TSP
ฝัง
- เผยแพร่เมื่อ 27 ส.ค. 2024
- In this episode Shahriar repairs an Agilent 8562E 13.2GHz spectrum analyzer which does not power on. Upon pressing the power button the power LED briefly flashes and turns off. Teardown of the unit reveals a compact instrument where the power supply is rather difficult to reach.
The power supply block diagram shows various internal block functions which include startup, DC-DC converter and output regulation stages. In order to diagnose the instrument, the CMOS driver ICs were removed and swamped. An error during the installation of the gate drivers caused a cascade failure of the supply components! After replacing all the affected components as well as the DC-DC converter PWM controller IC, the power supply is repaired.
After powering on the unit it becomes clear that there are several internal self-test and alignment errors. The errors point to YIG oscillator failure. The YIG oscillator no longer has the full range of frequency tuning. The core of the YIG block is replaced with a spare unit from a different Agilent instrument. After re-alignment of the YIG the unit functions correctly. The principle operation of the YIG oscillator is explained and the internals of the defective unit are examined under the microscope.
The Signal Path
www.TheSignalPath.com
/ thesignalpath
www.Patreon.com/TheSignalPath
Really appreciate your honesty when making mistakes...we all learn that way and hopefully they're not catastrophic ones (but even from those we learn "the hard way"!). Thanks for sharing.
I know this is a 5 year old video but I might add something interesting here. Years ago I needed to repair a Tektronix 475 scope that would not power up. I don't remember all the detail but I did determine which supply was not coming up and, of course, as it tried to power up, it shut down from over current as sensed by the regulator board. I ended up jumpering out the regulator board and when I turned it on with no over current protection, a little bypass capacitor somewhere on the PCA exploded - that was it - a shorted capacitor and that "fixed" it. I removed the jumper from around the regulator board and it worked just fine. A shorted .01 or .001 bypass capacitor on the buss that exploded when full PS voltage was applied to it. That's fixing a problem by pulling a rabbit out of the hat...
I love these 856#X (# is 0 to 5, X is A, B, E, and EC) series analyzers so much. Everything about it says portability (a huge plus over a 8566B), reliability, and servicability on it. Meanwhile it has specs I would label as "lab grade". I love it even more with CLIPS being widely available. This is the go to spectrum analyzer for those that can't afford buying a $10,000 instrument or even a $5000 one. I typically see these sold at a steal of a couple of hundred to around $3000 on ebay. I highly would recommend getting one as a starter or for educational purposes. The only things I don't like are the RYTHM and the fact that it isn't modern enough to include real time spectrum analysis. Its expensive or very labor intensive to repair a RYTHM. Its noted that over the years, for some reason the YIG balls will come out of alignment (heat affecting the epoxy keeping the YIG in place?). This ends up causing an "attenuation" of a high band signal (above ~2.9GHz). About the video you were right to assume that the YIG oscillators are the same with different bias boards. The microcircuit inside is interestingly a bit different but should put out the same frequency range. Power level might be a different story. It will still be a suitable repair because HP tends to reuse parts. Currently I'm in the process of repairing/replacing a low band assembly's ceramic microcircuit amplifier. I've ID'd several other units which I could obtain this amplifier from. So far I count 6 units. From VNAs to tracking gens to even a stand lone RF amplifier unit. The RF amp unit is one I hope to get to not only replace but also repurpose as a testing rig for these ceramic amplifiers. As for the repair of the power supply, you did a great job not only acknowledging your mistake but explaining how it would lead to failure of other components. Its a very good thing to document everything thoroughly. I'd like this video more than one time if I could.
Again a very interesting video!!! The adjustment of the new YIG would be interesting for me too. I try to repair a 8593E and maybe it is a YIG-failure. I hope I´ll find it out! Thanks a lot!
20 years ago I used one of these extensively. Been looking for a spectrum analyzer and thought of getting one like this for nostalgia reasons. I'm beginning to rethink that logic. Thank you for the video.
Nice video Shahriar. We've all made mistakes that sometimes make the repairs more extensive than they would've been otherwise - nice to see that it also happens to the best of 'em. I wonder what the original problem was with the supply before the lethal blow. It would've have been interesting (to me) to see the details of the YIG alignment process. I like the look into the oscillator - I've never seen the inside of one of them.
Thank you. :) I can always give you the YIG oscillator if you like to take a closer look at it or have play with it in your setup.
One of the best tear down videos I am seen in a long time. TY Shahriyar.
An excellent example of a professional at work.
These videos will continue to help people for a long time as these topics never get expired.
Thanks for posting.
Drinking Sauternes 1999 and watching a guy repairing Agilent spectrum analyzer on Friday evening, I love my life
Wow 75% of what your talking about I don't get but it's amazing thanks for the video I enjoy every minute of it
If I can ever understand 1% of what he says through osmosis I'm thrilled!
I’m a tech and I get about 75% as well lol no worries
amazing alien microwave thing in the end, thanks for sharing with us!
Wow, that unit is packed for sure. There is something special about old gear :)
10:54 you should replace the 4 transparent RIFA suppression caps on the input they are replace on sight parts in vintage equipment. The transparent (epoxy?) cases get hairline cracks in them. Moisture then gets inside causing the plates to corrode and short.
When they do decide to fail and trust me they will there will be a lot of bad smelling smoke.
Yeah, do replace those. They are nasty!
When they go...the smoke will linger in your lab for a week. It is horrible.
@@Factory400, Oy, and the noise! LOL
TRUE THAT!!! My TEK2445B smoked so bad I was headed for the garden hose - but instead put it on my covered porch room and ran for my special screwdrivers set! The smoking failure was exactly the clear plastic cases on the defective "Y" suppression caps and age/moisture caused dielectric failure and 115V AC to the 3rd wire ground! BTW in this magnificent Video, the 3rd wire, green chassis ground ring lug was left OPEN; should have been grounded to the power board. unless he was using isolation transformer. Also when poking around on any Primary voltages side, a much less expensive or battery Fluke scope would have been less risky! ALSO, follow rhe old Ham Radio adage to keep your other hand in your pocket, (and wear some insulating gloves). I will do Patreon here like the ham/repair genius Mr. Carlson's lab the past year. These guys have the key to encourage and inform us Experimenters HOW to fix and rescue great old test equipment designs which were still field repairable maybe into the 1990's - too proprietary now!!! Thanks, W4HDL.
Thank you for showing the YIG. I did not know it is that small.
Thanks for another nice repair video.
The whole new adjustment which you talk about at minute 31:00 would surely also have been a very intresting part to see in the video.
Amazing to see it running again! My 8563E‘s power supply still waits for its repair, I assume it has resistors which have gone high resistance and it‘s a pain to service. If you interested I can send you schematics and component positions of several 856x power supplies
What I wish is that I had an hdmi converter so I can see your schematics on my larger TV screen. In fact I may get one of those new oled 65" and I could see the schematics clearly since I am 73. Keep up the good work. Excellent. Tvs have such high resolution it is rediculous for regular TV but for me zzahh zzahh.
You produce so much high quality content with every single video, thank you very much! When i graduate i will become a patreon :)
Also i was really excited to see the inside of a yig oscillator! It was definitely a nice addition to this video.
A YIG oscillator is a cool thing.
Awesome! I learned something today, using a scope to troubleshoot a short! And the YIG stuff was fascinating, thanks for sharing your knowledge, I look forward to more :-)
Fantastic video thanks. I must find myself a YIG to play with now
I'm about to cry too. I think there's a problem with the voltage doubler for my YTF that's causing ERR 334 in my 8563E that was working just fine 6? years ago the last time I turned it on. I think the caps are on their way out. Wish me luck!
Unfortunately you have made an other mistake in the power supply. The capacitor labelled as one is placed back the wrong way around.
Yes, you are right. That part didn't make it into my edit. I changed it before I closed the unit. Good catch!
Extremly interesting video. I wondered what a YIG looked like. Great repair.
It takes a spectrum analyzer to fix a spectrum analyzer. A conundrum!
Beware of the RIFA safety capacitors across the mains. I have several units where these have fused. Especially if the unit is put on 240 Volt mains after a life on 120.
So glad I watched the whole video! This by far is my favorite TSP yet. The YIG teardown was completely unexpected when watching.
Was that a coupled line coupler with a dielectric brick for tuning at the output of the YIG oscillator?
You deserve more subscribers for this content.
I learned a lot. Thx
Really a outstanding video. Loved all 41 mins. I really appreciated you including the ups and the downs (lessons I'll remember!). SUBSCRIBED :)
Nice video. I currently have an 8563A with a dead power supply. The rectifier section is ok, howwever I was stuck as to where to look next. So watching yor video will help a great deal. Two questions though, first what was the yellow lead, are you joining two grounds together? Second any idea where I can get the full repair manual for the 8563A? I can only find incomplete information on the internet so far.
love the frontpanel layout of this guy
How was the pwm driver diagnosed to be faulty? That was the whole point of the repair
"At least three more" ... ok now I know why I can't find cheap defect gear to repair on ebay ;)
you don't have to buy defective gear on ebay, dummy! you can make your own!! :D
@@GeorgeTsiros, Heck, he did... that means I can, too! LOL
CERAMIC IS USED TO TUNE FILTER MICROSTRIP LINES.
INCREASE COUPLING FACTORS BETWEEN LINES, AND ALSO MINIMIZES SURFACE ACOUSTIC WAVE COUPLING.
That SA looks soo alien..yet quite familiar. Great episode.
Excellent video. Just one thing: might be interesting to be more precise about which driver MIC442x to select to avoid misery. Fortunately in my SA, IC marking was not HP references and states MIC4426C. Parts list in CLIP states TSC426CPA. Schematic reads ICL7667CPA (low current driver version: 300 mA). All these quite hard to find in Europe, except "China quality". Will check Intersil EL7212 which seems compatible (and improved) and available at Reichelt...
Thank you! Again great video. Did you try investigate what's wrong with this YIG oscillator?
The few (compared to you) times i've tried troubleshooting faulty (or badly designed) circuits, I am very, very, apprehensive about powering it on and poking around it.
Really wanted to see the steps of the YIG alignment, but the YIG teardown was worth it. Did I see a signature of the technician in the ceramic structure? Who was the manufacturer of the YIG?
Maybe it's on my end, but your video appears to be a little underexposed. Thanks for all the awesome videos!
Is there more information we can read about how you accomplished the alignment? The rest I understand, but without this knowledge, the repair seems incomplete? Thank you!
M@ , I believe the instrument self aligns if you get the frequencies close enough through those VRs (I believe there are three) when adjusted.
The adjustment procedure would be in the HP manuals.
Awesome video. Thanks!
Wow! Thanks for a great video.
Wow! Very cool - haven't seen inside a YIG before. You aren't worried about accidentally shorting something with the ground ring on your scope probe? I did that once - was very exciting. Whenever I probe power circuits since then I have used the little plastic hood.
I am always fearful of the ground ring. Its a pain and end up wrapping in tape
I think it's OK since the power-supply is floating trough isolation transformer so he can't short anything to ground trough that. Otherwise it would be a problem and yeah, one should tape it or use other probing method than freehanding it.
Isn't there an included plastic piece with the probes that isolates the ground? My DS1054Z probes came with it.
PERSONALLY, I WOULD WORRY FAR MORE ABOUT DAMAGING THE YIG SPHERE.
MECHANICAL DAMAGE IS ALMOST IMPOSSIBLE TO REPAIR WITHOUT MAJOR SURGERY.
Where did you find that kind of equipment to be repaired?
This makes me tempted to open up my 8590A to see if I can improve its frequency accuracy, it is off by a couple of MHz...
Would be a nice qol thing to have a sorta captive terminator for that lo output
Hello, I have seen your videos for a long time, they have been really inspired to me. Last week, I was received a used 8563E not working for years. When I powered it up, after the message "PREPARING TO REALIGN LO", it displayed error codes of: 319, 334 and 358. I have looked up the error code with some information about the YTO but to be honest, I can not understand clearly. May I receive any suggestions from you? Thanks in advance!
You have the best toys!
Good video, a faulty yig must mean the specy is uneconomical to repair. The inputs to the yig ass are */- 15v but the yig transistor must have less voltage across it. As someone has commented these secies are old there are much better transistors about now, such as those used in LNB,s
Super , Thank You
Hi Just a question for you today. I successfully repaired a HP 8565A spectrum analyzer. I felt this unit had a very unique problem to share with you and I hope you will give me an opinion on this experience. To make a long troubleshooting procedure short, this unit had the only three identical valued 2 watt resistors , 68.1 k ohm resistors open circuited. These resistors had no sign of overheating and were located in diverse portions of the instrument. The only thing in common was their power rating and value, all 1% resistors (all open circuited). Could this have been a manufacturing error with the resistors?
Yes, could very well be an issue with that particular lot of resistors. It happens a lot more often with capacitors but resistor manufacturing defects are not unheard of.
Can you tell me how to remove the front panel?, does that hinge without removing all of the connectors?, A colleague has one with a damaged 2.4mm connector
Wow!! Thank you!!
Is the Narrow Bandwidth option a hardware upgrade or software upgrade?
thanks for the video. if you mentioned it, i missed it, what was the year or around what year that the unit would have been made?
My 8596e had empty four memory bank..can explain me that function ..thx.
16:19 "interesting!" :D
👍👍
Great video, thank you.
Brilliant
Thank you for another excellent video.
Did you try using hardware accelerated encoding for your screen capture? A decent GPU with a dedicated .x264 encoder should have no trouble capturing your desktop at 60 fps.
Why was Shahriar using LoZ mode to measure the rectifier voltage? Was it to allow the caps to discharge after the test was over?
Hi, I have recently bought a used Hewlett Packard 8592A Spectrum Analyzer and I am wondering what happens if I don't terminate the 1ST LO Output ?
I see the unit you have is the 8562E and it says TERM in 50 Ohm. Should I terminate my 1st LO with 50 Ohms also?
My unit does not have the term message written beneath the LO connector.
I have not found anything in the manual yet to advise one way or the other.
Thanks.
Generally it is good practice to terminate the LO port when unused. It prevent poor VSWR from impacting the LO distribution amplifier.
@@Thesignalpath Thank you for your reply. I have terminated it with 50 ohms.
Question have you ever repaired a Yig, long time ago I did open a Yig filter to observe crack around the print, I was told to send up to pic & place repair side of the factory where they totally over soldered the socket, a broken solder joint was cover it up with Solder to the Yig :( went back & I observed the mess, one of those that I wish I had done the job, which scrapped the HP BER :( if I had done it it would have worked, but the manufacturing plant had a system in place. :(
Hi Shahriar
I have some ldi laser module and wanna know how to test them without damaging before fusion to my fibers ! both section of them laser diode and photodiode!
please do some video or give us an answer
the wave length of them is 1310 nm typically
Thanks Great Shahriar :)
Hello, You restore this YIGO oscillator ? (Olá, Você faz a restauração deste oscilador YIGO ? )
I had a 8590 that took an hour to get the 10MHz right.
Does anyone know if the reference is heated and if the heating might have been broken without it generating a message?
good afternoon, i have an AGILENT 8594E with a defective YIG, could you guide me on its replacement or repair, because here in Brazil we don't have companies specialized in this component. Grateful for the attention.
Olá, vc ainda tem esse yig? Venderia ele sou de Curitiba.
And no explanation for the yellow jumper wire on the PSU.
But I do think the buck conversion followed by an isolated DC-DC is interesting.
What are the advantages with this method? Or is it done purely for the fact that it is easier to make an efficient buck converter that can work with a large verity of input voltages?
Anyway, nice video, and I do hope more interesting content is one the way.
my guess is connecting two grounds that are isolated on the pcb but would be connected through the chassis
That were my guess too, were just thinking it is silly to not explain it in the video. As it could be confusing for people that doesn't think of it.
On its own buck conversion followed by an isolated DC-DC is less efficient than just using isolated DC-DC, but buck conversion act in same time as PFC (with lower power factor then if they used more common boost PFC) which give overall better efficiency then combination of boost PFC and isolated DC-DC (in time when that power supply was designed, now difference is negligible). In that time best overall solution was SEPIC followed (or buck boost converter) by isolated DC-DC (power factor same as boost PFC efficiency same as buck PFC), but is hardest to design and most expensive.
All that solutions are/was useful for small to medium power supply's.
Alexander Wikström the two wires on the bottom were to essentially function as a switch bypass. So that he didn't need to attach a switch to the PSU. This psu, much like the old AT PSUs, switch the entire mains input and do not use a standby function.
Could you just RTFM! Service Manual states linking both earth TP301 and --Vrail TP101 non isolated lower power line for the maintenance procedure. This makes it MANDATORY to use an isolation transformer (and NOT autotransformer)!
Since the YIG is a basket case, why don’t you try probing around to see if you can find the failure mode?
The failure depends on frequency, the frequency depends on magnetic field, and the magnetic field depends on the the coil assembly, which when in position makes the oscillator difficult to access. Perhaps one could scoot Neodymium magnets around until it starts resonating, continue to scoot until frequency drops below the minimum working frequency, and begin debugging? In any case, it wouldn't be trivial, but that hasn't stopped Shahriar in the past :-)
This big white cap on YIG - under microscope it looks like it`s cracked. Just guessing.
Do you have an electrical diagram of HP Spectrum Analyzer?
Interesting stuff again, thank you :)
It's a pitty you didn't filmed the failure.
It took a fraction of a second and didn't produce any visual result. However, there was a loud audio result which came from my screams of regret.
Not the best supply design, no gate resistors
The gate drivers they used have the resistors built-in. It says in the datasheet "7 ohms output impedance".
Vgs @ +3V...?
29:30 DUDE, you just happen to have that exact same YIG lying around??? I know you're into that kind of stuff, but come on, what are the odds?
Well considering that HP/Agilent used the same YIG in many of their instruments it is not that unlikely to find a spare. :)
Adds up. BTW, now that I got your attention... A while back, you said you're thinking about doing an episode on the design of RF filters on PCBs (like in the front-end of most spectrum analyzers). Is that still on your agenda? I'd love to see what goes into designing such elements and what determines their geometry (and maybe which tools can simulate the frequency response).
I don't remember what exactly those are called, but I hope you know what I'm talking about.
I absolutely love your videos, thanks for sharing the knowledge!
Hi Shahriar, how can I contact you.Thanks.
You can send an email to TheSignalPathBlog@gmail.com
I did, no response!
What was the topic of the message? Or subject of the email?
The Signal Path N5230A Netwok Analyzer, sent by raymondhashemi@hotmail.com
I hate this item when i calibrate it
how do I become as smart as you?
Died during The Incident....
Thanks for that. Good job.