Please do more repair videos in the future.. I like the way how you did it step by step and very logical. You're a fantastic man,a good teacher too! I've learned so much from your videos.
If you're beating yourself up over the utility of the video because of the trivial nature of the problem, don't. What you HAVE shown is a problem that will more likely be repeated in the field than if it was some obscure digital processing error. To put it another way, you helped more people with this outcome than you would have with a more rare failure. Thanks.
Every time I watch one of your videos it reminds me of a quote by Arthur C. Clark, "Any sufficiently advanced technology is indistinguishable from magic." Thanks for all your time and effort.
Thanks for another great video Shahriar. I'm a huge coward, so when working on switchmode powersupplies like that, I generally like to run them from a little 12V 50W mains inverter that I have. I run that inverter from a current limited bench supply with a low current limit. Although I could still get a shock I am 100% certain that it will just be a capacitor's worth of energy rather than a power-station's worth. Cheers!!
Another great video Shahriar. For many of your viewers this might have been a common sense repair, but for those less experienced it was quite insightful. Please keep doing these types of videos. I want to stress how helpful it is that you go through the schematic prior to actually diving in with your probes. Good stuff!!
When I work on power supplies, especially linear ones with big power rail filter caps, I always make sure the capacitors are discharged before I do anything else.
Very nice video. You need to take into account that active PFC steps up the voltage quite a bit, so a 115 VAC source might still generate 200++VDC at the cap. Not sure if that particular PSU has active PFC.
stopped by after hearing your amp hour interview, and very glad I did. Your videos are amazing and really interesting to watch. Thank you for sharing them! :)
18:40 - Exactly what happened to me when fixing a SMPS. 200-300V DC right across my chest. The solder on the pins on the back of the board pierced the skin on my fingers and it hurt like hell.
It looks like there was a screw in that place at some time. The heatsink had a round scratch by the hole. I hope you didn't find anything rattling in the case? ;) If it was a metal screw (most likely, because it scratched the heat sink) it was lucky, that it didn't short out and damage much more. Did you check the case for a loose screw? If they didn't put a screw in that place at assembly (which I doubt, because why else would they bother to put a hole in the heatsink), someone should really ask HP/Agilent if they pick their suppliers more carefully nowadays. It's already bad pratice to have a single TO-220 component free standing, but a whole board including a transformer...
I had to repair one of these.. The fuse had blown. I replaced the fuse and ágain it blew with a big bang so I noticed things could be serious, After much searching and fiddling I notice the thermal insulator behind the transistor @23:55 had a bit of burn mark on it like film left when piece of wire disintegrates. I replaced the thermal insulater and everything worked.
Dont you hate intermittent problems? My Hubsan X4 quadcopter drone had a problem where the camera module would sometimes not function properly, and EVERY TIME I took it apart to analyze the issue, it fixed itself after a bit of messing around with the board. The minute it was reinstalled into the frame, though, it started failing again! After weeks of head banging, I finally saw a suspicious lack of solder near a decoupling capacitor. After placing a liberal amount of solder there, it never gave me any problems there after! (this was not actually a manufacturing fault, I had previously modded the board with a microphone and that was very close to that capacitor! So it might have been be when I had to wick away solder from the microphone area to rework it.)
nice repair, is it just me or does it look like someone temporarily repaired it in the past by removing the screw to secure it to the heatsink? The 12v dropped out when it was pushed back to it's correct location. Also seemed to be some flux residue around the pads before you did your desoldering.
It's great that you show the whole process of tracing the problematic board. I have a question - are you sure the heat sink is not connected to ground (27:05)? Heat sinks are usually soldered to ground rail.
You are right that heat sinks are typically grounded. But sometimes when a series of transistors are electrically connected to the heat-sink, they can sit at the full voltage of the drain/collector. In this case, the heat-sink sits at 170 Vdc with respect to ground.
Hi Shahriar, Great video as usual. When you reviewed the Rigol Power Supplies, you mentioned that they were bringing out a software tool for "orchestrating" measurements and commands to them. I assume that was the ultra sigma software. It would be interesting to see how that package works given you have a large number of Rigol tools at this stage. Perhaps you could do an experiment with the control being provided by that software package. cheers Paul
The schematic you showed for the U3845 is very confusing, Vin is connected to Vcc, but I assumed that the Vin is the 170 V and the output I hope has galvanic isolation from the mains? Maybe the schematic you showed is only test circuit using a low voltages, and not mains voltages as is used in the instrument?
Nice repair. That thing is VERY different from my 33220. Even the PSU is completely different. Mine had bad T491 tantalum caps, I replaced them all. And I wonder if I am really really underpaid as an electronics desgin engineer. I could not afford a Rigol DS6104, only got the DS1104Z-S.
Good video as always - thanks! One thing - what did you do to discharge that cap to make it safe? Is it OK to just put a screwdriver across the terminals (with an insulated handle of course) or is such a sudden discharge likely to damage the cap? I guess a 10k resistor would be kinder or does it not matter? Also - do you need to keep the cap shorted while you work in case the voltage creeps back up on it's own due to the mysterious dielectric reforming effect?
Try to avoid discharging the cap with a piece of metal directly on the pads. This can likely damage the pads, traces or even the cap. I used a 10-Ohm resistor for 5-seconds every time I turned the unit off. :)
Very nice video. I didn't know you have DS6104. These top of the line Rigol scopes are very rarely mentioned anywhere (youtube or eevblog forum for example). Did you thought about doing a review and/or teardown?
It is often that SiGe BiCMOS processes do not have regular PNP devices, only lateral PNP devices. As a result, sometimes you are forced to use only NPNs.
Hi, could you let me know if those 3 cans (shielded coils) on the left bottom of the picture (jump to 43:42) have any ferrite core in them or they are empty? I have bought one of these recently and fixed quite a few issues with it and it now seems to be working but someone has been in there and done some repair on it and those three cans are empty. I wonder if they should have a core? The bandwidth of the instruement seems to be off the square wave after 10-15MHz is not recognizable. I appreciate your help
My guess is that the power supply failed to pass its verification test, that the 12V reference board was replaced by the factory technician and that the tech forgot to put the screw back in. The unit then failed due to physical vibration, or thermal expansion.
As always, great video. And damn, compared to the main board, that PSU was an ugly ugly mess. Sure it worked, but damn, someone signed off on that rats-nest?
Surely you repaired more equipment than Dave ;) pls use screen capture software - it will be cherry on top of your high quality videos. Thanks Shahriar!
+Blade Runner Nope, its called "google" In case google banned you, there is it: ridl.cfd.rit.edu/products/manuals/Agilent/function%20generator/33250-90011_service_guide.pdf Schematic with our switch at page 204.
But in this case it was easy, just typed agilent 33250 and third or fourth result was that service manual... Here in poland we dont like when ppl ask w/o trying to find it in google.
He spoke so fast. And I still dun understand why he unsoldered the little piece of PCB out of power supply unit and then resolder it back. No components are replaced so that the whole power supply unit is working without problems.
TheSignalPathBlog It's an easy to fix when you know what you doing, I bet you weren't the first person to try and troubleshoot that. Many people would have given up once they realised that it was an SMPS fault and the capacitors, diodes and transistor were OK and they couldn't get a replacement power supply.
That's a nice piece of equipment let down by an inferior PSU design, Why did that Soft start board have to be a daughter board and not just in with the rest of the PCB. Would have stopped this fault in the first place. Nice video though Shahriar. Would it be insulting to point out that the first transistor you looked at in the schematic was actually a PNP and not an NPN? sorry fort that.......
Last I year I've repaired my TDS540 scope. The problem turn out to be in the PSU. After a while I discovered that the main switching IC was the problem (That IC have a 5V ref voltage out that was not present). But another one and I solved the at least the psu problem. if any body want to know what are the problems I have found I have made 4 part repair blog on eevblog. part 1) www.eevblog.com/forum/testgear/tds-540-problem/ part 2) www.eevblog.com/forum/testgear/tds-540-problem-ii/ part 3) www.eevblog.com/forum/testgear/tds-540-problem-iii part 4) www.eevblog.com/forum/repair/tds540-repair-series-finale I still have the acw memory problem but unfortunally the sram ic isn't more avaible. Also I have mapped the acq mem 4 adress space: www1.tek.com/forum/viewtopic.php?f=5&t=6083
![BRASIL vs. URUGUAY [1-1] | RESUMEN | ELIMINATORIAS SUDAMERICANAS | FECHA 12](http://i.ytimg.com/vi/-hWBAuxF1eY/mqdefault.jpg)
Please do more repair videos in the future.. I like the way how you did it step by step and very logical. You're a fantastic man,a good teacher too! I've learned so much from your videos.
Thanks! You guys can always send me broken stuff to fix. :)
If you're beating yourself up over the utility of the video because of the trivial nature of the problem, don't. What you HAVE shown is a problem that will more likely be repeated in the field than if it was some obscure digital processing error. To put it another way, you helped more people with this outcome than you would have with a more rare failure. Thanks.
Every time I watch one of your videos it reminds me of a quote by Arthur C. Clark, "Any sufficiently advanced technology is indistinguishable from magic." Thanks for all your time and effort.
Thanks for another great video Shahriar. I'm a huge coward, so when working on switchmode powersupplies like that, I generally like to run them from a little 12V 50W mains inverter that I have. I run that inverter from a current limited bench supply with a low current limit. Although I could still get a shock I am 100% certain that it will just be a capacitor's worth of energy rather than a power-station's worth. Cheers!!
Excellent. Great as always. When I was an EE student, I wish I had teachers with your natural ability to explain things. Thanks for the great videos.
Not boring at all Shahriar.. Just like all the other vids..Very interesting.
I like the repair videos. You might think this repair was trivial and boring, but I found it quite helpful.
Another great video Shahriar. For many of your viewers this might have been a common sense repair, but for those less experienced it was quite insightful. Please keep doing these types of videos. I want to stress how helpful it is that you go through the schematic prior to actually diving in with your probes. Good stuff!!
Very good video. Always learn something new. You make troubleshooting very straight forward, and logical.
When I work on power supplies, especially linear ones with big power rail filter caps, I always make sure the capacitors are discharged before I do anything else.
I love repair videos! Specially the part about how to find the problem, without knowing where to start, a little problem can get very big. :)
These repair videos are super helpful. Every comment was useful. Keep up the good work!
Brilliant!
You not only know your stuff but teach in a very interesting manner.
5 stars!
Thanks for the video. You make debugging look easy!
Thank you for continuing to share your knolege
Very nice video. You need to take into account that active PFC steps up the voltage quite a bit, so a 115 VAC source might still generate 200++VDC at the cap. Not sure if that particular PSU has active PFC.
stopped by after hearing your amp hour interview, and very glad I did. Your videos are amazing and really interesting to watch. Thank you for sharing them! :)
Another excellent and well thought-out video. Thanks!!!
18:40 - Exactly what happened to me when fixing a SMPS. 200-300V DC right across my chest. The solder on the pins on the back of the board pierced the skin on my fingers and it hurt like hell.
It looks like there was a screw in that place at some time. The heatsink had a round scratch by the hole. I hope you didn't find anything rattling in the case? ;) If it was a metal screw (most likely, because it scratched the heat sink) it was lucky, that it didn't short out and damage much more. Did you check the case for a loose screw?
If they didn't put a screw in that place at assembly (which I doubt, because why else would they bother to put a hole in the heatsink), someone should really ask HP/Agilent if they pick their suppliers more carefully nowadays. It's already bad pratice to have a single TO-220 component free standing, but a whole board including a transformer...
I had to repair one of these.. The fuse had blown. I replaced the fuse and ágain it blew with a big bang so I noticed things could be serious, After much searching and fiddling I notice the thermal insulator behind the transistor @23:55 had a bit of burn mark on it like film left when piece of wire disintegrates. I replaced the thermal insulater and everything worked.
Great video.
Looks like you dont have the "curse of eevblog" when it comes to repairs.
He never gets easy fixes. :)
Nice video, thanks. I am looking for a function generator like that. Looks like a nice project to repair.
Keep up the good work 👍
Great video as always very informative
Sweet, another video already! Awesome, thanks for sharing :)
And it's a repair video. I really love those! :)
Dont you hate intermittent problems? My Hubsan X4 quadcopter drone had a problem where the camera module would sometimes not function properly, and EVERY TIME I took it apart to analyze the issue, it fixed itself after a bit of messing around with the board. The minute it was reinstalled into the frame, though, it started failing again!
After weeks of head banging, I finally saw a suspicious lack of solder near a decoupling capacitor. After placing a liberal amount of solder there, it never gave me any problems there after! (this was not actually a manufacturing fault, I had previously modded the board with a microphone and that was very close to that capacitor! So it might have been be when I had to wick away solder from the microphone area to rework it.)
nice repair, is it just me or does it look like someone temporarily repaired it in the past by removing the screw to secure it to the heatsink? The 12v dropped out when it was pushed back to it's correct location. Also seemed to be some flux residue around the pads before you did your desoldering.
It's great that you show the whole process of tracing the problematic board. I have a question - are you sure the heat sink is not connected to ground (27:05)? Heat sinks are usually soldered to ground rail.
You are right that heat sinks are typically grounded. But sometimes when a series of transistors are electrically connected to the heat-sink, they can sit at the full voltage of the drain/collector. In this case, the heat-sink sits at 170 Vdc with respect to ground.
TheSignalPathBlog
Thanks for your response, it's good to know you may get shocked by a heat sink!
Hi Shahriar, Great video as usual.
When you reviewed the Rigol Power Supplies, you mentioned that they were bringing out a software tool for "orchestrating" measurements and commands to them. I assume that was the ultra sigma software. It would be interesting to see how that package works given you have a large number of Rigol tools at this stage. Perhaps you could do an experiment with the control being provided by that software package.
cheers
Paul
Slower capacitor drain when other parts fail was good to see. Had read about that, but not seen it.
The schematic you showed for the U3845 is very confusing, Vin is connected to Vcc, but I assumed that the Vin is the 170 V and the output I hope has galvanic isolation from the mains? Maybe the schematic you showed is only test circuit using a low voltages, and not mains voltages as is used in the instrument?
The schematic only showed one half of the transformer. You can tell because the secondary and voltage feedback is not illustrated.
Cool video, thanks!
Nice repair. That thing is VERY different from my 33220. Even the PSU is completely different. Mine had bad T491 tantalum caps, I replaced them all. And I wonder if I am really really underpaid as an electronics desgin engineer. I could not afford a Rigol DS6104, only got the DS1104Z-S.
Still a great scope tbf.
Good video as always - thanks! One thing - what did you do to discharge that cap to make it safe? Is it OK to just put a screwdriver across the terminals (with an insulated handle of course) or is such a sudden discharge likely to damage the cap? I guess a 10k resistor would be kinder or does it not matter? Also - do you need to keep the cap shorted while you work in case the voltage creeps back up on it's own due to the mysterious dielectric reforming effect?
Try to avoid discharging the cap with a piece of metal directly on the pads. This can likely damage the pads, traces or even the cap. I used a 10-Ohm resistor for 5-seconds every time I turned the unit off. :)
Thank you for all your nice and high quality videos! I enjoy them very much. Keep up the good work :)
Nicely done! Where do you get the schematic for the device !!!
I’m new to your channel. Very informative and nicely done! Have you ever had a chance to repair an old hp 3314a function generator?
Very nice video. I didn't know you have DS6104. These top of the line Rigol scopes are very rarely mentioned anywhere (youtube or eevblog forum for example). Did you thought about doing a review and/or teardown?
It shows up on several of my videos. I haven't done a review of it. I have a few items lined up for review, I can try to review after that.
What did you mean when you said that the 2 NPN driver transistors are present because the chip could not really have a PNP?
It is often that SiGe BiCMOS processes do not have regular PNP devices, only lateral PNP devices. As a result, sometimes you are forced to use only NPNs.
TheSignalPathBlog I didn't know. Thanks! Happy TG.
Cheers!
27:03 Wrong, your heatsink is at 0V not 168v DC!
It's connected to the negative/(ground) output of the bridge rectifier!
Hi, could you let me know if those 3 cans (shielded coils) on the left bottom of the picture (jump to 43:42) have any ferrite core in them or they are empty?
I have bought one of these recently and fixed quite a few issues with it and it now seems to be working but someone has been in there and done some repair on it and those three cans are empty. I wonder if they should have a core? The bandwidth of the instruement seems to be off the square wave after 10-15MHz is not recognizable.
I appreciate your help
My guess is that the power supply failed to pass its verification test, that the 12V reference board was replaced by the factory technician and that the tech forgot to put the screw back in. The unit then failed due to physical vibration, or thermal expansion.
Genius!!
As always, great video. And damn, compared to the main board, that PSU was an ugly ugly mess. Sure it worked, but damn, someone signed off on that rats-nest?
nice dso
Surely you repaired more equipment than Dave ;) pls use screen capture software - it will be cherry on top of your high quality videos. Thanks Shahriar!
No need. You can see everything that needed to be seen without complicating the video creating process.
NICE. just an observation: Most of the power supplies in these agilents seem to be terrible!
It looks like you might run out of room for test equipment in your home lab soon, so let's have a giveaway! :)
Sure, I'll try another give-away near the holidays.
👍👍
where did you find the schematic for this?
Wiggle ALL the things!
I wonder, where did u get all that schematics?
+Blade Runner
Nope, its called "google"
In case google banned you, there is it:
ridl.cfd.rit.edu/products/manuals/Agilent/function%20generator/33250-90011_service_guide.pdf
Schematic with our switch at page 204.
Well, quite sarcastic answer:) Usually is hard to find some schematics. I am surprised that Agilent gives .sch files out..
But in this case it was easy, just typed agilent 33250 and third or fourth result was that service manual...
Here in poland we dont like when ppl ask w/o trying to find it in google.
He spoke so fast. And I still dun understand why he unsoldered the little piece of PCB out of power supply unit and then resolder it back. No components are replaced so that the whole power supply unit is working without problems.
Nice but.... I was screaming “Why he’s touching the PSU without checking the mains capacitor first?? 😨
video not in HD ?
Your just lucky it was easy.
It would have been more fun if the problem was more complex. :)
TheSignalPathBlog It's an easy to fix when you know what you doing, I bet you weren't the first person to try and troubleshoot that. Many people would have given up once they realised that it was an SMPS fault and the capacitors, diodes and transistor were OK and they couldn't get a replacement power supply.
That's a nice piece of equipment let down by an inferior PSU design, Why did that Soft start board have to be a daughter board and not just in with the rest of the PCB. Would have stopped this fault in the first place.
Nice video though Shahriar.
Would it be insulting to point out that the first transistor you looked at in the schematic was actually a PNP and not an NPN? sorry fort that.......
Last I year I've repaired my TDS540 scope. The problem turn out to be in the PSU. After a while I discovered that the main switching IC was the problem (That IC have a 5V ref voltage out that was not present). But another one and I solved the at least the psu problem. if any body want to know what are the problems I have found I have made 4 part repair blog on eevblog.
part 1) www.eevblog.com/forum/testgear/tds-540-problem/
part 2) www.eevblog.com/forum/testgear/tds-540-problem-ii/
part 3) www.eevblog.com/forum/testgear/tds-540-problem-iii
part 4) www.eevblog.com/forum/repair/tds540-repair-series-finale
I still have the acw memory problem but unfortunally the sram ic isn't more avaible.
Also I have mapped the acq mem 4 adress space:
www1.tek.com/forum/viewtopic.php?f=5&t=6083
Need test equipment to repair test equipment to have more test equipment. Hoarder.