Excellent repair series! Very interesting. I've only experienced resistor faults a handful of times. The last time was a 4.7ohm through-hole gate driver resistor opening up inside an AC PSU, causing two power MOSFETS to blow up (4x parallelled MOSFETs per rail in that unit). No corrosion or dirt, since this PSU was used in a cleanroom. I had a random temptation to measure those resistors as part of the gate drive circuitry inspection. Glad I did, otherwise I'd be chasing that fault for a while. This Agilent PSU seems perfect for analog development. Nice unit!
Nice work, it's good to see equipment repaired. I've personally found a failed-open resistor in the reference section of a E3631A, so I guess it's becoming a common fault.
It's more common than you'd think. Almost makes me wonder if there was a large ESD strike or something as these thin film precision resistors are highly ESD sensitive.
Very nice series. I am presently digging my way through a 66312A. Service manual diagnostics say the interface board (has all of the digital, DAC/ADC, talks to the front panel), actual measurements say the main board (analog board). Has factory options J02 and J03, which include replacement of some components and deletion of others, and make the feedback loop oscillate , and are not covered in the service manual, so it has been an interesting voyage. Your recent videos have been a lot of help, in large part reminding me that the analog side of the supply is repairable like any other analog system. Keep it up. You are doing good things. Thanks.
One of the best! This eye-opener tells me, what I need to go to learn, beginning with the ltz1000 at eevblog, what I already do with PX and FX. But the connection to the outer world to control precise voltages and currents, and control them back through the uP, is the next essential part. Which op's will do the job? All these analog electronics, these KiCad and maybe STM32Cube-things are relative new for me, I feel like i don't have enough time to learn all this stuff, and to be able to implement it in my own projects, anymore ;) But I will! What a fantastic world! Therefore, big thanks to you, Ian ... 👍👍👍🎄
Thanks! I know what you mean about lack of time.....it's dang hard! I have an STM32 project on the go and really like it.....but not CUBE, I just use the Arduino environment.
Perhaps that specific type of resistor is more sensitive to vibration damage? This power supply did suffer an impact due to a fall from great height if I'm not mistaken? I believe part 1 deals with most of the mechanical damage caused by the impact?
It's so useful to see and hear your explanations of how you use the circuit diagrams to guide your testing. It's like watching a maestro applying his art or a master teaching journeymen. Out of curiosity, what are the additional banana cables connected to the +25V output that I see at 20:12?
Very odd, these "precision" resistors. Maybe I'm a barbarian, but these modern supplies are over-engineered, too complex? Anyway, it was a pleasure to see how you find, test and resolve these faults. I wonder - is such (faulty) equipment cheap? Thanks for sharing! Merry Christmas and happy New Year!
Its an older power supply so uses the available technology of the day.....plus, cheaper to use 5 cheap components despite a more complex design than 1 more expensive component. Its all about the BOM cost.
A very interesting journey. Were the problems due to a defective batch of resistors do you think? If so there must be many of these crippled power supplies out there.
First of all thanks for a really informative video covering a lot of the diagnostic path I'm currently going down. I'm restoring an E3631A power supply originally with a '625' error code sorted by replacing the opto-couplers. Next problems was the supply functioned OK on all supplies in CV and CC modes but the monitor function (using the ADC) didn't work. Turns out it was the same resistors 30k resistors open-circuit in the +/- 10V reference supply as you found in this video. Now I've got the +25V and -25V supplies working OK and displaying correctly in monitor mode but the 6V is still displaying garbage either a higher than 6V or 'noise'. All the CV and CC 'mon' and 'ref' voltages seem to check out OK on all supplies. Any ideas of the most likely culprits to investigate next?
Excellent repair. Having schematics causes less hair loss! Why not tape the calibration process? I for one, would love to see how these are calibrated.
The calibration is pretty simple on these power supplies, you just set the voltage in the display to match what it’s output is measured as. I did some similar work on some Agilent E3647A supplies a year or so ago.
All that dirt on the PCBs in combination with humidity might have damaged the resistors over time. Your water rinsing might have done the rest. Precision Thin Film resistors like these have a conformal coating on top of the resistive film to protect it from corrosion. The resistive film is usually structured as a meander, or LASER trimmed with an I or L cut. So it's evident that if the coating was damaged, these structures have corroded and you get an open mode failure. Maybe that dirt was chemically aggressive and also damaged the coating, but that's no so probable. We had a series problem many years ago with a certain supplier of Thin Film resistors, where this coating initially had so called pin holes, because the coating lacquer had a chemical problem. This created exactly the same failure. If you put the failing chip resistors under a microscope, you might see such pin holes. Therefore it's necessary, that you optically inspect and replace all Thin Film resistors from the same brand on both PCBs.
Good info, thanks! The fault existed before I got the PSU, there was a label stuck to the case explaining the fault. There are just 3off those 'green' resistors remaining which i will change out, I have yet to look up the BOM to get the spec. of them. No choice with the water rinse I am afraid, the gunge on the pcb's was just horrific and I didn't fancy any foreign diseases.....LOL
Could it be that constant thermal cycling of power supplies, could contribute to SMD caps and resistors kicking the bucket earlier, than their through-hole counterparts, with their fancy led bends? To me, if the solder used doesn't expand and contract at the same rate as the two ends of a cap or resistor, a lot of force will be changing hands, and the story wouldn't end well.
Amazing job, never knew that we can wash circuits with water and not causing damage to the electronics ❤
Well done, great job. Very helpful and easy to follow.
Haha, it's like someone specifically set this one up for you. A puzzle/labyrinth for you to traverse, unravel and enjoy.
Thanks for sharing!
Excellent repair series! Very interesting. I've only experienced resistor faults a handful of times. The last time was a 4.7ohm through-hole gate driver resistor opening up inside an AC PSU, causing two power MOSFETS to blow up (4x parallelled MOSFETs per rail in that unit). No corrosion or dirt, since this PSU was used in a cleanroom. I had a random temptation to measure those resistors as part of the gate drive circuitry inspection. Glad I did, otherwise I'd be chasing that fault for a while. This Agilent PSU seems perfect for analog development. Nice unit!
Nice work, is very interesting to see equipment repair step by step, good instructor , thank for your information.😊
Nice work, it's good to see equipment repaired. I've personally found a failed-open resistor in the reference section of a E3631A, so I guess it's becoming a common fault.
Great repair!
Very much enjoyed this repair, nice work! Oh and seasons greetings.
Troubleshooting skills at its best! Thanks Ian. 🙌
Happy Christmas, thanks for all the great videos!
Good video.
This is the craziest device I saw with SMD damaged resistors in my life yet ... Who could ever think that large values could be faulty ??
It's more common than you'd think. Almost makes me wonder if there was a large ESD strike or something as these thin film precision resistors are highly ESD sensitive.
I think it's just aging / corrosion. Rare components mean less real life testing.
@@sdgelectronics Steve .. If there was an ESD issue then semiconductors would be the more faulty components than resistors..
Very nice series. I am presently digging my way through a 66312A. Service manual diagnostics say the interface board (has all of the digital, DAC/ADC, talks to the front panel), actual measurements say the main board (analog board). Has factory options J02 and J03, which include replacement of some components and deletion of others, and make the feedback loop oscillate , and are not covered in the service manual, so it has been an interesting voyage. Your recent videos have been a lot of help, in large part reminding me that the analog side of the supply is repairable like any other analog system. Keep it up. You are doing good things. Thanks.
66312A looks like a nice bit of kit.......hope you get that one up & running.
Was also looking forward to the calibration, but great videos. Merry Christmas and i hope your wife has bought you loads of faulty presents to fix! ☺️
Point taken......I'll do a follow up short vid.
Well done Ian🎉🎉🎉
Great videos! Thanks
2 for 2, nice job Ian
One of the best! This eye-opener tells me, what I need to go to learn, beginning with the ltz1000 at eevblog, what I already do with PX and FX. But the connection to the outer world to control precise voltages and currents, and control them back through the uP, is the next essential part. Which op's will do the job? All these analog electronics, these KiCad and maybe STM32Cube-things are relative new for me, I feel like i don't have enough time to learn all this stuff, and to be able to implement it in my own projects, anymore ;) But I will! What a fantastic world! Therefore, big thanks to you, Ian ... 👍👍👍🎄
Thanks!
I know what you mean about lack of time.....it's dang hard!
I have an STM32 project on the go and really like it.....but not CUBE, I just use the Arduino environment.
Happy Holidays Ian! Thanks for the rapid fire videos recently! :)
There's one more coming before New Year hopefully.....it's not a PSU.....Lol!
thanks for taking the time to make such great videos.
Best wishes for xmas and 2021
Perhaps that specific type of resistor is more sensitive to vibration damage? This power supply did suffer an impact due to a fall from great height if I'm not mistaken? I believe part 1 deals with most of the mechanical damage caused by the impact?
Well done Mate!
It's great to see such a detailed and successful repair. 😁
An early Christmas present =D Have a good Christmas!!! Incredible how many small faults there were on that!
Can I please donate this Agilent PSU to your channel?......once I get the precision resistors replaced.......let me know.
@@IanScottJohnston OMG!! How about I buy it from you instead?
@@GadgetUK164 Nope......free only.
@@IanScottJohnston That's incredibly generous of you =O What a lovely bench PSU!!!!
Awesome. You have really nice equipment... Im saving up to get one Fluke 87. Lol
It's so useful to see and hear your explanations of how you use the circuit diagrams to guide your testing. It's like watching a maestro applying his art or a master teaching journeymen. Out of curiosity, what are the additional banana cables connected to the +25V output that I see at 20:12?
Very odd, these "precision" resistors. Maybe I'm a barbarian, but these modern supplies are over-engineered, too complex? Anyway, it was a pleasure to see how you find, test and resolve these faults. I wonder - is such (faulty) equipment cheap? Thanks for sharing! Merry Christmas and happy New Year!
Its an older power supply so uses the available technology of the day.....plus, cheaper to use 5 cheap components despite a more complex design than 1 more expensive component. Its all about the BOM cost.
It's amazing "tutorial like" repair. Really nice job thank you! I wish you all the best and even more subscribers to the new year. THX
Merry Christmas to you and yours
A very interesting journey. Were the problems due to a defective batch of resistors do you think? If so there must be many of these crippled power supplies out there.
First of all thanks for a really informative video covering a lot of the diagnostic path I'm currently going down. I'm restoring an E3631A power supply originally with a '625' error code sorted by replacing the opto-couplers. Next problems was the supply functioned OK on all supplies in CV and CC modes but the monitor function (using the ADC) didn't work. Turns out it was the same resistors 30k resistors open-circuit in the +/- 10V reference supply as you found in this video. Now I've got the +25V and -25V supplies working OK and displaying correctly in monitor mode but the 6V is still displaying garbage either a higher than 6V or 'noise'. All the CV and CC 'mon' and 'ref' voltages seem to check out OK on all supplies. Any ideas of the most likely culprits to investigate next?
Excellent repair. Having schematics causes less hair loss!
Why not tape the calibration process?
I for one, would love to see how these are calibrated.
Point taken, I'll do a follow up video.
Maybe the mechanical shock when this unit was dropped has caused so many resistors to go open circuit?
Maybe the green resistors didn’t appreciate the garden hose treatment?
I'm a bit disappointed that you didn't show the calibration. Excellent video.
I did it in the last Agilent PSU video I think so didn't want to repeat myself. I also need to re-fit new precision resistors when the arrive first.
The calibration is pretty simple on these power supplies, you just set the voltage in the display to match what it’s output is measured as. I did some similar work on some Agilent E3647A supplies a year or so ago.
All that dirt on the PCBs in combination with humidity might have damaged the resistors over time. Your water rinsing might have done the rest.
Precision Thin Film resistors like these have a conformal coating on top of the resistive film to protect it from corrosion. The resistive film is usually structured as a meander, or LASER trimmed with an I or L cut. So it's evident that if the coating was damaged, these structures have corroded and you get an open mode failure. Maybe that dirt was chemically aggressive and also damaged the coating, but that's no so probable.
We had a series problem many years ago with a certain supplier of Thin Film resistors, where this coating initially had so called pin holes, because the coating lacquer had a chemical problem. This created exactly the same failure. If you put the failing chip resistors under a microscope, you might see such pin holes.
Therefore it's necessary, that you optically inspect and replace all Thin Film resistors from the same brand on both PCBs.
Good info, thanks!
The fault existed before I got the PSU, there was a label stuck to the case explaining the fault.
There are just 3off those 'green' resistors remaining which i will change out, I have yet to look up the BOM to get the spec. of them.
No choice with the water rinse I am afraid, the gunge on the pcb's was just horrific and I didn't fancy any foreign diseases.....LOL
On 17:31 You have offset 32mA on current mesurment. I have similar problem on E3633A,
Calibration does not help, unfortunately
Another great video, thanks. Can you use a different multimeter? That blue one isn't very clear, especially the minus sign.
I think you might be right......I'll try an get a new one in the new year.
But the blue one is cult ... 😉✌️
@@hwj1822 yes, it's sad considering what a milkshake duck he turned out to be
Could it be that constant thermal cycling of power supplies, could contribute to SMD caps and resistors kicking the bucket earlier, than their through-hole counterparts, with their fancy led bends?
To me, if the solder used doesn't expand and contract at the same rate as the two ends of a cap or resistor, a lot of force will be changing hands, and the story wouldn't end well.
U7 uses a symbol with open-collector output designation, as if it was a comparator. But per the P/N, it’s an op-amp. Minor boo-boo.
I wonder what happened to the thing.
Resistance was futile