TSP #77 - Teardown, Repair & Experiments with the Anritsu 37347A 20GHz Vector Network Analyzer

In this episode Shahriar repairs an Anritsu 37347A 20GHz Two-Port Network Analyzer. The unit does not boot up and as a result its internal state is unknown. The boot fault is traced to a bad RTC module which has a built-in integrated battery. A replacement unit is located which allows the unit to fully boot.
The instrument initially displays an unlock condition on the internal PLL. This problem is resolved by loading the PLL calibration files from the HDD. The complete block diagram of the unit is examined in detail and an unusual PLL intermittent locking problems is demonstrated. Various measurements on several of the internal PLLs confirm their functionality. As a final experiments, the instrument is calibrated and the S-parameters of a tune-able band-pass filter is measured.
The Signal Path Blog
www.TheSignalPath.com

[@micksshop]

Here it is 6 years later and I am watching this video again. I have replayed many of the videos on this channel. thank you.

[@thegoodies2]

Yes we are still watching right to the end!

[@Thesignalpath]

Great! Thank you.

[@ChipGuy]

+The Signal Path Blog Of course a big thumbs up and I watched it right through the end. I'd thought you are going to address the display issue again. Got an idea for that in another comment, replacing the CCFL with LED filament strings.

[@ironridgeaudio]

I stayed for the whole video. This kind of equipment is out of my league but you explain it very well and I enjoy watching you troubleshoot it.

[@peterkjellstrom3100]

So +1 for watched all of the block diagram walk through. It was well done and the coloured doodling made it much easier to follow (or more correctly, easier to keep the overall picture in mind as you added detail).

[@godfreypoon5148]

31:30 - Sure did. (With a dinner break.) 46:00 - We always like it, Shahriar. Your videos are an invaluable gift to us!

[@alandoak5146]

I was a system design engineer on this instrument '97-00, what a blast from the past!

[@godfreypoon5148]

10:32 - Shahriar has a light-bulb moment, and makes a FULLY AUDIBLE chime sound! (Back when I was a kid, I used to use a similar part, a battery backed RAM module, instead of burning EPROMs. Thank you for the free samples, Dallas/Maxim!)

[@Thesignalpath]

Haha!! Yes, the chime of ideas.

[@paulmoir4452]

I'm an electronics amateur for whom 10MHz is fast. I listened to and understood the block diagram description because it's a challenge to my understanding of electronic systems and often electronic fundamentals. Thank you for teaching this; I sincerely appreciate it.

[@Thesignalpath]

You are welcome.

[@gelderg]

Great video! As someone who currently works on other VNAs at Anritsu, it was neat to watch you walk through the block diagram and troubleshoot one of these older models. It's a rare treat to see people tear down our stuff, so I'm glad to see you got it working again - mostly :)

[@robertfoss803]

I very much enjoyed the detailed repair process. Overall it's not a skillset that I posses, so being guided through the entire process is very illuminating.

[@AI7KTD]

Very clear and easy to follow explanation on the block diagram.

[@cncdavenz]

The cat must be very intelligent as he sniffed out the problem right at the start. Although I am not a microwave engineer its good to understand the process of dropping the signals down to dc to allow it to be analysed with digital circuitry.Cheers Dave.

[@TheAmmoniacal]

I did not listen to the block diagram description, but I'm a biologist and I watched the rest. Should be a win? Great content.

[@electricalscience3354]

You are actually have the best channel about electronics in whole youtube. After your videos i actually learned english! Thank you for you work!

[@WillstDuNichtWissen]

"I hope you're still here" ... well that's why I am here. I really like the way you describe such diagrams and in general the cool stuff you show in your videos.

[@joopterwijn]

I liked a specially the way (use the colors) explaining the block diagram! Nice!

[@RogerBarraud]

Yes, that's a great old-timey trick - I'd forgotten it (last seen on paper, of course ;-) )

[@chuuni6924]

I just wanted to say that I don't even know what this instrument does, and yet I watched a 45 minute video about it from start to end. I'm not very good at it, but high-frequency stuff just fascinates me. :)

[@gamccoy]

I listened to your explanation of the block diagram. It was scary to look at but your explanation was quite good. I am not worried about fixing the 2nd PLL. As always, I learned a lot. Thank you.

[@k4rokirkpickering117]

The block diagram description with the colored paths was an excellent way to convey the analyzer's design. Your videos are very educational, and I appreciate the time that you spend producing them.

[@BuckCalabro]

Watched every minute. Particularly liked the multiple colours used to trace the signal paths. Glad there was a backup of the YIG coil parameters on the HDD.

[@dgaborus]

I just bought a faulty 3577A and so happy that I will have something to fix! :)

[@ytrewq6789]

I have always been very thankful for the extra time and effort you take to explain the block diagrams as I not only find the way you explain it very easy to follow but very educational as well.... I have learned more from your mini lectures in your 20 minute block diagram explanations then hours of researching through the text books and manuals. Thank you so much for going to the extra mile where most will not bother to do so, You personally made my life a whole lot easier and saved me a lot of extra time and frustration by doing so. You are an excellent tutor for me any ways :-D.

[@Thesignalpath]

Thank you. I am happy to hear that.

[@silvermica]

I’m a design engineer who worked at Anritsu during the 90s - I know the designer of the VNA instruments at Anritsu - his name is Jon Martens

[@trentspencer8189]

I know this is a rather old video, but I definitely watch the block diagram explanation, that is one of my favorite parts of your videos, you explain things very well. The only thing I might ask to be different is to be more in depth and go into more detail. I understand you have to keep things rather brief in order to entertain the majority of people.

[@xtevesousa]

Yes to the block diagram, It's extremely interesting. Now that service manuals don't have the circuit description, you're the only one that explains how high-end equipment works. Very usefull for those learning analog/rf electronics.

[@TyPoProteus]

Just a heads up, I did watch the whole video, and I think you should show us more of what you do, it is hard for us to say what we want to see, because we may not be fully aware of everything that you do, and therefore might limit ourselves to only seeing things we have experience with. Fantastic job as always, I am extremely jealous of all the equipment you have! Thank you for the time and effort you put in to the videos for us!

[@RogerBarraud]

Dithering would generally look like noise, not random jumps visible as such.

[@WTK444]

The block diagram explanation was great. The effort you put into your videos is much appreciated. I always learn something new each time I watch one.

[@nraynaud]

I love your work, I'm not a microwave guy, but you are demystifying the thing a bit to me.

[@trickyrat483]

And, after several weeks of watching, that's me caught up with all your videos. Went through each and every one from beginning to end. Now I won't know what to do with myself when I get home from work! :) It's been a long and fascinating journey. Many thanks for all your hard work on these. Simply nothing else of this class on KZbin. I enjoy the mix of theory and practice; much better than a simple "teardown" or "quick repair" type of video. I've learnt so much. You are a great inspiration, my friend. All the best and thank you once again.

[@JetNmyFuture]

This is one of my favorites. Thanks! It seems I end up watching all of the videos more than once.

[@smgvbest]

Watched all the way through and loved going over the block diagram and the whole troubleshooting. hope to see more of them like this

[@douro20]

That uses a Motorola MVME162-043 single-board computer, which has a 25MHz 68040 processor. Although it was originally designed to run FLEX OS it is most likely running Microware OS9 in this application.

[@pithandmoment]

I watched it all the way through and enjoyed it thoroughly! Thank you again for devoting so much time and effort to these presentations.

[@miceuz]

Hey! Thanks for the episode! I'm always waiting for your block diagram explanations, also, you explaining your intuition while troubleshooting is very valuable too - a lot to learn from!

[@PileOfEmptyTapes]

I spot a number of small electrolytics on the suspect 2nd PLL board. You may want to ask an ESR meter for its opinion of those (actually that Fluke RCL meter might just do the job). You might also be able to get a number of thin wires in there so you can monitor all the locally-generated voltages - I suspect one of them will be unstable or oscillating (bad cap or bad reg). Great things are often brought down by small ones, and this unit proves it once again. One of my retro PCs uses one of these pesky Dallas-style CMOS clock modules (which were popular up to about '97), and guess what, I can only get a boot loop out of it these days since it fails to save its CMOS data and stumbles over bad CMOS checksum over and over again. Good job breaking down the block diagram BTW. These higher-order PLL affairs can be quite intimidating at first glance.

[@btouw8558]

You explain things clear, I watched even multiple times the blockdiagram,

[@adk0212]

Always enjoy the repair videos, even when the subject matter is approaching the outer limits of my background. Thank you for producing them!

[@wither8]

If you don't want that gear taking space up in your lab, I'll put it to good use Shahriar for a fair price and pick it up in NJ ;) I've said it a bazillion times already but thanks for the high quality, educational videos. You + Paul from Mr Carlson's Lab + w2aew + mikeselectrical put out absolutely fantastic quality videos. Sure I've said it a million times already, but thanks so much for investing the time in making these -- they must take at least 5 hours of production for each hour you put out.

[@Thesignalpath]

+wither8 Thanks for the kind comments. Even though this is an old unit, they are still rather expensive. On eBay they are listed for $18k! I don't think anyone will buy it for $18k of course, but even at $10k, it is still above most hobbyist's budget.

[@AureliusR]

Hey Shahriar -- that portion where you said it would make a cool background... did you take a high-res photo of that before this went on its way?

[@PyroShim]

Perfect video. Thank you for taking the time and effort to make videos like this. And of course I watched the entire explanation of the block diagram.

[@gabeapger1846]

Love the videos and always watch to the end. Thank you for putting in the time to give back.

[@klassundberg4464]

Quite lovely as allways. ) Fun to see how a network analyser works, and it was really helpful with the run through of the block diagram.

[@SLEAGD]

Yes, I actually listened and followed through the entire Block Diagram description. Thank you, nice explanation and thorough approach, as usual.

[@preludelinux]

loved the video and the presentation of the block diagrams with color overlay helped keep everything understandable. thanks always look forward to your videos.

[@jacekpaw1]

I listened to the block diagram description, also I vote for further repair of the unit. THX for this video, it was one of the moast interesting I have ever seen on youtube.

[@donreid358]

Great video, thenks. If it wasn't working I'd clean the connections on those socketed ICs, and of course check the power supplies. I've found many old systems that just had oxidized connections.

[@billysdomain]

Hi Shahriar Many thanks for all your efforts in this channel Regarding the VNA, i have a video idea for you which im sure will help many, including myself, relief a major recurring headache! The problem: You have to match impedence on a trace between two devices In a normal situaltion, you may be lucky enough to have the device impedences, for example, a RFO pin to a saw filter at 868MHZ may consist of: RFO pin on IC, 30+j25 for max power transfer at 868MHz SAW filter datasheet shows a simple 50R input at 868MHZ so, how do we get, via a PCB trace, from 30+j25 @ 868MHz to 50R? How do we choose the matching components and most importantly how do we verify the result with the network analyser? How to we use the VNA to find the complex impedence of the saw filter? this would be an awesome tutorial. This is a great example because freqs arnt too high, so people can play along at home, FR4 is still ok, no strange other HighF stuff have effects....and its something that gets encountered a lot Thanks again for your great content Billy

[@EEEAmin]

was waiting for a VNA video for ages ...and for you 7 months in the making ...wow

[@PapasDino]

Enjoyed the whole video as usual. I vote for fixing if it's "broke", 'course that's the OCD in me!

[@RogerBarraud]

Durn tootin' it is! ;-)

[@Petex90]

Thanks for the very detailed and well explained video! There's not too many youtubers at these days, who have time to do videos like this or have enough knowledge to do so.

[@rcleveacp]

Great Video I love the explanations and troubleshooting. It is really nice to see peaces of test equipment that I am not familiar with. I appreciate all your time and effort and thank you. Keep up the great work. I did watch the entire video. Every minute of it!

[@wiredkat]

Of course I had to watch all the way through. Actually when I see you've posted a video it can be a bit of an inconvenience because I know that whatever I'm doing I need to clear my schedule in order to watch. Another great video. I vote for digging into it a little deeper to see if that frequency instability is normal or if it can be improved.

[@brickbybrick6150]

Thanks Shahriar - great video. I enjoyed your explanation of the block diagram, and the process of tracking down the fault. It would be interesting to see some experiments with this or another VNA. It is not easy to gain experience with this type of equipment unless working in a well equipped lab.

[@kungfumaster8171]

Excellent. I appreciate the time and effort that goes into doing this. Thank you.

[@stazeII]

Great video as usual. Really do like your repairs since you actually get into the theory of operation, how each part interacts, and then trying to figure out where to actually look for the issue. Rather than just the shotgun method that I often seem to use. =)

[@nogeek43]

I loved your block diagram explanation. Please keep up the tear-down/repair videos and subsequent experiments.

[@Mrjm5411]

Really enjoyed the block diagram walk through, Thank you. Watched in complete and enjoyed.

[@bobfuller]

Really enjoyed the video. Watched it all the way through. Great video format and great presentation style. I always learn something new from your videos, even after 30 years in the business. I always look forward to seeing them - they are the best technical videos on KZbin in my opinion. Thanks for the time and effort that you put into them. Your repairs videos are my favourites.

[@jonka1]

You speak for all of us.

[@jpdear3064]

Watched to the end. Crystal-clear explanation. Many thanks.

[@ketturi]

This was quite interesting troubleshooting. Simplifying and explaining function block diagram that way really helps to understand how equipment works and what could be culprit of problem. I recommend taking backup of that HDD, if that's gone, all calibration values and software could be impossible to recover. It could be that calibration values for PLL is still bit off that causes it to be unhappy. I don't know how much software controls YIG source and PLL, but it could be the case that while software is busy doing other things it just leaves it freeruning and locks it only during measure (I don't know either if that would even work, how fast it can recover stable lock etc.)

[@Thesignalpath]

Yes, the second PLL problem is of concern. I'll leave it running overnight to see if it gets worse.

[@vincei4252]

Yes, I did watch everything and followed it. I also learned that pin-diodes can be used for leveling and was surprised to see they behave like resistors dependent on the current flowing in them. I'm gonna buy a handful to play with. Thanks for all you do.

[@FakirCB]

Yup, made it through the block diagram too. Very well described, no problem whatsoever.

[@outagas2008]

I watched everything.. I'm a Mechanical Engineer working in telecomm so I can relate. Your troubleshooting methodology is fantastic.

[@rwils6333]

Thanks for finding time. Video is informative as always. As for the question did I listen to whole video. Not yet. I decided to watch only 1/4 way through then in 2 weeks watch till half way. Then if I can I'll finish watching in about 6 months.

[@YTDoedel0815]

Great video, as always. Please keep your explanations as detailed as they are! Although I, as a mechanical engineer, understand only half of what is said, I keep on trying until I get it. ;)

[@opelize]

I listened to the entire block diagram description - thank you!

[@JGunlimited]

I also listened to the entire block diagram description, and learned a lot from it. =P

[@cmdstraker]

Yes, I did in fact sit through the entire block diagram sequence and enjoyed it !

[@jpopwell1]

very good. i love watching both the work and the descriptions , thanks

[@Yrouel86]

You can actually sort of repair these potted RTC modules grinding the enclosure to access the internal pins so you can disconnect the internal battery and connect an external one. It's not pretty but you do it only once.

[@topherteardowns4679]

Watched it through twice actually . the block diagrams are one of the best parts

[@Romenadan]

Excellent video! Yes, I listened to the entire block diagram explanation, haha. I am curious whether that dithering is on purpose or not, but since it works I don't expect you to dig in again and try to figure it out.

[@7head7metal7]

I enjoy these videos really much! I'm currently in University and studying EE with RF, wireless and communication technology as my main focus. So it is great to watch these videos and learn a lot of from it. Thank you very much for this, and all your good and hard work! We have some old Wiltron units in our high frequency lab, and I really like the vintage feel with them. Although it's sometimes difficult to use them. Greetings from Germany :)

[@Thesignalpath]

These old units are nice sometimes because the operation can be a little easier to follow. And hello Germany!

[@7head7metal7]

The Signal Path Blog Thanks for the reply, you are absolutely right! Currently I'm trying to fix an old Tektronix FG 504 Function Generator, but still need to build a break-out-cable to trim a bazillion pots outside the TM 504 Main frame. And I really like it. Plus, there is a lot to learn from that.

[@jchelm1979]

another excellent video, your content is always of a very high standard

[@EthanBlanton]

I listened to the entire block diagram description. :-)

[@catt87]

As always, very interesting and educational. Thank you.

[@jovianconflict5366]

Another excellent video from my favorite channel.

[@RminusOR]

There hasn't been one video on your channel... i didn't watch through to the end. I think you can be proud of your teaching skills. Of course your sample of 'students' happens to be interested in this kind of stuff ;)

[@ironman7261]

yes very informative video we appreciate your effort to put these on utube you certainly know your stuff keep up the good work!

[@timturner7609]

ive watched it more or less all the way through, but not in 1 shot. lots of information to absorb

[@thekaduu]

Thank you! You are simply the best.

[@mghumphrey]

I really enjoyed this one. Thanks

[@toddphelps5030]

Yep, I watched all the block diagram bit!

[@mrkattm]

Very good video, you are the best you tuber in this genre, I really appreciate what and how you do it, you are a fantastic resource for those wanting to learn, I only hope people appreciate it. I wish I had something like this (and you) when I was working on my BSEE. Keep up the good work, you are awesome!!!

[@Thesignalpath]

Thank you. There will be more repairs I am sure.

[@1337NoMad1337]

Yup I watched the entire video. It's interesting and helpful. Can you do a Video on PLLs? Principle of operation, use cases, different types, features, important characteristics, common faults, troubleshooting? You could use the oddly behaving PLL board in this video as an example.

[@daveblane6442]

Always great, Shariar! I am addicted

[@joehubler4965]

great job, love it!

[@gfrias06]

Very nice video. Thanks!

[@xDevscom_EE]

Nice vid. Perhaps we can have video in future explaining a bit about basics of S-parameters, and few typical concepts of using them to characterize components or transmission lines? That can be handy for better understanding of high speed designs and such! I think you have lot of experience in this field, which we all can learn from ;)

[@vaualbus]

finally this video. i've waited for it fora long time!

[@leppie]

"And that is why there is a power amplifier BTW" So obvious, still have no clue, but I love it :D I would like a laymen's wiki for all these funky terms :)

[@leppie]

+leppie Not to be nitpicking, but your recent videos seems to lack a bit of optical focus in the recording. Even at 720p, it seems quite blurry.

[@erichislop8921]

I have several of these Wiltron/Anritsu analyzers that I use with millimeter wave extension modules, and I have observed the same problems that you have seen. With some software versions, a phase lock fail will result in the "warming up" error message, and in others it will say "phase lock failed." For example, if I forget to turn on the power supply to my extension module, one of the VNAs will say "warming up," whereas another will tell me the phase lock has failed. I think the phase locked loop jitter that you observed is common. If I look at the output IF frequency of my millimeter wave setups, I see the IF jitter even when in CW mode. However, if I press "hold," the jitter disappears. It is almost as if, when not in "hold," the machine attempts to re-lock between data points even though it is not changing frequency. To me, the major flaw in the design of these VNAs is the dependence on the battery backup RAM (the BBRAM) (The module you replaced on board A9) for boot-up. As you found, the hardware calibrations for the machine are stored in the BBRAM, and are necessary to obtain phase locking of the main oscillator. You were able to recall the hardware calibrations from the hard drive. I had one machine for which the hardware calibrations had never been stored to the hard drive (I believe this to be a serious omission by Wiltron), and when the BBRAM battery ran out, I lost the hardware calibrations. For this reason, I urge anyone who has a functioning 37series VNA to store the hardware calibrations to the hard drive to ensure that they are available if the BBRAM dies. I also store them to floppy disks. I also bought a pocket programmer and a bunch of M48T08 Timekeeper RAMs (from China) and copied one for each of my VNAs so that I could quickly be up and running when a BBRAM failed. In one case when I tested my copy, the machine hung up indefinitely at "could not set time and date" because an invalid date had been programmed into the timekeeper portion. It is unfortunate that a small error such as that would keep the machine from booting up.

[@staffanb8357]

Eric, Great to hear from one with experience from the 37 series. I have one too and like it very much, but guess the BBRAM is about to die. You mention just replacing the BBRAM after programming it. Can you please describe a bit more in detail? Sincerely hoping someone - with good answers - is reading this... just three years later...

[@OneBiOzZ]

i did sit through the entire video and it was really interesting! Network analyzers have always been somthing i wanted since i was a kid (not that i could have done much with one back then) ... maybe one day when i get a proper spectrum analyzer with a tracking gen i can justify the purchase ... But i would take that HDD out and image it and copy the contents into a more modern drive ... that one sounds rather EOL

[@Thesignalpath]

Yes, that HDD is not a happy one. I'll back it up.

[@admiraladonis3258]

super, super nice! keep up the good work

[@dabay200]

Superb video

[@mikeselectricstuff]

I wonder what the state the battery in taht replacement unit is in - does it have a date code? ISTR these have some sort of 'factory seal' mechanism to stop them drawing power until they are first used

[@Thesignalpath]

I'll check. I can send it to you if you want to CNC it out. :)

[@MrCrystalDB]

+mikeselectricstuff +The Signal Path Blog would love a video on these things !

[@Enzaie]

+Crystal Dabuu Me to..

[@RyanUptonInnovator]

If the signal in the PLL section 2 that is jumping about or dithering is still causing a problem then can I suggest, because PLL section 2 is created from two PLL parts mixed together, that only one of those parts are faulting. The PLL section 2 signal may have a course and refined part that are mixed together. The refined part appears to be faulting.

[@vitormhenrique]

Awesome video, watched the hole thing of course, even though i can only understand 20%....

[@thenoiselevel4943]

Nice video as always. I like the description of the block operation. I am curious how many hours of study are behind your electronics knowledge. PhD always make the difference. I give anything to spend one day with you at Bell Labs and just watching you working

[@superdau]

I watched the whole block diagram segment. I guess it was the candy colors! ;)

[@henningschaferhoff1533]

I've watched through all of the explanation, the colored markings made it possible to understand the signalpath (damn I tried to avoid this word but its not easy) Could you do some more fundamental videos where you explain some of the basic elements used in the block diagram? Filters are pretty easy to understand but there are some elements explained in this video that are at least harder to grasp. ;)