Datron Wavetek 4950 Transfer Standard Finishing Touches & Tests
Vložit
- čas přidán 25. 06. 2024
- Try PCBWays fast & affordable prototyping services pcbway.com/g/Tj5om7
Raspberry Pis back in stock @ www.welectron.com/
Raspberry Pi GPIB shield elektronomikon.org/
Precision torque meter by mark-10.com/
My data about 4950 github.com/marcoreps/datron_w...
4950 vs 5700A cal script xdevs.com/doc/Wavetek/4950/ca...
Noise spectral density plot tool github.com/macaba/NSD
00:00 Intro
00:53 Self test
01:37 ACV pre-preamp
03:59 Current assembly
04:48 Input connector
09:54 Worn out relay
11:17 Cal setup
13:30 Calibration test
14:59 ACV linearity
21:10 Special menu
22:14 DCV linearity
23:07 DCV noise - Věda a technologie
The ratio of how much I enjoy this vs how much I understand is off the charts
You definitely need to try black magic. I mean microwave stuff, that would be extremely interesting.
Wait, what he does isn't black magic already???
@@MCsCreations I'm not trying to start the holy war here... But things like dielectric heating or resonant cavities with shorted input are kinda impressive. And we are not even started to talk about sci-fi level RF stuff like long delay echo phenomena or nature-generated signals. So, I'm sure, he will buy something like spectrum analyzer eventually and we all will see the whole new world of weird things.
Wonder what happened to that spectrum analyzer he bought
@@gigigigiontis8well, I can see an old Network analyzer and an S-parameter test set, so I think he's onto it. From my experience, Spectrum analyzers start stacking up quite quickly, but getting a VNA running correctly and in spec is quite hard. I'm actually quite close to pay a couple of thousands to get a working one, so I can finally take care of my stack of faulty/dubious RF Equipment. But even the cables are quite expensive, and even looking at them or changing your tongue angle will affect the measurement. So yeah, definitely black magic, because even existing in its general vicinity or a wrongly torqued bolt has chaotic effects on the circuit.
I would love to see you do some RF stuff, it contains even more magic than nanovolts. Plus you often play with nanovolts, disguised as dB.
Considering the equipment present on the right at 11:30 ... maybe something's up? 😀
Noticed the roll of paper towels, handy for mopping up escaping PPM's......but luckily Mr. Reps didn't need many, he caged them back up safely.
Does anyone else watch these videos and have zero clue what he's talking about, but is still fascinated...
Yep! KEKW!
1000%. God I wish I understood half of it, lmao. I'm more familiar with mechanical things than electrical engineering and metrology. But hopefully through osmosis, I'll learn something useful here, lmao.
I'd be so thrilled to be an apprentice, though, even though this isn't in my normal wheelhouse. This stuff is just amazingly fascinating to me. I love this level of attention to details, and it's unfortunate that it's so hard to find people that also do.
@@WhiteVaille I've been here since day one, learnt so much! I just don't get how he has the time to get so knowledgeable about such niche subjects... 🤷
this guys sense of humor makes every vid a 10
*Overview*
- Marco Reps discusses the Datron Wavetek 4950, a multi-function transfer standard, detailing its issues, repairs, and tests.
*Repairs and Issues*
- Fixed a missing positive voltage supply rail on the analog board.
- Replaced a damaged op-amp in the AC millivolt pre-pre-amp.
- Addressed leakage current issues in the current assembly.
- Replaced the original front input connector with a more universal one.
- Replaced a worn-out relay affecting the AC attenuator and preamp board.
*Modifications*
- Created a new sheet metal holder for the front input connector.
- Used PTFE heat shrink tubing to add dielectric strength.
*Calibration and Testing*
- Used a Python script to calibrate the Datron Wavetek 4950 against a Fluke 5700 calibrator.
- Found some discrepancies in ACV linearity.
- Used thermal voltage converters to validate ACV linearity.
*Additional Features and Observations*
- Found that the Datron Wavetek 4950 has excellent DCV linearity and low noise.
- The device uses hermetic bulk metal foil resistors for stability.
*Conclusion*
- Marco is impressed with the Datron Wavetek 4950's performance and is glad it was saved from being discarded.
*Positive Points*
1. *Excellent DCV Linearity:* The Datron Wavetek 4950 shows remarkable DCV linearity, exceeding expectations for a seven-and-a-half-digit class device.
2. *Low Noise:* The device has low noise performance, even down to the least significant digit.
3. *Stability:* Uses hermetic bulk metal foil resistors for stability, making it reliable for its primary function as a multi-function transfer standard.
4. *Versatility:* The device can be modified to act like a normal multimeter, adding to its utility.
5. *Durability:* Built with quality components that guarantee long-term stability.
6. *No Cooling Fan:* Lacks a cooling fan, which is seen as a positive due to the longer time constant and low temperature coefficient.
*Negative Points*
1. *ACV Linearity Issues:* Some discrepancies were found in the AC Voltage linearity during testing.
2. *Complex Repairs:* Multiple issues required attention, including replacing worn-out relays and fixing leakage currents.
3. *Limited Documentation:* Lack of comprehensive service manual makes some advanced adjustments risky.
4. *Old Connectors:* The device uses older types of connectors, requiring adapters or replacements for modern setups.
5. *Settling Time:* Thermal voltage converters used for testing have a long settling time, making the calibration process lengthy.
These points are sorted by their impact on the device's performance and usability.
Thanks a lot for this "wetting current" hint. This might be the problem why those super expensive hermetically sealed relays on the voltage sense card (ADC) in the semiconductor test stations I'm going to work on are failing... "even there is literally no current flowing through them"... :)
I have no clue on what all this stuff is for, but i love your dry German humor and your technical explanation and your curiosity and that you like me take everything apart to see what makes it tick...
My big question is what do you actually use all this equipment for?
I know what ppm and a lot of the technical terms since i am en electritian and a ventilation technician, but still a lot of this stuff is way over my head 😀
When they build the multimeter you use in your day-to-day work, there are tolerances with all the components, and it has to be calibrated against known-good exact voltage sources. Those sources, in turn, have to be measured and calibrated to make sure they're not changing over time, which requires precision meters and sources, which in turn require calibration against something more precise, and on up the chain until you arrive at the offices of NIST / DIN / ISO and multimillion dollar standards based on nothing more than basic physics. These boxes are several steps up that calibration chain...the kind of stuff the manufacturers probably use keep their equipment calibrated.
That's what the original customers used it for anyway. What does Reps use it for? Everyone needs a hobby I guess. Before finding this channel, I'd probably have said that tracking tiny voltage fluctuations in precision calibration equipment would be as exciting as watching paint dry. Turns out it you dig into it, both ppm voltage stability and paint drying have a lot of really fascinating subtleties that normal people would just gloss over.
Check out Raymark labelling material for front panels. It’s a kind of polyester like backing with an uncured epoxy coating. It comes in label or A4 uncut sheets. You print on it with an inkjet and then bake it in an oven. The baked sheets are resistant to oils, solvents, fuels and quite wear resistant!
I still can’t believe your sense of humor in not your mother tongue. Love it!
Pretty impressive equipment indeed, Marco! 😃
Fantastic work fixing it!
Stay safe there with your family! 🖖😊
I love this content. I have a feeling that given an ossified claw from a T-Rex you could make it walk and talk and feed on small children like back in the good old days. Thank you for this work, and for sharing it.
I have to say I am jealous on you, to have the ability to dive so deep and sort things out. I have been fighting for more than a year to fix my oscillator in my Fluke 5100B without success. Thanks for a great video.
Happy new year 🎆
A marco reps upload. Exciting
“Lay the Re” 😂 Good one.
Excellent repair and modifications. Always wonderful to see a fine instrument given another lease on life. Also, that's a lovely $1700 torque gauge you have there!
No resentment from me, but I see way too many awesome looking tools made from unaffordium on CZcams.
Very nice Marco
I remember you as a child - you designed JBC clone soldering iron - that was cute
PVDF heatshrink + PVDF cables (SPEC-44/MIL-W-81044) make a very good pairing. The heat shrink fuses with the cable jacket and forms a perfect bond when using the right temperature, giving you up to 2,5kV voltage rating, as well as all the excellent properties associated with PVDF. Just unfortunate that the HV type can get very pricy per meter, but the 600V rated one is decently priced.
interesting thank you! that goes right into the notes under "good stuff"
PVDF is piezoelectric, I understand?
didn't know PCBway did sheet metal fab, going to try this out!
You are More than just GOOD!👍👍👍
ThX for a great vid with so much acquired knowledge and effort.
21:37 - Sir, you are the real treasure in my mind. 👍
At 16:47 the towel of the Physikalisch-Technische Bundesanstalt (PTB) Braunschweig is very pleasant 😂😊
*Badeanstalt
I might have something for you that has (theoretically) "surprisingly good linearity" as well...once I fix the 1mVpp noise in the conversions💀
I have no idea about 90% things you talk about. But I love this channel anyways! :)
Another awesome video, thank you. 👍😉
Pentaherz ;) Thanks for great videos as usual ;)
Yes, finally. Fame at last. Fluke Norwich is my workplace. We have many treasures indeed, including most of the guys who were there in the Datron days. More than that, I'm not permitted to say.....
20:14 is like watching marco do some secret Voltnut ritual
شكرا على الفيديو
Congrats! You could make a video on precision scales.
Damn, You insane! Me like it!
With this guy I don't even need to be reminded to like and subscribe. I just wait for him to crack a joke and I am bought. Lay the ree... wow... I know my like will be safe with you, sir.
16:50 The gender-fluid 874 connector you speak of was created by General Radio, not General Electric. A tiny typo. Another fantastic video. Thank you!
Do GHz next Marco! 😂
Norwich is pronounced locally as Norrij with hard J sound x
Thas roight my man.
Very neat.
How about using cadmium copper instead of pure copper? It is stronger and more wear resistant, with pretty good… ooh, right, RoHS… OK, not even mentioning beryllium copper then.
Wouldn't mind seeing Atto-amp measurements. I've gotten noise down to ~30fA at 30Hz with a LMC662, a discrete high voltage output stage, and a 10Gohm resistor array. (about $100 in parts) Stray capacitance to the shield from the 10Gohm resistor caused fun times. Had to get cheeky with driven shields to solve that problem.
The circuit made a very useful amplifier for photo-diodes. Would full-scale with micro-watt laser beams so you could do great measurements with poor laser setups.
GHz! Would love to see that SA you have hiding in the todo list :)
cool stuff! I don't know much about 'electric metrology' but I know physical word accuracy, its cool to see and you make it seem approachable.
Random ? but did you once have a video on some sort of distance sensor, which also had visible light camera as well? I distinctly remember a color map overlay of the video input, with distances called out (I think it was red close - blue far?). I'm almost positive that is the video that I found your channel with, but I can't seem to find it... Maybe I'm confusing it with all your thermal awesomeness... Thanks for sharing!
this one maybe? czcams.com/video/ZMxzRn9GrKY/video.html
Nice equipment! If I may ask.. do you have a particular need for all these ppms and digits? Or do you mostly use your equipment for measuring your other equipment?
Just wondering!
Gear Acquisition Syndrome is a service to itself.
Could you go into more detail on the construction of your Fluke-inspired test-leads?
Those were just copper coax cables with inner conductors brought out via short silicone cables to copper lugs and both shields of a pair connected together to stackable 4mm banana plugs for guard connection.
But actually a few days later now I made another short test lead from a pair of triaxial cables. Same as before but inner shields can be used for hi sense & lo sense, so that the core is always fully surrounded by a similar voltage and therefore there is lower stray capacitance and lower degradation of AC signals.
Next? Maybe you'd like to use the Swerlein algorithm on your 3458A, for measuring AC @ < 10ppm uncertainty up to 1kHz.
Have you considered dumping all data from the eeproms and such so that you could revert it to the current calibration and then doing the AC-lin calibration? Having dumped memory means you can revert it to pre-cal state if it fails
Those old GE connectors are quite hard to find! I replaced them with N-Type ones on my pulse generator, because I couldn't find suitable mates or cables I'd actually trust for sub-nanosecond rise times.
I spy an expensive 3.5 mm (or 2.92?) VNA test cable in the upper right corner around 12:00 ... Exciting to see what the future has in store...
I had no idea Datron used to be an electronic sensors company lol. They have an interesting story.
You have more LPM's (Laughs per Minute) than Seinfeld 😂😊
please test the new ts1c cordless soldering iron from miniware!
after going through the pain of adding PTFE shrink-tubing to the input connectors, wouldn't the metal plate around them also cause a shorter air gap with those fork terminals ?
What's the end goal of all these ppm/ppbs ? Any practical applications ?
I would have slotted out the damaged PCB area completely and used point-to-point wiring. Old Fluke stuff has it all over the place in the most sensitive circuitry.
Did PTB give you that towel?
I keep my wife faaarrr away from my eBay account for that exact reason 😅. Well, that and my server acquisition syndrome as well. 11:38
The Al business card came out really well! Surely it would be no worse at 1.42kV than the mounting bracket anyway? Unless the holes were smaller in the cover plate?
yes
The story continues…
I wanna see ghz stuff. specifically, somewhere from 70 to 90 ghz. plus some liquid nitrogen. that would definitely help elevate your voltnut obsession to the next level.
About the glitterbox... I do have to motivate someone to let me adopt something....
I would like to see some nH.
If you Still need an opa637. I have some in SOIC and I think at least one in the OPA637SM (metal can) happy to send your way for free if you like. Most of my work these days is multichannel audio and they are not of any use as a low noise error amp in a low noise regulator really, due to poor stability at low gains.
10:09 Cold solder joint in the transistor leg in the middle of screen ?🤔
Wait, can an input relay go totally bad as in always open from not being used or is it just a bit of extra resistance? IE can relay contacts go bad enough just from sitting to cause zero readings on a DMM but still pass internal calibration?
Haven't encountered that myself, but I'd say it isn't inconceivable ... maybe a different effect than "lack of wetting current" tho
mannnnn make something easy and fun in ur style !!!!!
How to enable so it will measure full range and not only around the cardinal points?
What would be a substitute for copper metal if I don't want the oxidized connection points?
For single digit nanovolts: Probably nothing, pure clean copper and thermal equalization mandatory! For most day-to-day purposes gold or silver plating on copper, maybe even with a third metal diffusion barrier, is not too bad.
I say Josephson Junctions. Per Wikipedia those are also how the Volt is defined!
A JJ would be cool to see in the Reps Lab, but liquid helium is not household friendly. The rest of it would be garden variety for him, he could probably build a microwave source out of an old oven and a piece of metal tubing.
AR488... GPIB to JSON adaptor
Have you ever shown that 40 GHz network analyzer in a video? 🤔
not in any detail but that might change
Watching this makes me wonder over the years how much that diagnostic function would be used. My guess is that the amount of hours of labor it takes to figure problems out means it's always going to be more cost effective to replace units. It sounds like you put several weeks of time in working on this and still kept running into new problems. It's not that I don't think corporations main motivator isn't greed, it's just that it's largely impractical to repair complex electronics.
I'm sure every kind of modern gadget had this kind of mode during prototyping or manufacturing - for debugging and QA. In this case they decided to double it as a service tool. Seems logical, having service tools for complex equipment is nothing new, they probably saved some money by doing it this way
Do you review 5.12Kwh, and 11.8Kwh battery?
The AC’ness 😂
Did you renew your KitKat club membership ?
TI are suprisingly good to but parts from directly. $23.017 for the OPA637.
Yummmmm yumm ppms
Im not surprised this showed up when i search asmr
Awesome! Now do something with microwaves or even better,... generate x-rays or so!
I‘ve heard Marcos evils twin Franz does x-ray stuff
@@WurstPeterl 🤫
you Rocco Siffredy of electronics, you...
Did you mean General Radio 874?
Was machst du eigentlich hauptberuflich?
24:06 almost choked laughing.
I sense some voltnut moistness
I need a girlfriend... I mean watching CZcams at 5.30 am... she would keep me in bed until a reasonable hour. Your videos are so good, they take precedence over all sorts of things......
The time is not long gone when measuring devices with 1% accuracy were considered sufficient.
do RF
RS PRO function generator ?? 🙈
Femtoseconds please. GHz is pedestrian.
gHz plz
Femtoseconds please ⌚
😱😱😱😱😱😱😱
arf arf
I guess this means you have to take both asterisks off your old thumbnail now
heh true
Moar CNC bitte
Marco, du hast eindeutig zu viel Geld übrig 😀
Whats next? Beard-seconds!
AH yes... carbonized pcb material. I still fondly remember digging out burned conductive material from a few macbook motherboards with a blade, drill bit, or tweezers. Fill with expoxy and rebuild the traces over it. It's like cancer, I've seen it travel horizontally through the board as things get spicy.
Rindfleischetikettierungsüberwachungsaufgabenübertragunsgesetz
fs femto time
Try Microwave.
wtf do you do with all these instruments?