*Thanks for the comments! I think as some of you have suggested, the MOSFETs are part of an AC generator circuit to power the sampling heads. Most likely this is transformer coupled into those boards. In the manual it says that the communication is done through pulsed transformers; but it appears that only the optocouplers are used for data.*
Yet another very interesting repair/teardown. I too once got PM3000A and found out that power supply was faulty. Tried to fix it but failed, so, what I have done was, use a standard power supply matching voltages and the current limits. The equipment is working just fine :) I would definitely re-visit the PCBs and the stuff after watching this video for one more time. Great work!
Leave that Cap in there. Remove the heat sink and chop off the part that touches the Cap. If removing the heat sink is too invasive, trim it right in place. Mask everything off and dremel it. Then put a heat insulator in between the heat sink and the Cap. I know I am not there and there may be conditions I don't see. I am only suggesting. :-) I know you will make it right. You are good. So glad it came back to life!! Awesome!! That is an intense unit.
When Shahriar theorizes about a potential fix, I always cheat by checking to see how much time is left in the video. If the video is ~70% over then I know that he was correct. Then I feel guilty about cheating. Shahriar, I hope you feel bad for doing this to me. :)
Great again with a new video, I love your work Shariah thanks always. Amazing how you repair and you find problems and diagnosis faults. You are one my favorites on RUclips doing electronics...Cheers
Every Voltech product seems to be a PITA to service. Even simple resistive heaters can produce substantial harmonic distortion in the current waveform. During each period of peak voltage the temperature of the element rises rises, and during each period near zero volts the temperature falls. This causes enough resistance change 100 or 120 times a second to cause significant distortion. An ordinary incandescent light bulb can produce 20% to 30% THD.
Thank you so much for this hint I have purchased .. not working unit and it would just not measure anything , after replacing that very same cap ( which when measured with rlc was ok) now unit measures and works.
I suppose that the pulse transformers are not for galvanic isolation of the data, but they are intended to galvanically isolate a powering of the board.
around 17:00 im thinking the mosfets on the heatsinks create a high frequency AC to feed those boards with power, the big cap might be to block dc in case a fet fails and the pulse transformers isolate the AC power, because there are diodes near the second transformer. digital stuff might be completely trough those opto-isolators.
Yep this is also what i was guessing. Those small ferite transformers need high freqency AC to get isolated power across to the cards. This is likely what is generating that AC since they can just feed all the transformers from it in parallel.
When describing the measuring boards, you forgot about delivering power into those OPAMPs and ADCs on the isolated section. Those inductors are the transformers to deliver the isolated power there. And the pair of power MOSFETS on the main board is what was generating a few kHz of AC voltage to drive those transformers.
Really nice video, thanks. You seem to be plagues by either really simple fixes on the T&M gear or really nasty ones that cost as much as the instrument. Still, thans for showing us the process
18:39 possibly POM (Delrin) It feels a little like Teflon, but it is black, cheaper than Teflon and a lot more dimensionally stable (easy to machine). It's got 500V/mil of dielectric strength.
And his videos are too long and he talks too slowly. I don't watch him any more. I get too frustrated. Just get to the point. I don't need the basic details. I already know that. Just get on with it.
@@simontay4851 every video gets 10 minutes longer. By the end of the year his videos will easily be 2hrs+ in length and not convey any more info. Even at 2x speed, and skipping through huge swaths of it I literally miss nothing. They're neat vids but yeah very redundant / drawn out. I thought I was the only one who noticed.
Very nice instrument. I have a PM100 with 1000:1 CT and 100:1 split-core CT clamp. Very useful for SMPS evaluation and compliance testing for EN61000. That cap you replaced I suspect is a series coupling cap on the transformer primary for volt second balancing. Probably has quite a high ripple current as well as being right next to a heatsink. Fire the PCB layout guy!
Great video and thank you. I would be curious to see an accuracy comparison with the very simple and inexpensive Kill-A-Watt or equivalent. I know this is about a high end tool and you wouldn't be able to do harmonic analysis but for under $30 you may be able to see what is going on with most of your AC appliances. Thanks!
Very nice! I also thought those pulse transformers must be for power of the ADC boards. Enjoying your videos a lot. Analyzes of different instrument architectures, how they work and so on, thats that I really like about your videos. Maybe for the future, some deep dive into specific high speed stuff: example, what happens if you dont length match and impedance control ethernet lines?
I would guess the higher current ranges you would have to add in external current transformers, as likely the internal shunts are only rated for 5A AC current, as that is a standard industrial control current, and the transformers are very common as well. Would just need to set a current transformer ratio to get a correct reading, and they can have quite a wide bandwidth, as they essentially are low inductance, and running into a near short circuit. Easy to get the 100kHz bandwidth there with common permalloy cores, though the common silicon steel probably will be running out of response at 1kHz. DC you probably will need to use a low voltage shunt, around 100mV or so, with an in circuit calibration.
Do you have any thoughts on why the capacitor blew? The tape on the old capacitor would indicate a known problem. Maybe socketing the capacitor to a different location would have been better?
Wet capacitors don't like heat and i suspect the original cap was not a high quality Panasonic FR series or Chemicon. He should've bent the replacement capacitor away from the heatsink to keep its temp down - thats what i would've done
As Simon mentioned, electrolytic caps fail - usually from heat, sometimes from poor electrolyte chemistry. I'm sure as Shahriar mentioned, the tape is probably there on the original cap more for isolation than anything. The very first thing you should do when diagnosing any electronics is check all the power supply rails, and look for any obvious bad capacitors. Then if nothing visually looks bad, test the caps with an ESR meter to find the ones that aren't visually obvious. SO many things stop working because of bad capacitors. You don't even really need to turn on the scope to fix a lot of stuff.
@5:18 Was that Cat 5? Oh, I hope it doesn't have twisted pairs. ;-) I Love Cats! Yes, they should be altered unless you want otherwise. I just subbed watching your awesome Neopixel video from 5 years ago. I learned much from that. Thank You.
All Voltech instruments seem to have been designed with no regard for maintenance. I have no idea how the development engineers debugged some of their stuff.
Im having a real hard time following all of these design decisions that were made for this product, some parts of it just feel a bit rushed and glued together. Capacitor right next to a heatsink aside
Each channel board has 4 Ferrite Core Transformers to power the different sections of the channelboard. On the channelboard you can not see any power transistors or chips driving their primary side.... I assume the AC for all the transformers comes frome those 2 MOSFETs and the broken Capacitor sits in line with the primary windings of all 12 transformers to block the DC from them. So this repair works for now. But how about the capacitors on the secondary side of those 12 transformers? Maybe some of them tend to have an increased quiescient or startup current which will stress the new capacitor again... Let's see how long this works now. Some day you will find that an over 100 year old Model T will start with a few turns on the crank and drive away. A 35 year old Model S won't make the slightest blink on its display and we will hate those error-prone electrolytic capacitors sitting in several dozen positions on every PCB.... We really need to invent some capacitors with a significant better lifespan than electrolytic capacitors but the energy density must stay comparable.
*Thanks for the comments! I think as some of you have suggested, the MOSFETs are part of an AC generator circuit to power the sampling heads. Most likely this is transformer coupled into those boards. In the manual it says that the communication is done through pulsed transformers; but it appears that only the optocouplers are used for data.*
5:17 Thank you for the uninterrupted flow of speech - "there's a cat, there it goes".
Yet another very interesting repair/teardown. I too once got PM3000A and found out that power supply was faulty. Tried to fix it but failed, so, what I have done was, use a standard power supply matching voltages and the current limits. The equipment is working just fine :) I would definitely re-visit the PCBs and the stuff after watching this video for one more time. Great work!
Leave that Cap in there. Remove the heat sink and chop off the part that touches the Cap. If removing the heat sink is too invasive, trim it right in place. Mask everything off and dremel it. Then put a heat insulator in between the heat sink and the Cap. I know I am not there and there may be conditions I don't see. I am only suggesting. :-) I know you will make it right. You are good. So glad it came back to life!! Awesome!! That is an intense unit.
When Shahriar theorizes about a potential fix, I always cheat by checking to see how much time is left in the video. If the video is ~70% over then I know that he was correct. Then I feel guilty about cheating. Shahriar, I hope you feel bad for doing this to me. :)
LOL!! :-)
Thanks for putting efforts in making these videos.
Great again with a new video, I love your work Shariah thanks always. Amazing how you repair and you find problems and diagnosis faults. You are one my favorites on RUclips doing electronics...Cheers
I'm surprised the fault wasn't detected at 5:15 with the traveling wave cat scan. :-)
Great video, love the break down of the boards and the processing of the unit, learned a great deal from just this one video, thank you.
Every Voltech product seems to be a PITA to service.
Even simple resistive heaters can produce substantial harmonic distortion in the current waveform. During each period of peak voltage the temperature of the element rises rises, and during each period near zero volts the temperature falls. This causes enough resistance change 100 or 120 times a second to cause significant distortion. An ordinary incandescent light bulb can produce 20% to 30% THD.
Thanks Shahriar, nice job. Enjoyed the video. You can hear the Burr-Brown engineers sigh around 16:10 ;)
Yay, new TSP vid!
Thank you so much for this hint
I have purchased .. not working unit and it would just not measure anything , after replacing that very same cap ( which when measured with rlc was ok) now unit measures and works.
I suppose that the pulse transformers are not for galvanic isolation of the data, but they are intended to galvanically isolate a powering of the board.
around 17:00 im thinking the mosfets on the heatsinks create a high frequency AC to feed those boards with power, the big cap might be to block dc in case a fet fails and the pulse transformers isolate the AC power, because there are diodes near the second transformer.
digital stuff might be completely trough those opto-isolators.
Yep this is also what i was guessing.
Those small ferite transformers need high freqency AC to get isolated power across to the cards. This is likely what is generating that AC since they can just feed all the transformers from it in parallel.
I am pretty sure TSP considers it low frequency AC, everything below 10 GHz is low frequency to TSP.
@@larslindgren3846 Good point. 100 KHz is practically DC on this youtube channel.
When describing the measuring boards, you forgot about delivering power into those OPAMPs and ADCs on the isolated section. Those inductors are the transformers to deliver the isolated power there. And the pair of power MOSFETS on the main board is what was generating a few kHz of AC voltage to drive those transformers.
Spot on.
Really nice video, thanks. You seem to be plagues by either really simple fixes on the T&M gear or really nasty ones that cost as much as the instrument. Still, thans for showing us the process
18:39 possibly POM (Delrin) It feels a little like Teflon, but it is black, cheaper than Teflon and a lot more dimensionally stable (easy to machine). It's got 500V/mil of dielectric strength.
Most of what I have thought is POM has been white so far.
When Mr Carlson buys a piece of old test gear, half the insides are missing. You should send him some of your luck! :)
And his videos are too long and he talks too slowly. I don't watch him any more. I get too frustrated. Just get to the point. I don't need the basic details. I already know that. Just get on with it.
@@simontay4851 every video gets 10 minutes longer. By the end of the year his videos will easily be 2hrs+ in length and not convey any more info.
Even at 2x speed, and skipping through huge swaths of it I literally miss nothing.
They're neat vids but yeah very redundant / drawn out. I thought I was the only one who noticed.
Very nice instrument. I have a PM100 with 1000:1 CT and 100:1 split-core CT clamp. Very useful for SMPS evaluation and compliance testing for EN61000.
That cap you replaced I suspect is a series coupling cap on the transformer primary for volt second balancing. Probably has quite a high ripple current as well as being right next to a heatsink. Fire the PCB layout guy!
Another great video 👍
Thank you sir 🙏
Excellent video and nice device! Thanks
Great video and thank you. I would be curious to see an accuracy comparison with the very simple and inexpensive Kill-A-Watt or equivalent. I know this is about a high end tool and you wouldn't be able to do harmonic analysis but for under $30 you may be able to see what is going on with most of your AC appliances. Thanks!
Very nice! I also thought those pulse transformers must be for power of the ADC boards. Enjoying your videos a lot. Analyzes of different instrument architectures, how they work and so on, thats that I really like about your videos. Maybe for the future, some deep dive into specific high speed stuff: example, what happens if you dont length match and impedance control ethernet lines?
Interresting video. Has everything someone like me needs : electronics and a cat....
Here I am , the same interest ...
Very good device
You could zag zig - lol - the cap legs to offset it a bit more, maybe even by a lot.
thats a nice piece of equipment.
That cap is way too close to the heatsink - I'd have mounted it further away for longer life
Yes, I did mention that in the repair. It has to be moved. Even the original place for the part was not good.
Black PTFE is available. Certainly not as common but commercially available.
Great video
I didn't think such a cheap glue gun would have anything other than a PTC thermistor. I think even my Thermogrip C300 just has a PTC in it.
22:35 When they go for a triac approach but the model is just slightly to much dollar store to use a zero crossing detector.
I would guess the higher current ranges you would have to add in external current transformers, as likely the internal shunts are only rated for 5A AC current, as that is a standard industrial control current, and the transformers are very common as well. Would just need to set a current transformer ratio to get a correct reading, and they can have quite a wide bandwidth, as they essentially are low inductance, and running into a near short circuit. Easy to get the 100kHz bandwidth there with common permalloy cores, though the common silicon steel probably will be running out of response at 1kHz. DC you probably will need to use a low voltage shunt, around 100mV or so, with an in circuit calibration.
On the label in the back it say 30A.
beautiful
Do you have any thoughts on why the capacitor blew? The tape on the old capacitor would indicate a known problem. Maybe socketing the capacitor to a different location would have been better?
Wet capacitors don't like heat and i suspect the original cap was not a high quality Panasonic FR series or Chemicon. He should've bent the replacement capacitor away from the heatsink to keep its temp down - thats what i would've done
As Simon mentioned, electrolytic caps fail - usually from heat, sometimes from poor electrolyte chemistry. I'm sure as Shahriar mentioned, the tape is probably there on the original cap more for isolation than anything. The very first thing you should do when diagnosing any electronics is check all the power supply rails, and look for any obvious bad capacitors. Then if nothing visually looks bad, test the caps with an ESR meter to find the ones that aren't visually obvious. SO many things stop working because of bad capacitors. You don't even really need to turn on the scope to fix a lot of stuff.
I want you to repair some 4K camera so we can have 4K videos :p
@5:18 Was that Cat 5? Oh, I hope it doesn't have twisted pairs. ;-) I Love Cats! Yes, they should be altered unless you want otherwise. I just subbed watching your awesome Neopixel video from 5 years ago. I learned much from that. Thank You.
All Voltech instruments seem to have been designed with no regard for maintenance. I have no idea how the development engineers debugged some of their stuff.
Im having a real hard time following all of these design decisions that were made for this product, some parts of it just feel a bit rushed and glued together. Capacitor right next to a heatsink aside
Sadly, this is actually more common than you'd expect.
As of viewing there's 616 thumbs up.... 6 for the video and 6 for the cat cameo :P :P :P
Интересно!
I think they put kapton tape on that cap not just to electrically isolate it from heat sink but termicaly as well
Looks like a last minute futile attempt to extend the inevitable failure of the unit just past the warranty period...
@@benbaselet2026 sneaky practices :) but I doubt. this is serious device made by people who know what they do but who knows :D
Hello can you help me, how to activate the other channels? when I press the Channel 2 and 3 it says invalid Function.
These internet cat videos are getting very technical.
5:16 Pooch! Now my evening is complete :=_
Bar graph say 500V pk?
155V*2*1.41 ~= 450V
Great Video, thank you so much. Please add a PayPal link to allow easy payments supporting your Lab/episodes. Regards from Germany.
Each channel board has 4 Ferrite Core Transformers to power the different sections of the channelboard. On the channelboard you can not see any power transistors or chips driving their primary side.... I assume the AC for all the transformers comes frome those 2 MOSFETs and the broken Capacitor sits in line with the primary windings of all 12 transformers to block the DC from them. So this repair works for now. But how about the capacitors on the secondary side of those 12 transformers? Maybe some of them tend to have an increased quiescient or startup current which will stress the new capacitor again... Let's see how long this works now.
Some day you will find that an over 100 year old Model T will start with a few turns on the crank and drive away. A 35 year old Model S won't make the slightest blink on its display and we will hate those error-prone electrolytic capacitors sitting in several dozen positions on every PCB.... We really need to invent some capacitors with a significant better lifespan than electrolytic capacitors but the energy density must stay comparable.
No views. 1st time I'm this early!
Show us how you buy this stuff on eBay. How to choose it and don’t loose money. Show us an example of successful purchase
Please DO NOT say "problematic"!!!!!