EEVblog
HTML-код
- Опубликовано: 2 авг 2024
- Peter Daly from Keysight takes us into the Metrology Standards Lab at Keysight in Melbourne Australia.
Looking at the RF standards and test rack and detailed looks at metrology grade RF connectors.
Standards lab vs calibration lab, and traceability.
Forum: www.eevblog.com/forum/blog/eev...
Support the EEVblog through Patreon!
/ eevblog
EEVblog Main Web Site: www.eevblog.com
The 2nd EEVblog Channel: / eevblog2
Donate With Bitcoin & Other Crypto Currencies!
www.eevblog.com/crypto-currency/
EEVblog Amazon Store (Dave gets a cut):
astore.amazon.com/eevblogstore-20
T-Shirts: teespring.com/stores/eevblog
💗 Likecoin - Coins for Likes: likecoin.pro/@eevblog/dil9/hcq3 
Наука
![Jordan Adetunji - KEHLANI REMIX (feat. Kehlani) [Official Video]](http://i.ytimg.com/vi/EeJ8n5PxFGE/mqdefault.jpg)
Please do more videos like this. It give so many clues and hints in every second sentence when listening to people with experience.
I love the old HP rack. We had the same at our university. Everybody had to learn how to properly connect RF connectors without damaging them. The RF power sensors remind me of my days as an intern at Nokia. I had a stack of gear worth 500k at my intern place. I‘m thankful forever for what I‘ve learned there. Too bad it is all gone now.
My university still has HP gear with nixie tubes.
I'd bet a LOT of universities still have labs full of old HP test equipment.
This is the good stuff. Mainstream TV shows never get into the nitty gritty like this.
That Agilent screensaver though... :D
Check out that name-tag while your at it XD
And the Hewlett Packard Systems rack. Good stuff.
Why did you change your name!!
I suspect the Melbourne lab will be getting a call from the Marketing Department on Monday.
I think it was shot before they changed their name - based on some of the cal date stickers on the equipment
I found this absolutely fascinating! The biggest takeaway was when connecting the various device bodies and turning them instead of just the ferrule and having the center pin conductors get damaged. Wow.. I worked for years with 18GHz MW links and never thought about this. Thanks Dave.
"Now Dave, you're a low frequency guy..."
This is great how he emphasizes connector care for high frequency stuff.
Interesting to see that Peter has the same shirt and tie as he used almost 5 years ago in EEVblog #424. And Peter seems not to be a day older now. Is this really a new video?
Well spotted..... looks like a younger version of Dave as well
Y'know, for a second there, at the beginning, i started wondering whether Dave started dyeing his hair, but i figured surely he can't be THAT vain...
21 degrees and a bonded ESD floor will do that to you. Plus all the 50GHz sources to regenerate your telomeres.
I think dave recently tweeted that he found some video files that he thought he lost, explains why he waited to long to upload them xD
If it ain’t broke…
Loving these interviews and site visits - hope to see many more to come!
Excellent. Thanks for the Keysight insight, outta sight, goodnight. I'm going to take much more care of my connectors.
The reference kit case at 26:00 has the place for a cross! When you need to expel the last demons of inaccuracy :D
Very interesting, love these lab tours.
Hi Dave. I really enjoy this kind of videos i hope we will see more of these. Perfect video. Thanks.
Would love to see a video about ESD, discharge from the Parka and rubber shoes, how to avoid, and save designs from being zapped
Thank you Dave and thank you to Keysight for giving Dave and his fans peak!
Thanks for the great video Dave. I always learn something.
fantastic job dave! loving this tech stuff!
Awwww it ended! I was enjoying that, now for the next instalment! I especially liked the part about the high end connectors.
very nice video dave! love the metrology stuff
Always amazed about the “little” stuff! Lesson connector, episode One!
Thank you very much, Dave, for videos like this from Russia. I'm a metrology engineer and it's really helpful to see the other countrie's metrology system working and lab's equipment. Thank you again
Динар Сулейманов hello soviet spy
It's really good that Agilent is willing to devote the time to talks like this.
Dave must be digging into the video archives. The video showing the portable lab is from Feb. 2014. Peter is wearing the same shirt and tie. Still love this stuff. Racks and racks of high tech precision equipment.
JPennDotCom yeah I was wondering if they somehow didn't spin off with keysight or this video was done in 14
Expensive kit still comes with a floppy drive 25:03
No USB on that sucker :)
My uni still uses a spectrum analyzer with a floppy drive. Good old HP kit😁
i had to google what metrology is; cool beans! make more mini documentaries like this
THANK YOU !
Peter said N-connector is the Navy connector which is a common misconception, N is for its inventor Neill.
Paul Neill worked at Bell labs and invented this connector, later on he worked with a gentleman called Concelman, together they reused the inner collar of the N connector and made the BNC and TNC, Bayonet Neill Concelman and Threaded Neill Concelman.
“People using these crappy [$20] connectors”
When 20 bucks is a crappy connector you know you’re not in Kansas any more.
Video of the DC/Low Frequency area please!!! Make it an hour or 2 long, while you're at it :P
Such a formal environment. Same type lab I did research for in my country, people walk in hard rock shirts and wear sneakers, including some of the managers.
Nice Lab and interesting video.
A lot of his instruments connectors didn't have covers on, to protect them. Could do with taking a leaf out of Shahriar's book at the Signal Path.
Michael Hawthorne
You have to be careful with covers, too. They may cause damage, leave behind crud, etc.
Ideally, you carefully clean high precision gear before each use (or at the start of the day).
The E4448A is just enough for audiophools to measure their headphones though ^^
Love these videos Dave - I bet he put that tie on just for the video.
@Dave, how often do you calibrate? Which devices?
>Brings us to a room full of very expensive and interesting machines.
>Talks about connectors for 18 minutes.
XD
lol. Great stuff though.
Not as bad as all that HP equipment
The connectors are easily over 1k USD each :)
It looked good. I expected he says somethink about the calibration procedures, about cooperation with other metrology labs... No, he killed the video by dumb connector that is more precise and bore expensive than standard... Thumb down for this video.
+PETMK I disagree. I found the video very interesting and educational.
Tyler Barnes
Connections are the most important part of the system when you get up into RF. The lab I worked for only went up to 1.1 GHz ‘scope calibration, but even there you have to be careful.
Try comparing some cheap Chinese BNC connectors to expensive variants. Chalk and cheese.
Even some of the dubious 4 mm banana plugs that you see in Dave’s videos can be a source of problems - high resistance and associated heating, physical damage to the instrument connector or piggy-backed plugs due to rough surface finish.
If you have some cheap banana connectors then you should wear them in before use - spray some contact cleaner on then plug them into each other several times until there is no ‘crunchy’ feel, then clean off the contact cleaner using IPA (as there will be a fair amount of fine metal present from the wearing in). Or buy better cabling; but cost is often a bigger problem than your extra time.
There are a few things said about RF connectors around 17:30 that are about as wrong as it's possible to be wrong: "Its' not 7mm wide, the area is 7mm wide" No and no. A 7mm aka APC-7/GPC-7 connector *is* 7mm wide between the inner walls of the outer conductor. An area can't possibly have units of length. The dimensions of these connectors can be found in IEEE Std 287-2007, IEEE Standard for Precision Coaxial Connectors (DC to 110 GHz). Annex D gives dimensions for 7mm. Likewise APC-3.5 is 3.500mm (Annex F). The standard is paywalled so I can't distribute, but some good info can be found here www.microwaves101.com/encyclopedias/790-microwave-connectors. Rant over, sorry.
I think he meant to convey that (it's the width of the waveguide not the width of the centre pin) but it wasn't explained very clearly.
You would expect someone in this position to get this right. It's the width of the interior of the outer conductor of the transmission line. It's not a waveguide it uses TEM modes.
Looks like a box for an iDirect line card on top of that cabinet right at the end. I do a lot of work on those :D
What was that last word Peter says in the video?
An thanks for showing all of this Dave an Peter! Could be a boring subject with the wrong person, but Peter is a very nice person to listen too and he bring the subject very well!
"Cool bananas". That might just be an Australia/New Zealand expression. It just means "good stuff".
At 26:00 he starts handling the precision check set without gloves. Probably a lab supervisor off camera waiting to sweep in with the IPA swabs and undo the managerial mishandling. :D
Neat stuff.
Ha... I lost it at his super serious face when saying "..can damage the front-end of your unit". That's it I'm off to bed.
For those that want more details on connectors, TSP #26 - Tutorial on Microwave and mm-Wave Components and Modules at ruclips.net/video/6kwamCh1QkE/видео.html gives a lot of good information on the various rf and microwave connectors.
Respectfully, the N connector is named after Paul Neill of Bell Labs. It does NOT stand for “Navy”.
See en.m.wikipedia.org/wiki/N_connector
Great vid, thank you!
Nice bit of kit there. How did you manage to keep from drooling? Thanks!
Wear a bib when you enter the lab
That signal generator is an Agilent 33250A!!
They have a new model that has been around for some time now and uses "Trueform" technology so I thought they would completely replace their 33250s immediately! (considering it's in a calibration setting and they'd want to put their best instruments to use)
I'm no expert in metrology but the costs of 'upgrading' devices in a cal lab probably prohibits frequent changes. If they change to different devices they have to change all their standard operating procedures and documentation, retrain staff, etc. Not to mention they would probably have to run the new equipment and old equipment simultaneously until the next calibration date when they can be confident that the old device hadn't failed and they can finally decommission it. Also the longer you use a device without it drifting in calibration, the higher the confidence that it will not drift in the future.
It's not that easy to just replace equipment in a cal lab, the existing instruments have a lot of traceable history.
Not sure if it's a factor in this specific case, but in a reference/calibration application an older device with nominally worse specifications may be preferred to a more recently produced one simply because the older instrument has already aged to a lower aging rate.
Justin Spencer - your eagle eyes may not have noticed that not only is there only HP and Agilent gear here, but that Keysight didn't exist when the video was made. LOL
I work in a RF Metrology lab ( dc-50Ghz). It is not uncommon to spend $$$$$thousands on connectors and cables. Due to the constant cleaning torquing on and taking off they all wear out.We have our cables on a PM and sweep them on a regular basis. However, on the video, it was boring.
been there. Got the T-Shirt
Why are BNC connectors used on equipment rated above 500MHz if the signal is compromised?
Cool video, but why do you upload it like 5 years later?
I need a set of these calibrated connectors and wires for my hifi setup
Now come tour keysight in Santa Rosa please.
Holy old-video batman! How many years ago was this filmed?!
Seriously, someone would have done those connectors up with stilsons? I think my advantage of learning electronics seriously at the age of around 30 odd was that I already learned the basics of other trades such as mechanics, Welding, building etc. To me it would be obvious not to use force on an obviously delicate piece of equipment even by just looking at the box it was stored in and probably everyone reading this has much more skill in electronics than I do. If you have people doing things like that in your lab then someone needs to give HR a kick up the butt!
These machines look like they belong in a museum or a 1990 disaster film.
go dave!
I think he wore that shirt just to mess with your auto-focus (Maybe the tie also).
So it's a bit 'how ya goin' with the focus at times, still loved it though.
zzzzzzzzzzzzz Good thing there are people out there that actually like this kind of work.
I am all for the highly regulated world.
Nice vid again.
just one question from a newbie,if the damn connectors can break so easly why dont make them easy to replace in such gear? maybe the losses from a extra cable on the input or a edge card or something detacheable on the frontend board would be too much loss for the aplication?...
Its not that they 'break' its that they wear down if not used correctly, spinning them as you tighten shaves a microscopic layer off through friction and at the frequencies they are testing that can have a significant effect on your result over time. They are also they are very expensive to make - there's probably only a few tens of places in the world that need them so its not like you can get mass manufacturing advantages. Breaking them out on a detachable card might work but apart from the losses you mention it means there's an extra piece in the chain that would need to be calibrated/verified.
regmigrant I see thanks for the explanation
muy interesante capo.
ESD is rigthly important, but cardboard boxes on the bench????....... I can tell you that in the defence sector that's a big no no, deal with a company like Rolls Royce and they'd tear you a new one for that.
Also, are the chairs ESD bonded? Typically man made materials, sitting off the floor on non-conductive wheels.
Laser printers a big no no too, as they use static to transfer dry ink and the paper always comes out with a slight static still on it.
I worked for a place that used to lie to customers about their ESD measures, we had a "fully bonded ESD floor" (or should I say standard vinyl flooring with a big helping of BS?), it's one of those things they usually get away with because companies or potential customers always send "idiots suits" rather than engineers. As an engineer i'm also a BS detector......
In this video they have clearly run around sticking ESD signs on things before hand, which in itself is not necessarily bad as long as they back it up.
Also, from looking, it is reasonable to assume that the clothes that both are wearing are not ESD compliant - I guess that's why they never did a foot test when entering the room.
Well that's another issue, I once worked for a major Japanese electronics corporation here in the UK, the production facility had a dress code imposed by British management of "ESD coat, ESD hat, collared shirt and black trousers, NO JEANS) most of the non-jean trousers you can buy will be man made materials, as a result production staff used to walk around discharging themselves onto custom made production/test equipment worth £1'000'000's, jeans would actually have been a better choice, or at least supply clothing made from natural fibres.......
Connectors? You go there and and only film connectors? Will there be a part two?
The connector technology owes a lot to Julias Botka at HP. www.mwrf.com/test-amp-measurement-analyzers/microwave-legends-2012
Why does he refer to the N connector as "Navy connector"?
Dat tie with dat shirt doe - tres gauche. Just add pocket protector for maximum nerdiness.
At lest Keysight is using HP equipment to verify there equip
OK so I just watched the other video with Peter Daly from 2013: ruclips.net/video/G3QK31zotoQ/видео.html. He's wearing the exact same shirt and tie and the same pens. Was this video shot in 2013 too???
He does say "We'll have a look into that in another video" (not we had a look or similar) so it does look like they were recorded together.
Both are fantastic videos.
I think the video is recent, it's just that you try to keep everything as constant as possible when you work in a standards lab. Temperature, humidity, shirt, tie, pens, company name, etc.
rofl XD.. imagine having to recalibrate everything cause of a new tie
It can get expensive having to calibrate new fashions every day, so like Einstein, a few copies of the same garb is all that's needed. Also, I believe access to military barber with a #2 clipper and ample supplies of Oil of Ulay for skin condition are standard requirements of any metrologist.
Floppy disk? do they even produce those anymore.
That HP rack equipment though...
He wasn't using a torque wrench!?
Floppy disks everywhere. O_O
Agilent ftw :D
Why aren’t the center of the connector square, so the can’t turn.
Ill assume its because a cylinder is easily manufactured to defined radius, a square needs all sides and angles to be the same, much harder to mahcine
ExtremeBlastoise sure but for this precision work? And a USB connector is square to., there is no friction on that connector.
I was thinking along the same lines but maybe because you then need precision in three dimensions instead of two so it might be cost?
regmigrant I think your are write
I think you are confusing coaxial connectors with waveguides! Trying to do metrology in the face of multiple
frequency-dependent modes would be, erm, tricky.
21 +/- 1.5 degrees C doesn’t seem that tight. High precision metrology labs are often +/- 1 or even 0.5 degrees C.
And 20, not 23
wait, his card still says Agilent Technologies!
HP printer, of course :)
Nope; that's a Zebra label printer next to the PC display
Where is Dougy :(
Awesome video.
13:17 "Your computer might be at risk" "Antivirus software might not be installed". 😬
It is a shame that these devices have their user-interfaces done with software like this. But they are "only" measurement equipment... It is more scary if you look at medical devices in hospitals, think about heart-lung machines running this kind of software. ;)
Lol yeah and its not just scary anymore it actually happened in the UK, half of our hospitals were effectively shut down because the IT departments weren't updating the PC's regularly if at all :/
NIST is National Institute of Standards and Technology, not science
Here I though that *I* looked like an idiot when I thought that's what it stood for when interviewing for a job at Guildline. Also, if you want to see some crazy metrology standards, Guildline is tops, all the stuff in this cal lab is ultimately calibrated by Guildline stuff.
well, i had been in that business over a decade here in Singapore so i can pretty much spell most of the national lab name back to front and details of inernational standard DIN,BS,JIS,ANSI bottom to top.
The new logo appears over the old logo during the opening intro?
13:01 """"""""""low frequency""""""""""
25C would kill me, 21C I would be sweating, the joy of living in Scotland.
Have to agree.......25degC too hot. Ian (also Scotland)
I've noticed the temp stuff gets calibrated at has dropped over the decades from 25 Celsius to 23 and now it seems 21.
I agree 21 is probably just right for a *shiny arsed bastard* unless female, in which case 25 is way too low. Please close all the windows and turn off the aircon.
34 idiots thought electronic wizard talk is boring,watch it 1,000 times and you'll be impressed.
Found a Floppy disk drives still in existence 2018!!! 24:37😐 What is that doing in that 3-50 ghz scope?
_"I'm not going sexist about this, that's what it's called"_
And that sums up 2017 everybody...
Although the rest of it is hilarious 11:57 you kinda understand why he had to keep repeating male and female _(connectors)_
What a time we live in. When an engineer working on this level feels he needs to stop and explain he's not talking about sex when he identifies, "Male and female" connectors. Somewhere, a snowflake was about to melt!!! Great video Dave. Thanks.
It is real bad in Australia.
@@ChickenPermissionOG I get much of my news from Australian TV networks. Here in California, the only thing I trust during the news is when they say goodnight! Everything else is propaganda. Some day we hope to visit Australia. Looks amazing. You could spend a lifetime exploring the American southwest and we have seen much of it. Australia and Ireland are next. Take care C. P. Tim.
"that much paper" - doesn't sound very precise ;)
Cd... this guy living in 1990..
Did you notice that the E4448A 3 Hz to 50 GHz Spectrum Analyzer still has a floppy drive?
$30,000 equipment has poorly configured windows installed... hehe
You're no good at metrology unless you've got that ultra-OCD geekiness in you. Right down to 8 decimal places. Or is it 8 and a half ?
surely 8 +/- 0.5%?
I loved the bit about connectors. But kinda wanted to throw lube in his face... I'm sure there's a reason they're not using any. Something something RF excitement
Like copper grease? I mean copper is a great conductor after all. lol.
NIST __ NATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY ----- CANT BELIEVE HE SAID SCIENCE AND TECHNOLOGY
How can a company that produces such extremely advanced measurement equipment.. produce such utterly garbage printers? This isn't a technical question, it's a marketing question. Why associate HP both with extremely high value measurement gear, as well as consumer shit? The mind boggles
Because, they' re not the same company anymore. Now for almost 20 years. They are not connected in any way. Some might say, the heart of HP became Agilent. The EMG of Agilent was once the core of HP. Computers and printers came waaayy later. Today this heart belongs to keysight. Although Agilent is still going strong with chemical analysis stuff.
is this a 4 year old vid?
....... Who said you could come to Melbourne ?
The traffic is NUTS already without bloody 'New South Welsh-men' adding to it !
Get back to the Giant 'Coat-hanger'.
ooh early comment
Cal labs stuff bores me once Dave and the guest confirmed (in their opinions) in the past that "Calibration of a device does not involve any adjustment or correction of that device". Given that, I would only use their service when some law or regulation requires me to, not because I think there is value in using their $$$ service.
I feel what they are selling is validation not calibration, as the unit being "calibrated" will perform the same when it leaves their lab as it did when arrived at their lab, only paperwork is produced.
You use the device as before, and its performance doesn't change, but in your processing software you update all the offsets. Many devices also let you update the display offsets after calibration, so that the displayed value will be accurate even though the measured value has the same error. Calibration tells you how the device is drifting with time, so you can update your error analysis.
Even devices that can't be adjusted need calibration. They mentioned rulers in the video: steel changes size with temperature and wear. If you have a precision scale (fancy machinists ruler) you can't adjust the values it reads but you can adjust the correction table you use with it to determine the actual length measured.
You are just not understanding what *calibration* is compared to *adjustment.* I'm pretty sure this lab is able to do an adjustment on some instruments if you ask them, but that would totally screw up all your carefully maintained history and aging characteristics of your device. You are paying for a traceable history with all the devices used to calibrate being traceable too. Think of a space shuttle disaster or something like that - how can we blame some other party? Traceable cal.
I agree the terminology is confusing, maybe call the desired process *recalibration*
But calibration as understood by mere plebs is not calibration as understood by scientists. Just like the *theory* of evolution is nothing but a fanciful notion to the plebs. lol.
One question. Why do test gears like these run Windows to begin with? Linux would provide low cost fully customizable solution. Is it for running Windows based analysis software like Labview right on these gears? Coz um pretty sure no one is gonna run MS Word in a scope.
Hasanur Rashid Ubuntu!
Sometimes it's because they use off the shelf parts for which drivers where only/earlier/more easily available for Windows. Things like graphic chipsets for instance. A Windows license cost is negligible for this class of kit. Add the savings from reusing as much as possible from earlier proven designs, and you easily get vendor lock-in even a decade (or two) after alternatives have become available.
Very knowledgeable guy, but he shouldn't assume the connectors' pronouns.