Really nice debug and repair. It is pretty amazing how one bad cap can take down the entire supply. Nice work isolating down to a single transistor fault.
No different than a car with no fuel. Some components, though of common type, are pretty crucial. Without them out it just doesn't work Or the navies working on the railroads. Common laborers, but without them we would not be where we are. Lowly but very crucial.
I’ve seen a diode somewhere take down a whole piece of RF kit (i.e. mixer, amp circuits and the 12V PS rail) b/c it decided to become a wire instead. Wasn’t a HP but a big brand transceiver.
Excellent diagnostic techniques. Impressive how fluid your knowledge of the tech and theory of operation blend together to take you step by step until you find the solution. That can’t be taught and can barely be learned. Skill like that is embedded in the soul . Great video. Thanks
I always love how you use the schematics to pinpoint the problematic component, but I think this is one of the best examples so far. The way you excluded all the working blocks and the thought process used to find the faulty one was really impressive.
I have learned a TON watching him debug equipment and just get a better handle of debugging thought process. I'm not in RF, just audioland and older gear repairs (multimeters, the occasional lower Mhz tube scopes, etc.). However just seeing his approaches to these sorts of really tough repairs is super helpful for that, and to get what some of this higher up gear is even for, or what it can do for you. He basically taught me PLL.
Thank you! The X-Ray machine is a Faxitron MX-20. It is the latest model they made I believe with the largest sensor. Make sure you get the PC with it, otherwise the whole thing is essentially unusable.
Love your videos! Super inspiring! Watched an in-person presentation of yours during the Sunday 6G workshop at IMS 2022. You could clearly tell you were a RUclipsr from how relaxed you were during your presentation haha. I also liked how realistic you were about the technology and it’s future. Didn’t get the chance to meet you in-person but maybe another time.
Thank you for the video. I especially liked the procedure of analyzing the error on the block diagram. Here I could learn a lot! I found the X-ray image of the circuit board very impressive. Kind regards!
Finally, a worthy opponent! Oh well, at least it tried. Thanks for the analysis, very interesting signal path :). Actually, my take on your repair approach is that you always analyze in depth the cause and just then you proceed to the next step. That definitely influenced my repair approach, so I have to thank you. Last time I managed to cut down the disassemble - measure - reassemble - test to only 9 iterations
Always happy when you get the notification of a new video Nice video with all the traditional tsp thought process and component hunting that we all learned a lot from o Thanks a lot for your time and effort Y
I love those repair videos! Especially systematic work and explanation and I was able even in the middle of your explanation pinpoint the wrong component based on your previous words :)
in 1997 when I joined GSM operator, we were fully equipped to ... repair RF components. It was a nice time, we had all the equipment needed. (Un)fortunately nowadays nobody repairs it - worse - BTS are in the cloud (ok, except the RF part of course). Time flies ;-)
Great repair, thank you. It's funny, I just bought a handful of 2N5179 RF transistors that also come in that 4 pin TO72 package. It appears Central Semi also manufactures those too.
Signal hound recently released the SM435B, which is practically a frequency extended SM200 up to 43.5 GHz. Would be nice to see that in action as well! It would also be nice to see a pro/con comparison of the signal hound SM series to a traditional spectrum analyzer. Specifically, the spurious responses, image rejection, and IMD in the front end. Typical spec ans use a fancy preselector, but I believe the signal hound does most of the filtering in software.
@@kevina.4036 It's only the old, cheap SA44 SignalHound that leans heavily on software image suppression. I'm pretty sure the BB60 has preselection (maybe an early one didn't have low-freq preselction?) but the SM series definitely has good preselection. I have one and it's brilliant. The preselector is a sub-octave filter bank, so it's not as narrow as a YIG, but it's blazing fast because they just have to drive some PINs rather than a big honking coil.
From the condition of the VFD, I would guess that this unit has been switched on 24/7 for quite a few years. I wonder if the reduced performance of the fan caused the PSU internal temperature to be slightly high, accelerating the electrolytic cap drying out. It is possible that they original owner never experienced the unlock/unlevel if it were in an integrated test system generating higher frequencies, and never rebooted.
If it's not gone already, it might be nice to file the top of the transistor off (upside down) and have a look at the die to see what could have happened to it. Perhaps mechanical damage (dropped machine) or bondwires blown off.
I picked up 3 sig gens, 2 x 8257d and one of these 83752. I repaired the 8257ds but have been lazy on fixing the 83752. Would be coincidental if it has the same exact issues!
@22:45 , there are some trace on pcb just hanging without any connection , the x-ray show it too , any idea Shahriar why they make such stub trace ?? is it filter function ?
I'm an R&D engineer in a small company. Due to some standards we cover, we have our own manufacturing facility. Whenever they have a problem there, they tend to come to me for advice on how to fix it. I've tried teaching the guys there how to look for problems in first runs. I wonder if showing them your videos will help them. I'm just a bit sick of "well we assembled the board and it does not boot. Gave it a visual inspection and it was all good" *20 mins of skyping later* "OHHHHH ... this IC is actually the wrong way around".
As long as you dont spoonfeed them and they learn thru failures, they will get there. Give them hints about how to solve the problem without pointing out the obvious. Ask them questions that will invoke exploratory thinking. Just a software dev, but that is still the best way to learn (and remember). Problem solving is universal.
At first I thought this would be just another simple mains side fuse having blown. But the caps for the PWM controllers deciding to dry out over time isn't uncommon either, sitting around all those fairly warm components isn't a good life for an electrolytic cap. Especially when PWM chips like these at times takes fairly large gulps of current to start the power transistors. Also, is it just me. Or wouldn't it be great if a frequency counter had a power meter as well inside? Doesn't have to be accurate or high res, but seeing in one go that the signal seems a few (or a lot) db lower than expected without having to dig out another piece of equipment would be rather nice.
@@borisscernovs1952 That is indeed a frequency counter with power meters functionality. Although, its price is a bit steep, but going to 20 GHz explains the cost.
Fascinating! Just wondering what ratio of repair success you have? Surely you just post successes (or interesing failures), but I guess duds will not get posted and sure you must get them.
Good question. I have never heard of there being a issue. There could be a very small possibility that EPROM or EEPROM could have the memory holding time reduced by a tiny amount. We know most components are happy with x-rays, they are used in routine inspection of high end PCBs these days and then there are baggage and customs scanners. Basically most things that are not light sensitive (like x-ray film in an intensifying cassette) might have a spurious soft error or bias change but will shrug it off when the x-rays are turned off.
What is wrong with the Peak tester? It works very well and it is portable. If I wanted to completely characterized the device, of course I would use a different set of instruments.
@@Thesignalpath nothing as such, that's why I said they where nice devices. I have one myself. I was observing that in a room with so much high spec equipment it seems rather out classed. It's like when I go to the gym and do my thing and work up a sweat and then one of the trainers in his off hours does his round. I know I should be happy about my own performance but it still makes me feel inadequate some time. I am projecting this sentiment on the peak tester and find that humorous.
I know what you mean. There is a lot of gear elitism around engineering, especially in the hobbyist communities. It just kinda comes with the territory. It seems to happen in most hobbies, whether it is playing golf or painting a landscape. A lot of people would be pretty shocked to see just how simple / cheap / old / clunky the test equipment is on professional work benches - but it works, and works well. :)
Really nice debug and repair. It is pretty amazing how one bad cap can take down the entire supply. Nice work isolating down to a single transistor fault.
No different than a car with no fuel. Some components, though of common type, are pretty crucial.
Without them out it just doesn't work
Or the navies working on the railroads. Common laborers, but without them we would not be where we are. Lowly but very crucial.
I’ve seen a diode somewhere take down a whole piece of RF kit (i.e. mixer, amp circuits and the 12V PS rail) b/c it decided to become a wire instead. Wasn’t a HP but a big brand transceiver.
it is also true with people....for every 1000 fully functional people there is often just one bad cap (person) who takes down the entire society
Only notification on my phone always make me happy: "The Signal Path" new video
You never get cat videos? Or puppies? 😀
THESE are the type of video i prefer. And your repair videos are the kind of content that makes me still a Patreon. Thank you very much.
Excellent diagnostic techniques. Impressive how fluid your knowledge of the tech and theory of operation blend together to take you step by step until you find the solution. That can’t be taught and can barely be learned. Skill like that is embedded in the soul . Great video. Thanks
I always love how you use the schematics to pinpoint the problematic component, but I think this is one of the best examples so far.
The way you excluded all the working blocks and the thought process used to find the faulty one was really impressive.
I have learned a TON watching him debug equipment and just get a better handle of debugging thought process. I'm not in RF, just audioland and older gear repairs (multimeters, the occasional lower Mhz tube scopes, etc.). However just seeing his approaches to these sorts of really tough repairs is super helpful for that, and to get what some of this higher up gear is even for, or what it can do for you. He basically taught me PLL.
Congrats on your repair! Can you share the make and model of your X-ray machine?
Thinking about a new toy?
Thank you! The X-Ray machine is a Faxitron MX-20. It is the latest model they made I believe with the largest sensor. Make sure you get the PC with it, otherwise the whole thing is essentially unusable.
Impressive how you zero in on an issue so quickly. Makes it so easy to follow along, as well as the block diagram overviews
Love your videos! Super inspiring! Watched an in-person presentation of yours during the Sunday 6G workshop at IMS 2022. You could clearly tell you were a RUclipsr from how relaxed you were during your presentation haha. I also liked how realistic you were about the technology and it’s future. Didn’t get the chance to meet you in-person but maybe another time.
Thank you for the video. I especially liked the procedure of analyzing the error on the block diagram. Here I could learn a lot! I found the X-ray image of the circuit board very impressive. Kind regards!
That was insanely good! The xray picture *chef's kiss*
Imagine a world where not everything you buy is disposable garbage, but stuff you can actually repair... great work, interesting video.
Finally, a worthy opponent!
Oh well, at least it tried. Thanks for the analysis, very interesting signal path :).
Actually, my take on your repair approach is that you always analyze in depth the cause and just then you proceed to the next step. That definitely influenced my repair approach, so I have to thank you. Last time I managed to cut down the disassemble - measure - reassemble - test to only 9 iterations
Always a pleasure when you upload new videos! Will enjoy it later as my evening program. :)
those xrays are jaw droppingly good!
Always happy when you get the notification of a new video
Nice video with all the traditional tsp thought process and component hunting that we all learned a lot from o
Thanks a lot for your time and effort
Y
I love those repair videos! Especially systematic work and explanation and I was able even in the middle of your explanation pinpoint the wrong component based on your previous words :)
in 1997 when I joined GSM operator, we were fully equipped to ... repair RF components. It was a nice time, we had all the equipment needed.
(Un)fortunately nowadays nobody repairs it - worse - BTS are in the cloud (ok, except the RF part of course).
Time flies ;-)
hf brain jogging at its best! Thanks for your explanations and dedication. So much joy.
Great repair, thank you. It's funny, I just bought a handful of 2N5179 RF transistors that also come in that 4 pin TO72 package. It appears Central Semi also manufactures those too.
None of my repairs are ever straight forward like these,, you make it seem like magic..Thanks for such a great video.
Thank you! I love repair videos. Nice addition to your bench.
Cool I liked that repair one cap and one transistor all done!
Siempre es un agrado ver vuestros videos
Man, that Signal Hound is so fantastic and it looks like they have put some effort into improving the GUI.
Signal hound recently released the SM435B, which is practically a frequency extended SM200 up to 43.5 GHz. Would be nice to see that in action as well!
It would also be nice to see a pro/con comparison of the signal hound SM series to a traditional spectrum analyzer. Specifically, the spurious responses, image rejection, and IMD in the front end. Typical spec ans use a fancy preselector, but I believe the signal hound does most of the filtering in software.
@@kevina.4036 It's only the old, cheap SA44 SignalHound that leans heavily on software image suppression. I'm pretty sure the BB60 has preselection (maybe an early one didn't have low-freq preselction?) but the SM series definitely has good preselection. I have one and it's brilliant. The preselector is a sub-octave filter bank, so it's not as narrow as a YIG, but it's blazing fast because they just have to drive some PINs rather than a big honking coil.
From the condition of the VFD, I would guess that this unit has been switched on 24/7 for quite a few years. I wonder if the reduced performance of the fan caused the PSU internal temperature to be slightly high, accelerating the electrolytic cap drying out. It is possible that they original owner never experienced the unlock/unlevel if it were in an integrated test system generating higher frequencies, and never rebooted.
please review SM435B from Signal Hound
Will do.
thank you so much for the repair video.
If it's not gone already, it might be nice to file the top of the transistor off (upside down) and have a look at the die to see what could have happened to it. Perhaps mechanical damage (dropped machine) or bondwires blown off.
I picked up 3 sig gens, 2 x 8257d and one of these 83752. I repaired the 8257ds but have been lazy on fixing the 83752. Would be coincidental if it has the same exact issues!
Excellent video, thank you!
Excellent.
@22:45 , there are some trace on pcb just hanging without any connection , the x-ray show it too , any idea Shahriar why they make such stub trace ?? is it filter function ?
Thanks again
I'm an R&D engineer in a small company. Due to some standards we cover, we have our own manufacturing facility. Whenever they have a problem there, they tend to come to me for advice on how to fix it. I've tried teaching the guys there how to look for problems in first runs. I wonder if showing them your videos will help them.
I'm just a bit sick of "well we assembled the board and it does not boot. Gave it a visual inspection and it was all good" *20 mins of skyping later* "OHHHHH ... this IC is actually the wrong way around".
As long as you dont spoonfeed them and they learn thru failures, they will get there. Give them hints about how to solve the problem without pointing out the obvious. Ask them questions that will invoke exploratory thinking.
Just a software dev, but that is still the best way to learn (and remember). Problem solving is universal.
@@leppie This works great when a designer has problem. Tried and tested.
For production people ... not really.
At first I thought this would be just another simple mains side fuse having blown.
But the caps for the PWM controllers deciding to dry out over time isn't uncommon either, sitting around all those fairly warm components isn't a good life for an electrolytic cap. Especially when PWM chips like these at times takes fairly large gulps of current to start the power transistors.
Also, is it just me. Or wouldn't it be great if a frequency counter had a power meter as well inside? Doesn't have to be accurate or high res, but seeing in one go that the signal seems a few (or a lot) db lower than expected without having to dig out another piece of equipment would be rather nice.
There are, indeed, frequency counters with integrated power meter, e.g HP 53150A
@@borisscernovs1952 That is indeed a frequency counter with power meters functionality.
Although, its price is a bit steep, but going to 20 GHz explains the cost.
Fascinating! Just wondering what ratio of repair success you have? Surely you just post successes (or interesing failures), but I guess duds will not get posted and sure you must get them.
I post everything I record, failures too.
Wow...if I were half as good as you I'd have my Tektronix 465B with Short Traces working!
Impressive! Btw that sencore lc103 checker is $2500.
Ingenious...
Great investigation! I wonder how long it took behind the scenes?
why it has a lot of distortion on 100Mhz ? not so clean 100MHz , is it mater ?
What is that metal pencill thing? And where did you get it/where can I get one? :)
A carbide tipped scribe. Used for marking by machinists. They come in various styles.
What does the A16 Sampler do?
It’s basically a harmonic mixer allowing for the detection of the oscillator output frequency/phase for the PLL loop. A poor man’s divider.
@@Thesignalpath Thanks for the reply.
I was wondering, are there components in any of your instruments so delicate that you can't X-Ray them because they could be damaged?
Good question. I have never heard of there being a issue. There could be a very small possibility that EPROM or EEPROM could have the memory holding time reduced by a tiny amount.
We know most components are happy with x-rays, they are used in routine inspection of high end PCBs these days and then there are baggage and customs scanners. Basically most things that are not light sensitive (like x-ray film in an intensifying cassette) might have a spurious soft error or bias change but will shrug it off when the x-rays are turned off.
Cat?
lol how long have you had that roll of radio shack solder hahaha
Making it look easy so a Gen-Z can do it :P
sweeps are nice but swoops are better 🤠👍🤦♂️
ok put your hand up. How many of you went on ebay to search for an x-ray machine after watching this video?
So much incredibly expensive equipement and he still uses as small cheap peak tester. They are nice devices, but it still seems weird.
What is wrong with the Peak tester? It works very well and it is portable. If I wanted to completely characterized the device, of course I would use a different set of instruments.
@@Thesignalpath nothing as such, that's why I said they where nice devices. I have one myself.
I was observing that in a room with so much high spec equipment it seems rather out classed.
It's like when I go to the gym and do my thing and work up a sweat and then one of the trainers in his off hours does his round. I know I should be happy about my own performance but it still makes me feel inadequate some time. I am projecting this sentiment on the peak tester and find that humorous.
@@bzuidgeest I see! Thanks.
I know what you mean. There is a lot of gear elitism around engineering, especially in the hobbyist communities. It just kinda comes with the territory. It seems to happen in most hobbies, whether it is playing golf or painting a landscape. A lot of people would be pretty shocked to see just how simple / cheap / old / clunky the test equipment is on professional work benches - but it works, and works well. :)
🅿🆁🅾🅼🅾🆂🅼 😇