Author Topic: EEVblog #727 - How To Kill An Opamp (Read 22309 times)

EEVblog · « **on:** March 24, 2015, 11:38:26 pm »

It's all fun and games until an innocent opamp gets hurt!
Dave proves he's not the sharpest tool in the shed by killing an opamp in his µCurrent calibration test jig, whilst checking the calibration of course.
Doh.

EEVblog · « **Reply #1 on:** March 25, 2015, 01:28:24 am »

Quote from: wilfred on March 25, 2015, 01:26:28 am

Still happening for me at 480 SD

One other person on youtube complained about this too.
No idea, uploading the same Handbrake output like I always have.\
Must be youtube.

EEVblog · « **Reply #2 on:** March 25, 2015, 01:31:13 am »

Quote from: wilfred on March 25, 2015, 01:26:28 am

Still happening for me at 480 SD

Ok, I see it now at 480p

Hmm, is it just this video?
I did render and transcode this with my new Xeon machine, so that's new installs of Handbrake and Sony.
No issues in my original HD source material I uploaded to Youtube.

orion242 · « **Reply #3 on:** March 25, 2015, 01:42:11 am »

Excellent video! Learn just as much off your fail...likely more, than my own.

I didn't notice any difference in the quality. HTLM5 @ 720p...Flash is evil. When the hell is youtube going to support full res in HTML5??

scopeman · « **Reply #4 on:** March 25, 2015, 01:44:37 am »

Hi Dave,

I feel your pain! Been there done things similar. What Sparky hasn't?

Once upon a time I had a really nice Motorola 1KHz-1GHz RF generator that I picked up cheap at a hamfest due to a phase lock problem. A friend of mine said that he had the Systron-Donner version and that I should bring it over and we could swap the cards between the units to narrow the problem to one board.

Well (as you may have guessed) we pulled the same board from his good working one and I put it in my unit and threw the power switch and smoke rolled out from the board. They were, from initial inspection identical (so we thought) but with a slightly different rev letter IIRC but the board assembly information matched.

As it turns out the Motorola boards (and obviously the frame) had the + and - 15V supply rails swapped. It was done with vias near the card edge as the supplies were on opposite sides of a 22 pin dual row connector so it was not visually obvious. Well we looked at each other and decided we needed a beer (or perhaps a six pack!).

He said "I guess this is our quota". I asked what he meant. He said "I need to blow something up at least once a week to keep my troubleshooting skills sharp". We maybe so. I felt about a inch tall about 1 nanosecond after I flipped the switch. Well we started with one working generator. Now we had none, and no spare working boards.

We went though both boards and replaced all of the chips, put them back in the respective frames and fired both units up and they both came to life. Neither generator exhibited any other faults which amazed me.

Neither of us liked to shotgun such a problem, but we did not have extender cards and the card did not lend itself to probing without them. Sometimes a handful
of cheap chips is the best solution, especially after you run out of beer.

I still have that generator and every time we see each other we have a chuckle about our reaction to the power up smoke event.

Needless to say I was a lot more careful after that event.

Sam
W3OHM

brilitin · « **Reply #5 on:** March 25, 2015, 02:53:15 am »

Hi
Thanks

Sent from my iPhone using Tapatalk

Ed.Kloonk · « **Reply #6 on:** March 25, 2015, 04:35:21 am »

"<----- goes bang."

hahahahaha. We've all done it!

Howardlong · « **Reply #7 on:** March 25, 2015, 08:12:43 am »

Too many times I've done that, but not enough times so that I never do it again.

And when are enough spares enough? Do you get parts anxiety when you've only got one or two spares left?

Say you think you've found the problem, think you've fix it, replace the part with your penultimate spare, switch on and... oops, guess you didn't find that fault after all.... Only one spare left...

Muttley Snickers · « **Reply #8 on:** March 25, 2015, 08:13:38 am »

I only ever watch youtube at 480 and just dont see much difference at higher rates. For me it's not specifically Dave's video's but most video's regardless of the uploader. On some it's the first 10 to 30 second's, this current one of Dave's was one of the worst that I had seen and I paused and then refreshed and it came good immediately.

I watched one of Tom Lipton's just prior to Dave's and that also was low resolution for the first 10 or so second's, it paused then continued at a higher more normal level. I do occasionally get a message above or below some video's about the problematic viewing with a link to the ISP, in my case Optus cable.

DH2ID · « **Reply #9 on:** March 25, 2015, 08:16:10 am »

Hi Dave, good video! You try lo learn from your mistakes.
I haven't kiled an opamp - yet, but I once plugged in my
ham radio/maritime radio transmitter in the wrong way,
at night. I have banana plugs on them... I had the luck
that my YAESU transmitter had a good design and came
to life again after some minutes of nailchewing...
Out at sea tend to be a bit dependent on your transmitter
working 8-)
73, Alex DH2ID

electr_peter · « **Reply #10 on:** March 25, 2015, 08:52:09 am »

On the issue of video quality - I watched on 480 and ~~noticed no irregularity~~ saw few inconsistent glitches at 0:23-0:28 mark. Glitches are on the parts with maximum movement and inconsistent - probably some YouTube coding/flash thing.

However, for some time (month or so) I have such problem with any YouTube videos - on the first try, video either does not upload, does not play, loses few of resolution options (resolution selection goes blank), plays few seconds with no sound, then distorts the image and goes to blank CRT like screen with "An error occurred, please try again later. Learn More". On second try, video plays nicely with no weird effects.
I guess it has something to do with YouTube being too clever with checking available bandwidth and failing at that horribly.

miguelvp · « **Reply #11 on:** March 25, 2015, 08:56:56 am »

Quote from: electr_peter on March 25, 2015, 08:52:09 am

On the issue of video quality - I watched on 480 and noticed no irregularity.

You probably blinked because at the time mark in that previous post it does have a glitch, my player however defaulted to 720p but I re-watched it in 480p and sure enough the bad encoding is there.

Must be a problem with youtube's encoding when they down sample.

Fungus · « **Reply #12 on:** March 25, 2015, 09:19:29 am »

Showing some love to the Fluke 27...!

EEVblog · « **Reply #13 on:** March 25, 2015, 09:55:31 am »

Quote from: miguelvp on March 25, 2015, 08:56:56 am

Must be a problem with youtube's encoding when they down sample.

In that case we have to figure out why.
I have uploaded the exact same file on my 2nd channel:

I see no problem on the HD source file I uploaded.

EEVblog · « **Reply #14 on:** March 25, 2015, 09:58:34 am »

Quote from: Muttley Snickers on March 25, 2015, 08:13:38 am

For me it's not specifically Dave's video's but most video's regardless of the uploader.
On some it's the first 10 to 30 second's, this current one of Dave's was one of the worst
that I had seen and I paused and then refreshed and it came good immediately.

Good to know it's not just my videos. My panic level has gone a notch

Do the other channel videos have 50/60p source material like mine?

rs20 · « **Reply #15 on:** March 25, 2015, 10:00:59 am »

Throwing out a random suggestion to anyone wanting to investigate further: use youtube-dl to download the videos as encoded by Youtube (480p is clearly the one in question here), and try playing it with VLC or whatever your favourite player is. That should help clarify whether the problem is with the encoded data (i.e., bug in YouTube's encoders on their servers), or with the flash/HTML5 player in the client.

miguelvp · « **Reply #16 on:** March 25, 2015, 10:03:12 am »

Well on the other channel the glitch is still there, but I was using 720p so it must be in the original video, but it's just a couple of frames

Edit: But I don't see it at 1080p

george graves · « **Reply #17 on:** March 25, 2015, 11:13:05 am »

It's glitch-ing all over the place for me. Every 3-4 seconds - also goes to black for a few frames, at random. It's like a stobe light. - as if the frames aren't there - good thing I don't have epilepsy.

I have to guess it's part of the 50fps stuff - no other videos cause that for me (firefox, win 7 pro, 32 gigs of ram, i7, SSD)

rs20 · « **Reply #18 on:** March 25, 2015, 11:47:36 am »

Everything looks perfectly fine to me, across 480p/720p/1080p. (chrome, win 7 pro, 12GB, i7, SSD, GTX 670)

Fungus · « **Reply #19 on:** March 25, 2015, 12:04:30 pm »

Quote from: rs20 on March 25, 2015, 11:47:36 am

Everything looks perfectly fine to me, across 480p/720p/1080p. (chrome, win 7 pro, 12GB, i7, SSD, GTX 670)

I think Chrome uses different source video than Firefox (Firefox uses flash, Chrome uses HTML5).

If so, it seems like the problem might be at Youtube's end. Hopefully it will go away.

Excavatoree · « **Reply #20 on:** March 25, 2015, 12:18:13 pm »

Quote from: Fungus on March 25, 2015, 09:19:29 am

Showing some love to the Fluke 27...!

I should send Dave an 8025.

Flump · « **Reply #21 on:** March 25, 2015, 01:22:35 pm »

dave sell me that fluke 27 pleeeease

miguelvp · « **Reply #22 on:** March 25, 2015, 02:37:28 pm »

It's so subtle that you really have to be on the lookout for it and if you blink you miss it.
If I wasn't looking for it because of he previous screen captures I would have missed it.

@george, that's weird because it's the same content so it must be in your end. Maybe the codec your system has goes out of whack after it encounters it and never recovers?

Show a capture to see if it's before or after the one found and if it can be reproduced by others.

AlfBaz · « **Reply #23 on:** March 25, 2015, 03:50:34 pm »

Nearly as dumb arsed as me the other day when I blew not one, but 2 LM399's by hooking the heaters up in reverse

Fungus · « **Reply #24 on:** March 25, 2015, 04:29:07 pm »

Quote from: Flump on March 25, 2015, 01:22:35 pm

dave sell me that fluke 27 pleeeease

The price of those things has gone waaay up on eBay lately:

http://www.ebay.com/sch/i.html?_nkw=fluke+27

Lucky I got mine when I did.

(You can still get the grey ones for a reasonable-ish price, but who wants a grey one?)

Flump · « **Reply #25 on:** March 25, 2015, 06:04:04 pm »

The yellow ones are faster !

Excavatoree · « **Reply #26 on:** March 25, 2015, 06:33:14 pm »

Quote from: Fungus on March 25, 2015, 04:29:07 pm

The price of those things has gone waaay up on eBay lately:
Lucky I got mine when I did.
(You can still get the grey ones for a reasonable-ish price, but who wants a grey one?)

I always wondered why that is.

Quote from: Flump on March 25, 2015, 06:04:04 pm

The yellow ones are faster !

OK, that explains that, but what is the benefit of the black sticker?

mtdoc · « **Reply #27 on:** March 25, 2015, 06:50:32 pm »

Quote from: Fungus on March 25, 2015, 04:29:07 pm

Quote from: Flump on March 25, 2015, 01:22:35 pm
dave sell me that fluke 27 pleeeease

The price of those things has gone waaay up on eBay lately:

http://www.ebay.com/sch/i.html?_nkw=fluke+27

Lucky I got mine when I did.

(You can still get the grey ones for a reasonable-ish price, but who wants a grey one?)

The real value there is in the kits that include the high voltage probe. A Fluke 27 with a high voltage probe included for $75 is still a good deal IMO.

Fungus · « **Reply #28 on:** March 25, 2015, 07:26:44 pm »

Quote from: mtdoc on March 25, 2015, 06:50:32 pm

The real value there is in the kits that include the high voltage probe. A Fluke 27 with a high voltage probe included for $75 is still a good deal IMO.

Yep. For $75 it's still a nice meter to have, even without the probe. Very solid/accurate/safe.

I'm just saying I got mine for $60 sometime last year, in case, with HV probe + all paperwork, in as-new condition.

Edit: I just saw this: http://www.ebay.com/itm/221724077442

He has three ... grab 'em while they're hot!

Fungus · « **Reply #29 on:** March 25, 2015, 07:40:42 pm »

Quote from: Excavatoree on March 25, 2015, 06:33:14 pm

Quote from: Fungus on March 25, 2015, 04:29:07 pm
The price of those things has gone waaay up on eBay lately:
Lucky I got mine when I did.
(You can still get the grey ones for a reasonable-ish price, but who wants a grey one?)
I always wondered why that is.

The grey "FM" models are actually better than the yellow ones - they have an extra chip for true RMS readings.

Excavatoree · « **Reply #30 on:** March 25, 2015, 07:47:16 pm »

The true Fluke 27 type connoisseur doesn't consider the yellow meters FMs, regardless of what the US government or a label says. If it's not TRMS, it's not an FM. (Ok, it's really only me, but I think everyone should agree with me - LOL)

mtdoc · « **Reply #31 on:** March 25, 2015, 08:07:14 pm »

Quote from: Fungus on March 25, 2015, 07:40:42 pm

The grey "FM" models are actually better than the yellow ones - they have an extra chip for true RMS readings.

Good point but be aware some that say 27/FM on the back label (not the front) apparently are not TRMS.

See this for more info on Fluke 27s

HeywoodFloyd · « **Reply #32 on:** March 27, 2015, 10:55:14 am »

Thanks for this interesting little video. I am amazed how my small electronics projects have been resilient - when I have inadvertently picked up the wrong transistor - it didn't work, but the magic smoke did not escape and the component kept working (when used correctly). However I now tend to set the current limit on my power supply to well below 400 mA because that's where the fuse in the DMM blows

tszaboo · « **Reply #33 on:** March 27, 2015, 02:09:45 pm »

So just for completeness:
It is generally not a good idea to run precision equipment without input protection. An ESD event not necessarily kills the device, but it can silently knock it outside the calibration range. Add that it is a bare board handled by people with their hands, hopefully with ESD strap.

I know, that Dave knows this, and he mentioned that he needed the board faster than possible.

Fungus · « **Reply #34 on:** March 27, 2015, 03:25:45 pm »

Quote from: NANDBlog on March 27, 2015, 02:09:45 pm

I know, that Dave knows this, and he mentioned that he needed the board faster than possible.

And that it's only meant for use by testers in controlled environments.

Neilm · « **Reply #35 on:** March 27, 2015, 06:09:58 pm »

It happens to everyone. Many years ago, I was testing a system and reversed a pair of leads from the old calibrator I was using for a test. Unfortunately for me, I had forgotten that the calibrator did not have isolated outputs - the result was it shorted the mains supply out with a nice loud bang - and vapourised a 3mm track.

On the plus side for Dave, he has managed to get a successful repair video made.

drcheap · « **Reply #36 on:** March 28, 2015, 03:56:14 am »

I recently had a capacitor that, during prototyping a mains voltage level circuit, I forgot to bleed after dis-assembly. Later on when I was cleaning up the bench I found the cap, and it wasn't well marked, so I stuck it in my component tester to get a quick answer on the rating. Well I got a quick response from the tester alright...ka-pow!

Nothing like a mains-charged cap discharging straight into the GPIO of a microcontroller expecting 5V

At least I had picked it up by the casing and not the leads! Although I might have been taught a more memorable lesson from a personal zap vs. the $18 worth of tester damage.

AlfBaz · « **Reply #37 on:** March 28, 2015, 06:45:27 am »

Quote from: drcheap on March 28, 2015, 03:56:14 am

I recently had a capacitor that, during prototyping a mains voltage level circuit, I forgot to bleed after dis-assembly. Later on when I was cleaning up the bench I found the cap, and it wasn't well marked, so I stuck it in my component tester to get a quick answer on the rating. Well I got a quick response from the tester alright...ka-pow!

Nothing like a mains-charged cap discharging straight into the GPIO of a microcontroller expecting 5V

At least I had picked it up by the casing and not the leads! Although I might have been taught a more memorable lesson from a personal zap vs. the $18 worth of tester damage.

I had a bit of a fright years ago when I racked out three banks of 600V DC BUS capacitors from a large variable speed drive. Tested the first 2 and saw the DC voltage rapidly falling thanks to the bleed resistor. Didn't bother with the third and presumed they were all discharged after a few minutes. Proceeded to work and put my 6" shifter on top of the 3 bank, I near shit my pants, that banks bleed resistor had failed open

Murphy is indeed a prick


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: EEVblog #727 - How To Kill An Opamp (Read 22309 times)

Share me