Or use a memory recorder. Memory recorders have isolated inputs.

Or use isolated probes, but those aren't cheap. Probably the cheapest but safe solution for hobbyists is to use an inverter powered by an isolated DC source.

I use two isolating transformers, one for DUT and another one for instruments.

@@DmitryOrlov_dimorlus that is certainly possible. Get the right tool for the job! Or as my friend Ed always said: ‘because you could doesn’t mean you should’

@@NiHaoMike64 Transformer is much chipper and reliable, and doesn't create noise (EMI).

The signal to noise ratio is excellent!

I see this failure mode all the time as well, although I often find a design will splurge on a name brand 400 V bulk filter capacitor, but the bootstrap capacitor will be some Dongguan back alley special in the smallest and highest ESR package they can find (much smaller than the ones in this video). Makes me rethink where I should be spending my BOM.

The reason the startup cap has such a high voltage rating is that if the power supply doesn't start for any reason/fault the voltage in the cap can go way higher that 50V

Thank you!

Looks like AIN to me

AlN is quite common for these types of applications. You can find it almost in every mass product butterfly-like assemblies. Zerodur will be overkill.

@@enclis yes also from the color, an i think it is hard on other material to make solder contact surfaces that hold well. Zerodur is way too expensive ^^

That would be very interesting. I hope we can find the time!

Yep. Our applications of gas lasers would have a simple ammeter on the driver input. Once the gas started to get very low, you’d see a high draw of current and know it was soon to fail. In our case, we went through hundreds of the exact same unit, so the OEM would remanufacture them - however the reman units had a substantially shorter life.

Most wavemeters nowadays still use HeNe lasers because a stabilized HeNe laser has a relatively precise absolute wavelength to ~ppm level. Most solid state laser are not absolute and can vary up to several nm so it’s easier to use a HeNe as a light source so the instrument doesn’t need to be designed to accommodate calibration over several nm.

@@wadehsu2347 Solid states laser can be made as frequency stable as HeNe lasers, but those are extremely expensive and HeNe lasers also have very tight beam that is hard to get on solid state lasers. So thus its way cheaper to get a HeNe laser which it very stable as is, only disadvantage is that is bulky and need HV but they can be made quite small if not need any real power from them.

dtiydr yes, solid state laser can be frequency stable, but they are not inherently accurate and not as stable as stabilized HeNe laser. So each laser will need to be measured and calibrated and the range of that calibration can vary up to 10nm. Unless the laser diode are lock to an atomic reference then the accuracy becomes absolute, but that come with a price of complexity and wavelength typically in the IR for common atomic reference such as Cs or Rb. Also, long term stability of single frequency diode laser are poor if not locked and they are also temperature sensitive. So to maintain the wavelength stability the temperature and current of the laser diode will need to be regulated to better than 1mK and ppm stability.

@@wadehsu2347 All correct and thus their price and in many case a HeNe is easier to control that part but not as wide band and also bulkier which could be a constrain sometimes but otherwise its pretty much the best choice and no doubt cheapest.

At 50mA it was putting out about -3dBm.

We make phased arrays to not have to do that! ;)

I appear to be second. Where's the party?

3

There are a lot of visible light tunable lasers on the market. If you meant Santec only, it turns out that relatively low power visible light CW+external modulation (not pico/femto-, supercontinuum like, etc) tunable lasers that are probably can be done by Santec is not that of an interest for the market. We have a lot of cheap mass-product fiber coupled NIR tunable lasers due to telecom WDM systems.

If you have the money: www.newport.com/mam/celum/celum_assets/resources/Vantage7100_datasheet.pdf?1 Tuning ranges are quite narrow though.

​@@LesLaboratory Or you can just take supercontinuum source ( www.yslphotonics.com/Home/Index/Product/details/id/50.html ) and filter out almost any frequency you want from 410nm to 2400nm

This is due to a phenomenon called stimulated emission. It is the fundamental thing that makes lasers work and produce coherent radiation. Basically, a photon that is reflected back into the LD will stimulate the recombination of electrons and holes inside the LD. The photons emitted in the recombination are in phase and of the same wavelength as that original photon, and they stimulate further recombination and so on. Eventually, the large majority of the photons are produced this way and you get laser light.

It is a power supply, and the output is DC. The transformer runs far higher than 400Hz, usually in the hundreds of kHz or even MHz.

Also interesting that some agency (government funded lab or possibly college lab) disposed of an instrument that any reasonably experienced tech could have easily repaired. Hey, the taxpayers have endless amounts of cash available!

Which DMM are you referring to?

@@Thesignalpath Keysight U1282A 😁

Yes, it was a wavelength meter as well.

It was TSP#29, Teardown, Repair and Experiments of an Agilent 86120B Multi-Wavelength Meter kzfaq.info/get/bejne/rb6Gfqt9vbWRoKM.html

Thanks for the point. I am not sure if I agree with it. The term _calibration_ in this industry has a well-defined meaning. The calibration process ensures that the instrument produces or measures the desired parameters based on its specifications. The calibration process achieves this by either making adjustments to the unit or by applying appropriate offsets, lookup tables, etc. in software. Regardless on how it is done, this process is called calibration for test & measurements instruments. If you ask me _how_ calibration is performed on a _specific_ instrument, then I can describe whether it is done by adjustment, or otherwise.

@@Thesignalpath yes I agree in the industry it is quite common to say calibration to both. I even noticed when I was younger I often heard it and though calibration also meant adjustment in the equipment, but I learned only later that people often used it that way. And I think even a lot of technical people who are not directly involved with calibration of Equipment probably think that they get adjusted when they get calibrated, but that's rarely the case. For most industries, at least. I also agree that we use the term as a whole of the process, but by definition an adjustment can be only a subsequent result from a Calibration result. As also NIST Handbook defines, "Strictly speaking, the term "calibration" means just the act of comparison and does not include any subsequent adjustment.". But yes, I'm splitting hairs, maybe. Also, it is not difficult or even hard to change the usage. Would be interesting to ask engineers but also hobbyist, I think you would be surprised on how many don't know what happens when an instrument gets calibrated, maybe that would be a video idea, also if possible with a cal lab as a cooperation.

The calibration chapter in every device's service manual ever printed disagrees with you.

@@stargazer7644 does not mean it is correct Nist and the German PTB see it the same way i do it, i also quoted it before. i think if PTB does say it so i would think they have a point. you will also see if you buy a calibration service you won't get a adjustment you have to buy a adjustment separately form every service provider, proving the point further

@@heinrichhein2605 Well, how about checking a dictionary? According to Merriam Webster, one definition of calibrate is to measure precisely against a standard. That's the sense you're using it in. Another definition of calibrate is "to adjust precisely for a particular function". That's the sense we're using it in.

Thanks, I added a pinned comment about this.

As the immortal Q mused to Picard: "If you can't take a little bloody nose, maybe you ought to go back home and crawl under your bed?" He went on to warn that: "It's not safe out here. It's wondrous, with treasures to satiate desires both subtle and gross - but it's NOT for the timid".

Said this because he's immortal, right ?

@@thomasw6169 Thanks, I added a pinned comment about this. And yes, I am also immortal.

@@Thesignalpath Haha, right. And it's easily forgotten during work that the whole case (if metal) can become live depending on how you connect your probe ☠️