TSP #95 - Teardown, Repair & Calibration of a Keithley 2400 Source Measure Unit (SMU)
ฝัง
- เผยแพร่เมื่อ 27 ส.ค. 2024
- In this short episode Shahriar investigates the failure cause of a Keithley 2400 SMU. The unit powers on and responds to controls however produces no output signal. Further examination reveals that the VFD display shows no activity when the output is enabled; normally the measured output voltage and current are displayed. This failure points to a digital problem, potentially a failure in the analog board FPGA.
Upon close inspection the FPGA clocking circuitry appears to have a failed solder-joint. A quick re-flow of the affected part corrects the problem and the unit reports the appropriate readings. The unit is then upgraded to the latest firmware and calibrated using a Keithley 7510 DMM.
The Signal Path
www.TheSignalPath.com
/ thesignalpath
www.Patreon.com/TheSignalPath
You know it's gonna be a good day when The Signal Path pops up with "teardown, repair, and calibration". You know an interesting puzzle is coming, you know it has a good ending, and you know it's going to be a good watch.
Another great video. I really like these repairs videos where you show your thought process as you work through the problem. I'm always amazed at the level of intuition you seem to have as you zero in on the issue.
You are lucky: The USA/Canada Ebay market for instruments is so big, that finding something to repair is not that hard. I comb through Ebay in Europe for months and rarely find anything of that quality. And once there is something people pay stupid prices for broken stuff like 600 EUR for a broken Agilent 34401.
Yes, you are right. The eBay US market has much more variety in it.
The way you broke down the problem was really impressive. thank you for the tips and tricks.
Looks like you ended up with a good working unit after little bit of work. Thanks for sharing.
Mike
If you didn't have so much experience and talent... someone would say you are a lucky man!
Great video, thanks! I had similar problem when repairing audio DAC/ADC. The DAC chip was lifted because there was too much glue under it (the whole board was wave soldered). So it worked when it was tested at the factory, but after some time the thin pieces of solder (that connected legs of the IC and pads) broke off and the whole DAC portion died.
So much want (:
:)
as a non patron I would suggest a random draw with a chance based on the support. someone that supports you 10 times as much has a as much higher chance to get a reward
Check in the camera settings for a volume input control, that will allow you to adjust the volume as its recorded onto the video. Its common for this setting to be in the maze of menus on Sony cameras
Sadly it doesn't have that when this external microphone is attached! The selection for microphone level is disabled.
Sony Enough Said
If you ever open the case then I recommend that you also change the battery which is on the bottom board. If the battery is dead then on some cases the equipment becomes unstable and shuts off after 10-20min.
I don't usually watch entire repair videos but this one really sucked me in because the issue seemed so specific, I was curious to see what the problem was. Great video. What isolation transformer are you using? I've never seen one with a current reading.You are working in a lab environment so I'd expect it to not be referenced to earth ground so you don't become "a part of the circuit" but just to confirm is it a transformer that only rejects high voltage spikes and noise or is it separated from earth (as well)?
This video is a must watch I won't spoil it for anyone else 😂😂😂
How many of these have you pulled off now, 3 or 4 maybe more, you are the luckiest dude! Keep it up!
I use the same Sony Bluetooth mic for my TH-cam channel. One is the transmitter and one is the receiver but really if you use an ear bud on each you can use them as full duplex radios good for about 300 feet or more. I connect the receiver to my camera and after syncing with the transmitter I then push the volume up button on both units about six times to make sure volume is max on both. Then through testing I found giving the receiver about 2 or 3 volume down clicks gives the best audio level to my camera input. One warning. Check your batteries before starting a shoot because if they get even a little low the blue indicators will still be on solid like they are happy but the audio into the camera will be fuzzy crackling and sometimes failed. I get about 3 hours on a battery shooting. But for a simple full duplex radio not recording I go 5 hours on a battery. I made a review on this item.
The signal Path... I watch your videos all the time and I think you are pretty amazing the way you solve these problem, Great Video. I would like to start recording some videos as well and I need to know what camera and microphone can I get to do this that is not too expensive. Any ideas? Also why don't you do like EEVblog and do a random draw. Should be fair.
The relatively short runtime of the video was almost a 'spoiler' - despite the mysterious nature of the fault, you just KNOW that the penny's going to drop at any moment, and it's going to be something simple after all - entertaining and enlightening as always, of course.
Impressive 👍👍
Thank you so much sir i had also same issue and I rectified it with this millions dollars video ;-)
Perhaps you could have verified that the crystal oscillator was in fact functioning using a simple "sniffer" loop and the spectrum analyzer? In less complicated gear I do that as a quick and dirty first check. Glad you saw the IC; I saw the moment the chip sucked back down on the board as you were heating it, that's always an interesting event.
Lucky find, I'm surprised this unit made it through the factory quality control/testing.
With my luck, had I bought that unit, it would have been the most expensive obsolete part that required unavailable firmware to get it working... nice find for sure!
Another great video, keep up the good work!
Some repair places, desoldering disconnect the faulty component and temporarily leave the faulty components without (correctly) soldering, pending/until hang of working with the a replaceable component.
Nice
Good find.
I find funny that the chip date code is matching rest of the board - seems to be really something that slipped trough Keithley output QC.
LOL, that was like The Signal Path, ASMR Edition.
I liked the video, but please do something to this mic :)
Few question from my side:
- I many test equipment one convert very high frequency to a lower one, to make it easier to process. Why does conversion process consists of many steps, for example 2 GHz to 500 MHz to 120 MHz and eventually to 60 MHz. Why cant we convert 2 GHz to 60 MHz directly? What could be potential benefits of taking intermediate steps?
- I recently finished my Master Degree in Nanotechnology. We learned a lot about fundamental physics of semiconductors, how transistors, lasers, leds, photovoltaic devices work , how to manufacture them etc. Now I want to get a job as IC designer. Unfortunately even junior positions demand lots of practical knowledge of specific software, such as Cadence and already an idea how to design digital/analog IC-parts. The problem is, I cant learn how to design IC or to use Cadence at home. Any suggestions how to get such job? (Jesus, that is a lot of text...)
1. Two issues coming to mind:
a) Image rejection - doing 2 GHz --> 60 MHz directly would pretty much require I/Q mixing for any kind of decent performance in this regard, and without good calibration it could still end up being less than stellar. Clearly, you do not want something like a spectrum analyzer to show things that aren't actually there.
b) Frequency generation - it is not at all easy to generate high frequencies with very small steps (say, GHz range with 1 Hz steps), so multiple nested PLLs with a fair bit of internal mixing going on have to be used.
2. That sounds kinda familiar. Companies always want people that they can use as-is without having to invest a lot into them first (which is understandable in a way, if problematic in a field with so many specialized niches), but what you write implies that you may have been at a traditional university without too much practice involved, like yours truly was. You may have to look for an intermediate position to get you some practical experience first, maybe as a support or applications guy or something else more suitable for a neophyte. ("Junior" anything does not necessarily mean it's suitable for someone who just graduated!) If you can drive a car, you should be good.
is there way to measure power out of an audio player (maybe function of audio level)? a multimeter is not fast enough to log that (frequency up 20khz), can be done by math integral function in an oscilloscope?
RMS voltage and current should give you the power output.
You'd normally measure in Watts RMS. The number will be much lower than Peak Music Power Output (or Watts BS as I like to call it) that commercial amps are often quoted in, but more meaningful.
IIRC, classically done by putting a resistive dummy load of resistance equal to the nominal impedance of the speaker that would normally have attached to the amplifier, then putting a 1kHz sine wave in the input and turning it up until it starts to visibly distort on a scope (not the best way to tell if it's distorting, but better than nothing). Then you can just measure RMS voltage with a multimeter or scope and calculate power from voltage and resistance.
How to reset the overtemperaure alarm the moment one switches On
I would really like to have the LCR Meter; I've starting building my home electronics lab - and this would be a very nice addition. I would be also interested in the HP Frequency Counter (HP sold me a terrible 6000V HP Laptop $1200, and almost worthless - HP definitely owes me,IMO!)
Great video, great channel! I watch this video to fix my own K2400, which presented an output issue, but it was helpful. Now I am curious: could you upgrade the firmware to C32 or superior? Mine is C17, but I saw in someplace that it has different hardware, so I cannot evolute to any superior firmware release... is it true? Many thanks in advance.
Did you end up getting a new camera? Even right from the start it looks much sharper than before.
Yes! I did thanks to my Patreon supporters.
How to do internal calibration on 2306...my instrument showing overflow error while measuring current....
Good Stuff
I have the same Sony mic on an a6000. It has worked as well as could be expected based on the low cost. I end normalizing the audio in the end. Not hideous and a certain improvement over on-camera mic.
Yes, that is what I have been doing.
Any idea where I might get hold of a replacement VFD for one of these...Tektronix won't sell me one...!!!
man you got lucky on that one for sure.
That's a strange thing. Did it ever work ?
And what could cause that? Just speed/rushed through the soldering process ?
The large crystal might have shaded the spot during reflow? (unless the crystal was soldered afterwards).
This inside corner is also an area of high mechanical stress on the board, you can see it flexing. Vibration or impact can break weak joints in areas with a high bending load.
ok, cheers
Oh man. As I told several times you are lucky man :)Again same story.
Yes, it must always be luck! There is no other explanation.
The Signal Path I wish I had seen videos like yours about 8 years ago. I had ended up with a stack of busted electronics (a basic analog function generator, some digital/analog guitar gear, and a 70s era tube hifi amp) and I ended up gutting or parting them all out because I didn't have enough knowledge to troubleshoot and repair them properly. I look back now with some regret at the missed opportunities, especially the hifi amp. I know now after watching your videos that I'd have been able to make a much more logical and informed go at it.
As I believe Jack Nicklaus (the golfer) once said: "The more I practice, the luckier I get."
Hai. Let me know if you are selling any of your used equipments..
Hi, if you ever want to sell the measuring source (that is, if you still have it after all these years) and the price would be reasonable, I would be interested - that is, if it is within my financial means. I am from the Czech Republic and the exchange rate of the dollar against the crown is quite a massacre. I lost all the equipment I collected in my life because I was robbed and so there is a lot that I have to buy, I have already put some equipment together, I made a lot of things myself, but I am not in danger of buying a new measuring source.
That glitch was really good 👍
Nice day 🙂Tom
Could you please enable community subtitles on your videos?
This is like watching a Superman movie, you know right from the start how its gonna end. Have there been anything you were not able to figure out whats wrong?
Sure! :)
Something went really wrong with the audio on this one, all the background/power supply noise has been amplified.
Why do you have so many different instruments in your lab?
So I can produce many different experiments, reviews and repairs.
the real question is why HoxTop doesn't have so many different instruments in HIS lab... P.S I love the videos!
You could sell the fixed stuff and use the money to buy more broken stuff to fix to make more fix videos?
I'm only a few mins in, but why not try a firmware update first?
Because not have done a firmware update would not break a instrument.
Altera MAX CPLD, not FPGA.
worroSfOretsevraH Thanks! Slip of the tongue.
"some points will have 400V between them" * continues to poke around with a metal stick *
Not that I'm complaining, but before you start offering things to the highest patereon supporter, have you checked it isn't always the same guy?
Please see my next video. :)
derp. Thanks Shahriar.
Regarding what you talk about in the next vid, please don't be disheartened. Most of us know you're being very generous in sharing your wealth of knowledge with us. A wise person once said "haters gonna hate hate hate hate hate". ;-)
No cat in this video. Thumbs down.
Just kidding. :)
First.. Again...
No
Get help.