TSP #49 - Teardown, Repair and Analysis of an Anritsu MS2721A 7.1GHz Portable Spectrum Analyzer

In this episode Shahriar takes a look at an Anritsu MS2721A portable spectrum analyzer which does not completely boot up. After displaying the message “Application Running!” the unit no longer advances to the main measurement screen. After intentionally corrupting the firmware, a system menu is observed.
During the teardown of the unit, various DC-DC converters are examined along with a basic overview of the Linear Technologies DC-DC converter datasheet. The main problem is traced to a faulty SRAM IC. After replacing the IC the unit completes the boot process. Various simple experiments are performed to ensure the unit’s functionality and correct operation.
The Signal Path
www.TheSignalPath.com

[@Starchface]

I am subscribed to over 250 KZbin channels spanning a wide range of topics, but it is your videos that I look forward to most. I can't adequately express how great they are. You have an understanding of all these systems and subsystems, and an uncommon ability to place them in context. It is the latter quality that I think puts you well into the "genius" domain. Plenty of people know about electronics or microwave system design, but how many can look at a PCB full of surface-mounted components and immediately recognize all the subsystems from a block diagram? How many can bodge up a completely different circuit on a PCB to replace unavailable components? I wouldn't be surprised to see you build a spectrum analyzer out of stone knives and bearskins if the need arose. You operate at such a high level that it is an honour to watch you work, sir. You may be working with electrons and waveguides, but you're a real-life Sherlock Holmes. Is this odd behaviour a red herring? How many blind alleys will you be led down? Is this bad component colluding with another? Will you catch the perpetrator before he-or it, rather-does more damage? Will all the loose ends get tied up this week, or is this episode going to be a 2-parter? Okay that's enough of that. I will never approach your level of ability, but I consider it a success if I leave each of your videos with more knowledge than I arrived with. Thanks for the brush with greatness, Shahriar.

[@yaghiyahbrenner8902]

pillsnotbills Agreed I second that.The man holds a PHD in EE and these videos are just a brush of he`s depth.

[@Thesignalpath]

pillsnotbills Thank you for the kind comments. You give me too much credit! :)

[@toddevans1]

pillsnotbills yes

[@Starchface]

***** Let me help you with that. Shahriar for you is the carrier, and you are the modulation...no? It's horrible, I know, but it wrote itself.

[@Starchface]

***** Would you believe a new signal path blog video just appeared? My Saturday night just got a lot better.

[@Shyam-ym3em]

Great work Shahriar, You guessed things very accurately. I was part of the Analog team that made it almost 15 years ago. The ADC and the 100 MHz sits in that metallic section on Digital board. Nice video, great presentation.

[@yaghiyahbrenner8902]

31:00 Held my breath. ahhah!The benefits of self testing gotta love it. Your repair analysis of the SRAM was spot on. Great repair.

[@amtpdb1]

I always enjoy your videos with the way you present them. This is considering that I don't understand 90% of what you are saying! Still, I always watch them. Thanks for the repair videos.

[@viesturssilins858]

Thank you for taking the time to re-shoot the lost footage for us!

[@lwfeagan]

As usual, I am in awe of your repair skills. Very impressive.

[@rysend]

Excellent as always...a highlight of the day. Cannot thank you enough for sharing your knowledge.

[@contextualelectronics]

Love love love seeing the troubleshooting stuff that you do!

[@MauroSedrani]

Great and very interesting work Shahriar!

[@TheTugge]

Ahh.. Friday night and new episode of total awesomeness :) It's time to pour beer in the glass and start relaxing. Thank you for these wonderful videos!!!

[@batterydudellc961]

Yeah do that! I would love to hear a rant on DC-DC buck boost converters. Great video as always.

[@nathanpc]

Another great repair! I think it would be nice if you made a video about building a DC/DC converter using a microcontroller instead of a dedicated IC and go over the theory of it.

[@stazeII]

Well done, and excellent video, as always!

[@nirajkini7838]

You are Mastermind, Brother. I love these videos. please upload more.

[@mnelson10000]

Dude, you're a bad ass...! Nice work!

[@douro20]

The large IC next to the two memory chips is the main CPU, a Hitachi SH3. There are actually two CPUs; the second is implemented in one of the two FPGAs and is most likely a SuperH based softcore.

[@rubusroo68]

Excellent video & outcome

[@AF6LJSue]

Good Video; Thanks for showing this repair.

[@DAVIDGREGORYKERR]

I found this very interesting and informative.

[@nplanel]

You have a good logic repair :) As advise, you should glue some heatsink on top of these SRAM. I'm pretty sure that a thermal design issue common on this series of Spectrum Analyzer.

[@videobruceb8879]

Update to the model year, in the front of the user manual, the compliance declaration is what showed the 2004 date.

[@stevepoppe8433]

Another cool repair job! How about using some of your GHz-range gear to do a survey of a typical EM envt? What all is out there? Cell phones, wireless routers, motors, fluorescent ballast, garage door openers ... It would be interesting if we could "see" in this range.

[@aerofart]

Good job! Nice repair video.

[@opelize]

Well done sir!!

[@PaulHigginbothamSr]

I think the reason for the delay on boot was correction for halting the previous update so you could find the menu for error errata table.

[@DoRC]

it's amazing how replacing the RAM chips fixed the firmware you corrupted ;)

[@drstrangelove09]

Well done!

[@johnnymartini1072]

looks so easy when u are doing it :D

[@Citronitroify]

Awesome!

[@KirkOfBellevue]

good job

[@MrMac5150]

EXCELLENT.

[@dtiydr]

6:31 xD Btw very interesting to watch, tnx! :) Where do you find these instruments? Not exactly in a dumpster I would guess?

[@gacherumburu9958]

👍👍 very inspiring

[@brig.4398]

May I ask how you replaced those chips? Was it with a hot air gun?

[@TheFool2cool]

How did you corrupt the firmware to access that hidden menu, got this exact model on the bench with the same sort of boot loop.

[@rolanddietrich5776]

Please ( should you find the time ) do make a video on the DC to Dc converters

[@videobruceb8879]

Great review, a few comments/questions; 1. I would of tried it with just the one memory chip replaced first, you can always replace the others later if it still failed. 2. I would of also done a test boot (not recorded) after you replaced the memory instead on a live 'discovery', 3. This model apparently came out in 2004 according to the owners manual. 4. What I don't understand is why would those voltages be at or near zero off the PS given this type of failure? Thanks.

[@Thesignalpath]

videobruce b The voltages are zero because those supplies were not enabled yet. They only get enabled at the end of the boot cycle.

[@RyanBlace]

The datasheet is dated 2006. :)

[@stefantrethan]

Nice repair, but you can stop pretending now, we all know it's really the cat finding the fault. ;-)

[@superdau]

Do you only show us the successful repairs? Because it seems you can always get the things you touch working again ;)

[@Thesignalpath]

superdau So far I have posted all the repairs I have attempted. There was the frequency counter repair which I could not complete because a replacement part could not be found.

[@vaualbus]

The Signal Path Blog And it way the core of the entire RF system. Ah Ah Ah

[@kartikarya6669]

Hi Shahriar! Here is your new fan from India. I loved the way you described the things and made them very easy to Understand. But Still I am Stuck in Upgrading the firmware of my Spectrum meter as it’s Ethernet port was damaged badly ( On Board PCB Prints are not there for fixing new Ethernet Port.) I tried very hard to get firmware on web but couldn’t find it. So I request you to send the weblink for its firmware.

[@ToniT800]

Hi SignalPath.Your channel has no logo/picture. Is it a bug or a feature?=)

[@Thesignalpath]

Toni T800 Thanks for letting me know! I don't know what has happened. I'll take a look.

[@steveone]

Theres another one of these for parts only on ebay . US $1400.00 .

[@Escaladist]

is that a russian blue cat?

[@TheAmmoniacal]

Your cat's name is Bush?

[@Thesignalpath]

TheAmmoniacal Haha... No his name is Pooch!

[@mohammadr797]

TheAmmoniacal I think it mean's null or nothing in farsi

[@TheAmmoniacal]

The Signal Path Blog Haha, good. Got scared for a moment! (Was told that Pouch means void/empty in farsi, maybe related?). Great content, love the videos!

[@BMRStudio]

You are a freeeeeeeaky lucky guy :) Simple power failure. Daaammm... :) like find a gold on the street

[@feedback-loop]

BMR Studio What power failure? Did you pay attention? It was a memory chip.

[@BMRStudio]

Yes You are right, but just half. Most of the memory failure is coming from power failure. And when I wrote this message, I just on the first few mintes in the video. So He has to check again the power supply unit before he got a new memory issue :) But maybe the new modules will be ok. The old modules by the time get more sensitive for the power voltage drifts. I think this error is generated by owrshoot when the accu is uncharged around 5%. This is always happened with dc-dc converters witch are using small inductors with high frequencies. So yes I know is memory issue, but not one bank, almost all the bank. The reason still somewhere in the power module.

[@RupertHandford]

I would give it a good home if you don't need it!

[@TheAmmoniacal]

First!

[@phoenixcrp]

We have a hardware failure. YAY...