TSP
Вставка
- Опубліковано 14 жов 2024
- In this episode Shahriar investigates an Agilent E4981A high-speed capacitance meter which fails self-test during startup. The instrument reports a series of errors associated with the CPU board. The teardown reveals a multi-board construction where the main analog assembly employs an FPGA for control and interface to the main CPU board. The unit has a disconnected ribbon cable between the analog board and the main CPU assembly. After correcting the problem the instrument completes self-test without errors.
After the repair the unit is used to measure several known capacitors for functionality verification. Finally the instrument is used to measure a hand-made parallel plate capacitor. The capacitor can accept various types of dielectric material whose impact is studied.
The Signal Path
www.TheSignalPath.com
/ thesignalpath
www.Patreon.com/TheSignalPath
Sometimes you get lucky, or unlucky, depending on your perspective. Nice unit though. Note - I especially like your hand-tool holder that is sitting on the bench behind the unit. ;-)
Well, you always make up for the "uninteresting repairs" by providing us with a wealth of knowledge revolving around the equipment that we may not be familiar with...thanks for sharing!
I would say you are one of the luckiest guys with these repairs but I know after watching your videos you would have fixed it even if badly broken. Love watching your videos. Always informative ! Thank you for sharing your knowledge.
The last communications test set I purchased for parts had the same problem. Had a ribbon cable partially unplugged from the hard drive. Plugged it in and the darn thing fired up with no faults. Now back to searching for another parts unit. :-) Everyone deserves a easy repair every once in a while.
Same here, bought an IFR with bad video only to find the ribbon cable slightly unplugged.
Wow, I wish I had repair problems like this one once in a while. You got a lucky charm over your head Shahriar . Always entertaining and informative. Better luck next time with some non working machine, I do like the longer repairs more, I hate to admit.
I once had a ribbon cable and the ID connector at one ende had one wire that didnt make proper contact. After opening it, shifting the cable 1mm further and closing it back up... works. But finding that took many hours...
Sure, if you can see the connector is lose or just reinsert it thats no big deal. But cables or such connectors can be a pain in the ass.
Love this updated high-speed TSP channel posting frequency :D
And 150$ for shipping is insanity? Eh, try 350$ for small box with Fluke resistance standard from USA to TW. Even little DIP40 IC will cost at least 40$ in shipping to here..
Atto Farads! I know you don’t like opening the front ends because it changes the calibration, but I think all of us watching these precision meter tear downs are dying to see what’s under the can. Could you do a video on what a low noise, low current front end looks like and experiments of how much calibration can be thrown off by dust, flux residue, etc.?
You are a lucky guy! (Americans would say "You lucky bastard") I saw that connector not being plugged in correctly and thought "OMG, may that be the problem?" I think you can even be fairly sure that the calibration is still valid. Good find.
I was screaming at the screen, "Shariar, look at that cable!" Mind you, if anything "real" had been wrong with it, I would have been hopelessly lost.
The variable capacitor experiment at the end was particularly interesting. I thought you were going to go the whole way and actually plot out the relationship between how far the dielectric was inserted and the capacitance, just to show that it really was linear...that would have been even better.
Ha! That's crazy - just the connector. Yes do some experiments! Yet another nice addition to your ever increasing hoard of amazing top-end gear :)
I was trying to determine what the wooden items were in the background when I saw Alan had already commented.
On a side note: changing dielectric on fixed plates is the basis behind aircraft fuel quantity indication systems.
Given the seals were still on the case, I presume it was assembled on a Friday afternoon or something, locking IDE connectors don't vibrate loose...
When I saw the 22 min length of the video, I said it's a damn fuse...
Always interesting, with an good lesson ... always check connectors! Would like to have seen the change in capacitance as small items were placed on top of the cap. Also curious how the Test Fixture was actually connected. It looks like a D9 connector on the far right but cant really see.
You must be the only person in the world complaining about an easy fix ! Hehehe ! Thanks for sharing !
When you get the easy repairs like this, it would be nice if you still kept going with the teardown a bit so we can learn more about the architecture and circuit design,
I am curious. Why no ferrite on the PSU cable to catch random switching noise?
What a GREAT deal!
BTW, $150 shipping for this is not bad.
You got to teach up how to pick equipment for repair.
Femtofarads, that is measuring attofarads, the display changes with your voice moving those PCB plates. Sample at 1MHz and get the data out, and you can make a microphone out of that PCB material.
Why would the series resistance be hundreds of K for the air capacitor made with pcb material? I guess that value doesn't correspond to any actual resistance in the setup.
What, no front end? *sniff*
I 'fixed' a friends amp a few days ago with this exact problem lol. Unseated cable.
Great video! Nice equipment! I have been thinking - why does almost every piece of HP test equipment have an "A" on the end of the model number? I am sure there is a reason - does anyone know? Thanks for the great videos!
David DeVillers
According to an ex HP employee that I know, HP's marketing people assumed that an instrument with an A revision would be judged by the unwitting purchaser to have at least one revision since introduction, and therefore a better unit. They didn't do this at the very beginning, but you are right, now they all seem to be marked with an A or better.
Al
what were the dimensions. I am guessing 25mm x 100mm with a distance of 3mm. with permittivity of air at 8.84 x 10-12 F/m approx 14.75pF
Those are actually shielded bearings.
I am in no way an electrical engineer etc.. but what is a good Q Factor? Reading online, bigger Q is better. But is there a maximum value? Is there a range for desired value?
Its the ratio of stored energy divided by energy loss per cycle, ie higher Q means lower loss at AC. It depends on frequency
can you use this instrument to show the microphonic effect of different types of capacitors ?
I would really like to learn something about capacitive touch, maybe this test instrument can help explain the principles.
Dear sir l have some problem with the power on. the device keeps to the orange led and can't turn it on. (fuse ok)l don't know what
That's a nice capacitance meter! This is equipment, witch i can't afford, because i can't buy broken one (no skills for difficult fault finding). :)
You should see the prices of used Tek 371 curve tracers...they are absolutely insane!
I’m surprised the Q isn’t higher given it’s just air and copper. Maybe at 1 MHz it goes up.
Nice video
Thank you
It appears to not be plugged in all the way, I could be wrong.
Frustrating? I wish my last repair attempt at a docsis tester was that easy. Bought with keyboard nonfunctional, turned out an internal battery spilled its guts all over the RF section, destroying everything.
The left idc ribbon doesn’t look like it is plugged in properly at 8:20
at 5:15 he is opening it for the first time and it pretty much looks like factory
5 decimal places on the pF range. That is nuts.
We use these meters in the mlcc systems we build and a have repaired a couple of them. I have not watched the video yet but the weak point is the power supply. Same thing fries, luckily it’s a stock PSU at digikey.
And watched it. Funny it is a loose cable. Yes, these are very fast, we use them to measure up to 1mil components per hour on the machines.
Oh, and it is completely normal for the agilent ones to boot with the keysight logo. I think all of ours do.
@Kevin A. MPB125-4350RG
12:30 "we are into the femto-Farads" With 7 digits, the last one is displaying tens of atto-Farads. The manual lists the accuracy as 0.042% (420ppm) which is very impressive, but I would only expect the first 4 digits to be useful. (in the 1kHz speed, at 1V, the 100 pF range accuracy falls to 0.055%=550ppm). Why display 7 full digits? You can't even be sure about the last 3.
Still want to see the cpu-pcb.
But!.. easy fix or not, 1 clic on a cable makes it, at least!! triple worthy :D
9:46 He complain about that it isnt broken so he could put some time on a repair. :D
If he does not like it I can break it for him, so he can be happy.
👍👍
I sent you an email about an old Tesla BM 595 programable RCLG meter... did you see it or did I send it to the wrong adress
It's still not plugged in properly.!
Have you seen Robin Renzetti (Robrenz)'s videos on making some very low resistance measurement kelvin clips? watch?v=wwgO6Lg4RZM&list=PLRe7NGdsY7s6sOTsRahdb_sx2oDwGbh_O&index=1 first two vids. Combine a master machinist/toolmaker with high-precision electronics and you get a really incredible device.
heh you know my comments , so far you are so lucky :P
You could've just replace the ball bearings and it would be as good as new.
Антон Южаков I find those bearings hard to source. Also I had a similar fan and no bearing numbers on them.
You don't really need any numbers, just measure it and find similar one. IIRC they are usually marked as 693ZZ.
Антон Южаков Yeah I could look it up in my little bearing book but I found it to not be worth it sourcing these for 2$ a piece plus shipping. I'd rather use a more silent PC fan even though they usually don't have ball bearings they hold up reasonably well because I don't run my instruments 24/7
Well, fan shown in this video cost 9,93 € so bearing replacement is quite reasonable thing to do IMO...
I would take an easy repair like that. My experience is just the opposite. The
11:41
Because there is a "HP" on it.
So what does that mean on this fixture?
Nothing.
So why is it so expensive then?
No reason other then there is a "HP" on it, with other words just extremely overpriced.
$12500!! Holy fuck...
LOL,
seriously :(