EEVblog
ฝัง
- เผยแพร่เมื่อ 22 พ.ค. 2024
- How ceramic backed precision thin film resistor networks are constructed and laser trimmed.
Examples come from the resistor networks in the 7.5 digit Keithley DMM7510 Multimeter, manufactured by Fluke.
Forum: www.eevblog.com/forum/blog/eev...
EEVblog Main Web Site: www.eevblog.com
The 2nd EEVblog Channel: / eevblog2
Support the EEVblog through Patreon!
/ eevblog
EEVblog Amazon Store (Dave gets a cut):
astore.amazon.com/eevblogstore-20
Donations:
www.eevblog.com/donations/
Projects:
www.eevblog.com/projects/
Electronics Info Wiki:
www.eevblog.com/wiki/ - วิทยาศาสตร์และเทคโนโลยี
Hello Dave,
This is an amazing video. I have extensive background in laser material processing in the medical field.That being said, I wanted to point out that the reason for the clear glass cover is for final precision trimming. The clean trim lines have been made with open substrate before the glass he been installed to rough in the values to a closer range than after the glass is installed the network could recieve the final absolute tune in as laser can still travel through the glass with minimal distortion. You can tell by the evident burning marks trapped behined the glass which is soot from the laser vaporized material that would not be possible to evcuate after sealed. I would l also point out the that final triming process could also be done in circuit after board assembly.
that is the difference between a $$$ precision device and a $$$$ precision device.
Cheers.
Final trim is not done in circuit, it's done on each part independently before being assembled.
Yeah! Lasers. I guess you beat me to it.
This was actually a lot more interesting than I expected, thanks Dave !
Ian Clarke Thanks, glad you liked it.
Cheers for all your work, Dave! these close up photos are simply mind-blowing!
Ignyte Thanks, did the best I could with those, it wasn't easy.
EEVblog What camera/lenses do you use for the ultra-close-ups?
Prehistoricman An Olympus E-510 with 30mm macro lens. Hard to get into such gear when the lens extends during focusing.
Very interesting video! I worked in laboratory sputtering thin-film resistors for hybrid integrated circuits crystal oscillators. We sputtered alloy resistive silicon-nickel-ferrum (RS-1004) on substrates from Sital.
Earlier sputtering was performed by the explosion - of a resistive alloy powder poured out of the bunker on a hot plate of tungsten and evaporated in it, then we have modernized, instal a magnetron, and started sputtering target resistive alloy.
Sorry for my illiterate English.
Demyanych88 This sounds quite interesting! :)
Watched this on my smart tv. Amazing quality!
Thank you Dave, a bit thumbs up to you - you deserve a lot for your effots.
I know next to nothing about electronics but your videos are changing that. Thank you sir for taking the time to share your knowledge with others.
Surfing the edge of the tangible and the microscopic... beautiful, love your work :)
I enjoyed this very much. Component technology is extremely fascinating!
Good work Dave. Very interesting indeed.
Dave, this was really interesting! I enjoyed learning about the ladders especially. Thanks for this! Very, very interesting indeed.
It seems I learn something new every day... Thanks again Dave. :thumbsup:
Resistors just got a lot more respect from me, amazing, thanks Dave.
I recently had to deal with ESD materials at work, and didn't know much about the Ohms per square. Thanks for clearing that up!
Great video Dave, I always wondered what high precision resistors looked like and now I know
So simple, but so brilliant! Very interesting.
I have dealt with what we called CPA (ceramic precision arrays is just what we called them in house) and i think you hit the nail right on the head here! great work as always!
I don't know why, but I always get excited when I see components that I recognize! Like the 2n3906 transistor! :)
Daniel Astbury Me2 :-) And sometimes I remember the iron curtain commy era, when It was a pain in the ass to get at least some info about those parts and trying to replace them by the only available soviet and TESLA (CZ) made components.Aaargh what a nightmare...
Stunning! thanks Dave :)
Absolutely amazing topic and informative video. Thank you
I just bought a t-shirt to support you, you earned it.
Whoa, this was way more interesting component than I thought.
I wonder why the resistances drift over time in general...
Brilliant video, just like the "Ultrasonic Delay Lines" some years ago.
Awesome stuff!
That was very interesting! Thank you!
Thanks, each time i learn something interesting.
Great lesson!
great video!
I never get more than a few minutes into these videos before the video ends up in the background and I'm off Googling things Dave is talking about
sandsifter149 go elsewhere.
sandsifter149 Says the actual whiner....
+Dee Jay Same here. If we were computer algorithms, we'd be "depth first search" algorithms.
interesting stuff, thanks Dave :)
thanks dave
Very nice, thanks
Awesomely interesting stuff. Great hi-def pics. Thanks!!!
Really interesting, thank you
You know so much.. Are there really anything that you don't know?!? Great video Dave, keep up the great work =)
Chris Fredriksson Eh, sometimes he doesn't know when to shutup. lol xDSometimes it looks like his eyes will pop out! xD
beautiful
These are widely used in the automotive industry, in components such as : engine management computers , air flow meters,, lambda probes (oxygen sensors ) , electronic throttle units,pressure sensors of all types, . also electronic accellerator pedals which are quite comp!ex variable resistors. Each sensor individually laser trimmed to give the desired accuracy. So when you buy a non genuine part which is half the cost you get problems !
The Vishay one seems physical/mechanical cutout on the bottom instead of a laser trim. Very interesting.
All those zigs and zags must have a significant amount of inductance and capacitance. Are they resonant? Do they ring at some frequency? It'd be great to see S-matrix measurement with a network analyser.
I always thought resistors were simple things. How fascinating when you get to the nitty gritty details
Martin Ziener And there is a lot more detail than what I covered here! The video was already much longer than I wanted though, so I called it quits.
lol, when Dave thought 18 min was long.
Good times.
That second resistor network would make one hell of a strain gauge!
TheBigBigBlues Could well do if it wasn't for the ceramic backing!
Most common example of these I can think of is the blower motor resistor in automotive applications. I wouldn't necessarily call the ones used in cars for that purpose precise though.
Very interesting video. $4000 goes a long way... Imagine all the R&D that went into this multimeter.
My old Fluke 77 even has one of these as well.
I wonder if the 2 different colour materials have opposite tempcos to get an overall near-zero tempco.
mikeselectricstuff i suspect it's unlikely, as the tracking should already be excellent by virtue of the same film with tightly controlled parameters, and that is the inherent advantage of resistor networks in the first place - the very tight ratio tracking. They can drift all they want with temp overall, it's all about the ratio. Vishay combine tempcos on their high end individual resistors though.
But still, it's a possibility, would be tricky business though and a fairly fixed thing though photo masking.
EEVblog My theory is the darker ones have a coating for the the second phase of laser trimming. Perhaps to absorb the laser better or to sequest the material the laser vaporises, so not to contaminate anything else under the glass.
EEVblog Excellent video! I wish You could make more like this in the future.:)
I fully understand what your saying, that trimming the "less" resistive parts of the conductors could achieve crazy amounts of tweaking possibilitys (I think we see some near the pins). But. it seems to be to much effort to do another layer process when it has the ladder networks etc. I noticed the "silk screen" was upside down at the top and the top right has a smudgy browned number cut in.
mikeselectricstuff As usual the know-it-all engineer(s) (ahem) think otherwise. Don't we get praised to "think outside the box"?! :(
Wow I learned a bunch in this video. I kept thinking about the tiny electrons following the shortest path without flying out ;)
I do not get the part about the number of squares making up the resistive value though. I feel like I could have divided that rectangle in a different number of squares than you?
Nice lesson Dave. How can you tell the voltage and wattage they can handle or did I miss that?
Got good laughs at around 09:00 for those red resistor drawings you made. No I couldn't do any better and that was right way to do it, but still somehow got me laughing. Thx
Sometimes they use two different kinds of metal with different temp coefficients. That may be why there where light and dark areas. Also, there were a large number of routes on this board. Very interesting.
This was really interesting, thank You, Dave! One thing that comes to my mind is, that these resistor networks seem to have rather huge parasitic capacitance and inductance? I'm working on some microstrip filters, so I guess these resistor networks aren't working as well in high frequency applications, more being used as precision ratio in DC- or low-frequency-applications. Is there any way You might get to use some high precision resistor networks in highest frequency applications? Maybe with special geometry to compensate the parasitic effects?
Sorry if my english isn't that well written, greetings from Germany!
@EEVblog they are glas shielded cause the laser cut is done after shielding. The burn dot is caused by the laser. They burned material away for fine trimming the resistor part. Measure while burning away the material toll the right value is reacht.
How does the resistance work with the curved lines if it's strictly based on squares and independent of actual area?
Dave, Did you write the Wiki article on laser trimming? Same picture and from the same source, the Keithley multimeter.
You tended to waffle on a little bit on this video... good stuff, I've always wondered what these were after I saw one in my fluke 23
I didn't believe you about "counting squares" till I worked it out myself. That'll make a nice intro physics problem.
Do they wait to trim the resister once it's on the PCB? There is still resistance on the traces in the board.
Interesting !
Might the glass be so that they can do QC on it or see if it's been damaged for repair?
I like the format of "narrated slideshow". The highres Pictures are really nice to look at, and doing the same with a (mantis) video, wouldn't be better. But maybe add some more pictures of other angles, if you'd do something like this again. But as usual: two thumbs up
Schwuuuuup Thanks. Yes, I kind of like the "narrated slideshow" thing too, it works quite well.
What are the specs on these for temperature drift ?
Greetings!!! What happens with inductance in those types of resistors? I assume that they will have a lot of parasitic inductance... is that correct? Furthermore, do they wave problems at high frequencies, i.e. microwaves (because of the length the signal has to travel once inside that chip)?
may i add (you might have said but i missed it) in our case at the time we purchased these fairly generically (not sure if you can just get them off digikey) and laser trimmed them ourselves (hence the glass back ... that i'm told is actually a crystal but not my department)
Alyx BioHaz I suspected it may be some form of mineral crystal glass, but I'm not sure why they'd need that.
it has somthing to do with the laser used, im not sure what the laser is all i know is its not a visible laser (at least in our case) but apparently glass would have distorted the beam ... my guess is the size of particles in glass but i have no real idea
Alyx BioHaz I think it's because glass absorb non visible light quite a lot...hence over-heating problems. For example in UV-Spectroscopy you have to use quartz and not glass
Yeah i believe it uses a UV spectrum laser
May i also add that those little nubs in the long twisty high resistance resistors are for laser trimming down ... the nub is a short and can be broken by a laser
I suspect the reason for using glass was to determine how much mileage it has on it, unless glass has more desirable insulator properties
Why does HD usually mean tinny low bass low audio level? If you could boost your audio level to match your older videos?
Quick question guys.
On the subject of precision, there are plenty of programs available on the web which will calculate which two preferred range resistors in parallel will give close to a required resistance value. Is there a program available which will calculate the best or most economical combination of more than two preferred value resistors in any combination (parallel, serial, or both) to reach a desired value?
Please add a link to the teardown video.
Nice
they may laser trim across glass after final assembly?
Could theyve chosen glass for adjustment using lasers?
While laser trimming the array, do they have to let it cool down to measure new values? Or they trim and measure simultaneously?
Normally they trim and measure simultaneously, but when trimming a area to slow it can head up. We use a Temp Table that has always the same Temp while trimming to measure precisely and compensate head ups
I think the burn marks are from being laser fine tuned though the glass after assembly of the pack
The detail of thin resistors is much needed in case Kevin Flynn is transported into the digital world, where he has to face off against Dave's computerized likeness,and the imposing Master Control Program, The EEVblog.
Good luck finding a replacement part for that.
if such a part fails, the adjustment of the meter is voided anyway, so you almost need a manufacturer repair anyway.But this part should be designed fairly robust to begin with. One thing that can kill such a network is a high voltage high energy burst. An arc in those ladders can and does burn trough it, increasing the resistance.
Which is the video # of the tear down of this unit? Thnks,
How do these resistors perform at high frequency? I suspect (given the geometry) they're intended for very low frequency / DC only.
Joe Desbonnet very low frequency is usefull too )
Yeah!
Are there issues with this type of package bending or getting sheared off during an impact? It seems like if you dropped the instrument on the right axis, these tall skinny devices would flop right over. Of course lab instruments shouldn't be dropped, but just thinking about mil, aerospace, etc...
dwDragon88 The answer to this is DON'T DROP YOUR DELICATE ELECTRONIC INSTRUMENTS. What are you *thinking*??? I could see this kind of concern for a hand-held DMM, but not a high-precision bench meter...
Ken Peek yeah LOL.
Well, the only other thing I could foresee is thermal stress causing fractures in the ceramic. Like in MLCC capacitors.
A modern/digital version of a Centralabs Couplate?
That circle looks to be extra glue/adhesive/silcone/epoxy that might have dropped onto the glass before installation. Color is very similar.
Glass (or quartz) protective cover allows for trimming through it. It can imply that the resistors are made in one location and trimmed in another. Perhaps a fluke factory for production and keithley one for trimming. What would be the reason is another question.
Secondly, trimming these resistors is ok, but means they already have a more accurate instrument at hand when trimming. What kind of standards would they be using to manufacture 8.5+ digit instruments? Any insight into this would be interesting. Metrology is a fascinating art in itself
These are constructed and trimmed by FMO (Fluke Microelectronics) in Everett. The "glass" allows for multiple passes after sealing to get the resistors into insane tolerances. Also provides the best hermetic seal possible. These networks are rock solid and won't drift more than a few PPM over several years.
Are these things uniquely trimmed by a technician? or they all trimmed with the same pattern by a machine?
Well, with all the laser trimmed and pre baking of the resistor networks, it's nice to see what the major, multi gazillion dollar corporations can use in their labs. At $4,000, USD, I'm curious as to whether you bought one or two of these for your bench, Dave?
hi I was wondering can you explain how a jewel thief works pleas.
At the moment Im drunk. This is very exciting!
Are they affected by light?
10:52 maybe a slip of the finger by the tech in charge of trimming these networks?
I don't understand how the resistance per square works on the NiChrome film. How would a mm^2 have the same resistance as a cm^2?
Jake Leonard The total resistance is resistivity*length/area, where the area in this case is the trace width * the film thickness. A square of any size has a length/width ratio of one, so it has the same resistance.
Very interesting, hadn't actually thought of it by applying the equations. Thanks!
Jake Leonard You could also think of it as 4 resistors in series/parallel, which arrives at the same value as the individual resistors. This would apply if you made one larger square out of 4 smaller squares of the same size. The more area a thin-film resistor has, the more power it will be able to dissipate, so you can only make them so small...
Hi Crashpilot^^
where is the teardown video?
Is the glass there so they can use a laser to fine adjust the resistances?
ipullstuffapart Interesting question. My guess would be they trim without the glass, as heat dissipation might be an issue?
EEVblog
Perhaps, but it seems as though they might do it with the glass on because of the depositing of what you said like a "burnt" material?
My thoughts are they could have the glass there to keep it protected from transit between the manufacturing house and the trim house (if they use different locations) whilst maintaing the ability to trim it
EEVblog
I would think you would want to laser trim without the glass or that vaporized metal would be trapped and go somewhere you wouldn't want it. It's probably similar to LAAPD which is a laser ablative plasma process.
***** If so, it works!
ipullstuffapart The whole idea of having a transparent window is so that you can LASER-trim *after* the resistor is assembled [not in the equipment necessarily]. The reason for this is that the absolute values of and the strains on the resistors will change after the high-heat bake cycle to attach the "lid" to the ceramic. They use a paste made of an organic binder mixed with borosilicate glass, called a "frit". This melts at 390C, so you can imagine that the resistors will change in value after this. So--- the need to trim *after* the network is assembled. This is a 100% hermetic package, so they can use high-stability NiCr alloys, and there is no worry of electromigration problems.
sure you waffle on a bit, but that's what makes it so interesting :)
"Try Shift for straight line"
Someone send Dave a tablet and stylus so we can have DaveCAD run on the computer :D
Hello to anyone that's reading this.. (Or Dave EEVblog ) I find these electrical circuits and all very interesting, and I've always wanted to get into Electrical Engineering since I was a little kid, my uncle had these learning books and things, but I never got to play with it. I currently am thinking IT would be the perfect thing for me, but I'd also like to get into Electronic Engineering. (I currently am taking an Electrician course sort of thing, and do computer stuff as a hobby.)
What would you recommend for me as a starter to get into this hobby? (I know about arduinos and all of that, but I'd like to start experimenting with breadboards and all of those simple circuits). Any websites you'd recommend me going to and read up on?
Thanks!
You can use Arduino with a breadboard, it's essentially just a MCU programmer device. I use Arduino, but I always do projects without the Arduino board itself.
Here are some excellent circuits I started with, you can find 555 timer, shift register and logic gate tutorials that you can do:
th-cam.com/users/DerekMolloyDCU
But I recommend you get to MCUs(Arduino, Launchpad) early on. You can get so much done with those, and they're so cheap nowadays that there's not much reason to avoid them.
MCUs are way easier and more handy. In our university we started off with Arduinos first, without touching any general purpose ICs.
RealationGames thanks! Will look into it. ^^
Mike Fox Look up a book called "Getting Started in Electronics" by Forrest M. Mims...
The programming in C or Assembly as well as Digital electronics is worth nothing if you don't crack your head with the analog part.
And as with everything... Learning by doing... And erring a lot... Just remember that you learned something new whenever Dave's so called "Magic Smoke" rises... :)
Also... Tipp for beginners: Think of electronics as if it was water... Helped me a lot to understand the basics (google "electronics water analogy"
Are you the same guy that talks on the Mindcraft games ?
Now take a picture of that with flash while its running and see what happens! :)
them Thin Film Resistor, don't cost alot to buy. smallest resistors about $0.02 ....
The new type of Tutorials is really nice but you should get yourself a Wacom tablet... ;)
👍👍
I'm sorry, but you haven't caught me yet.
Explains in english why these ridiculously precision resistors cost so much. Having to tailor/trim each and every one is a big step in production, not to mention a rather niche item when it comes to mass production crap. Only the machines actually testing/building said mass production crap will contain such precision parts.
I thought this one was reall neat.
Why do you need precision resistors like these?
A.J Norris Go look at the specs for the DMM7510 7.5 digit meter!
So they are just used to obtain a higher presision when measuring then?
A.J Norris
Car ignition systems.