter responses ? At the moment I have very modest needs just to learn and I'll worry about high end stuff after I finish school. I just want to know if this thing is worthwhile for simple spectrum viewing ?
What is your opinion ?
Peter B This is a very basic, entry level spectrum analyzer. It has limited RBW settings, which will keep low level signals hidden in the noise floor. There is only one spec for spurious which is -45dBc - which means that you may not be able to distinguish spurs from your mixers, oscillators, etc. from instrument generated spurs that are at similar or lower levels. That being said, it is a very inexpensive spectrum analyzer - so you'd get what you pay for. It certainly will give you a view of the spectrum.
What is a typical circuit for a mixer, in VHF?
A gilbert cell mixer, such as that found in the NE602/612 is very common.
@@w2aew Thank you for your help. Just pulled up the data sheet for the NE602, a few good circuit examples of using the chip are included.
@@koggism You might want to check out my videos on the Gilbert Cell:
czcams.com/video/7nmmb0pqTU0/video.html
czcams.com/video/38OQub2Vi2Q/video.html
where did you learn this? I am guessing the military. INTERESTING.
No, no military training. I learned most of this stuff over the years "on the job" over the last 25-30 years or so.
How much did you pay for your 1401A? Did it still have a working battery pack?
I think I paid $40 for it. It did not have a working battery pack, and I had to replace a few capacitors just to get it working. I gave it to a friend years ago now.
@@w2aew Thanks for the reply.
Good stuff man !
74 EA2CTK
Wish you were my neighbor, I sure could learn lots more, Thanks DAVE W4GSM
The link to your notes does not seem to be working.
Updated the link - fixed now.
U need a HackRf one and learn to use GUN and show us how we can use the HackRF one's hardware as a SWR meter. Loosely a Antenna analyzer.
I buy electronic test equipment at the estate or yard sales
thanks for the advice Be nice to pick up a plug in like you have 73s Jim KD7TU
Jim - I found mine at a hamfest several years ago. I am thinking of selling the one I have since I rarely use it anymore.
Holy shit that mixer thing is smart. Why can't i come up with ideas like this? :D
Very interesting your explanatios but, stop moving the camera each seconds. We couldn't look at more then 2 minutes. Thanks.
My video production techniques have improved a lot in the 5+ years since I made this video...