Is any real need for a modern LCR to go more than 100K Hz?

[Kiriakos GR]

The question is a bit generic and close to the scientific spectrum.
But I like to read your opinions.

As far I know the 100K Hz is the today's limit about the highest sampling of the most advanced analog to digital converters.

The necessary circuitry of those converters demands the electrolytic capacitors to operate that high too.

My wild guess is that the latest LCR meters which are capable for ESR measurements up to 100K Hz, will never get out of fashion (usable for ever), because they are touching the limit.
And I am talking about the Agilent U1733C or the DE-5000, and the very few others, which are coming close enough.

I am all ears about your comments.

[tom66]

Re: Is any real need for a modern LCR to go more than 100K Hz?

As far I know the 100K Hz is the today's limit about the highest sampling of the most advanced analog to digital converters.

Huh? My Rigol has a 1GSa/s ADC. Admittedly, it is comprised of 10 x 100MSa/s ADCs, but there are oscilloscopes out there using single ADCs to get gig-sampling rates. It has a 100MHz bandwidth.

Also, LCR meters do not use 100ksa/s ADCs. They generally generate DC voltages using op-amps to calculate average RMS voltages for sine waves.

The necessary circuitry of those converters demands the electrolytic capacitors to operate that high too.

Rarely. Many do not use electrolytics on the output stage. It's far too imprecise.

My wild guess is that the latest LCR meters which are capable for ESR measurements up to 100K Hz, will never get out of fashion (usable for ever), because they are touching the limit.

In general it's not necessary for the average technician, but Agilent produce LCR bridge meters which go to 2 MHz:
http://www.home.agilent.com/en/pd-71...20-hz-to-2-mhz

Of course, the price...

[Kiriakos GR]

Re: Is any real need for a modern LCR to go more than 100K Hz?

My example of 100K Hz is about sound circuitry.

About RF, I will agree that the specific universe has different limits.

[severach]

Re: Is any real need for a modern LCR to go more than 100K Hz?

I've run across 300 Khz DC to DC converters and have read in datasheets that the green movement will have some of them running at 1 Mhz.

[tom66]

Re: Is any real need for a modern LCR to go more than 100K Hz?

I've designed a 3 MHz buck converter, but it's nothing to do with being green or energy efficiency. At higher frequencies efficiency suffers a lot. It's to do with making components small. So your phone can be smaller at the expense of slightly shorter battery life (but -- on the flip side, if the motherboard is smaller, you can fit in a bigger battery...)

[Kiriakos GR]

Re: Is any real need for a modern LCR to go more than 100K Hz?

Interesting feedback, but now I have to ask what is the highest frequency which the electrolytic capacitor is capable to operate ?

Edit: Yes the question is a bit scientific - curiosity, because I am asking for the limits of the design.

[tom66]

Re: Is any real need for a modern LCR to go more than 100K Hz?

Well they can operate at very high frequencies but they essentially become inductors and that's not a useful effect. VRM switching frequencies used to be around 200-300kHz for single phase design. But that's really pushing the limit. So you see the move to multi-phase design, at first 3 phase (which is pretty much the standard today for any basic board), and later 6, 8, 12 and even 24 phase (which is overkill, but it makes for good marketing.)

[Kiriakos GR]

Re: Is any real need for a modern LCR to go more than 100K Hz?

And so with an optimal switching frequency in the range of 250-
500kHz, the bench type LCR at 2 MHz they do have some work to do.

One last factor that I am exploring, is the resolution that is needed to have one modern LCR.
I have in front of me one specification chart of suggested ESR measurements VS capacity.
And it looks that any capacitor larger than 470uF needs a high resolution LCR.
For example the 330uF at 100KHz should have an impedance of 0.10 Ohm.
Interestingly enough the the 470uF goes down to 0.080 Ohm.
And the largest in my document, at 4700uF has an impedance of 0.022 Ohm.

[Behemot]

Re: Is any real need for a modern LCR to go more than 100K Hz?

Modern motherboard DC-DC converters already operate at frequencies as high as 1 MHz. Not only modern transistors in low-Rds(On) packages, but also industry-wide adoption of polymers allowed this. You can achieve as high as 95 % efficiency with these toys.

Even old electrolytics ultra-low ESR types go much more below your 0,022 ohm. Samxon GC 3300 uF/6,3 V has as low as 0,02 ohm. Polymers, such as X-CON, can achieve as low as 0,006 ohm depending on type (capacitance/voltage).

Good that you opened this area, just recently I got really suspicious abotu or LCR arround. Generaly, it measured most precisely arround 100-120 Hz. Going 1 kHz meant sometimes as much as 40 % lower capacitance, in the case of ultra-low ESR it measured bullshit. At 10 kHz, it was nto able to measure anything. I asked my friend for his good ESR meter and despite what that crap of ours showed me, at 100 kHz all my caps showed 100 % condition. Even those 7 years old Rubycon MBZ caps from eBay.

[Kiriakos GR]

Re: Is any real need for a modern LCR to go more than 100K Hz?

Thank you Behemot for the valuable feedback which confirms that a rich in resolution LCR, is a must have for LOW ESR capacitors.

My LCR meter DER EE DE-5000 arrived this morning and I had to run for the customs clearance paper work, and up to now I did not had the time to touch it, other than a quick inspection of the shout case.
Comes complete with tweezers (4 wires Kelvin), and a basic module with alligator clips.
Plus PC connectivity Kit for large scale Pass/Fail tests and data logging.

Shortly I will post one link with some teaser pictures for any one who are interested to see them.

[Behemot]

Re: Is any real need for a modern LCR to go more than 100K Hz?

Friend of mine has that violet UK meter, not sure what type it is…but seems fine. But my favourite still seems to be ESR Micro v. 4.0

Man so many typos I made in that previous post

[Kiriakos GR]

Re: Is any real need for a modern LCR to go more than 100K Hz?

As I promised, teaser pictures from the first DER EE DE-5000 LCR ever imported to Europe.

[mariushm]

Re: Is any real need for a modern LCR to go more than 100K Hz?

I doubt it's the first. Might be the first DER but the meter itself is available under the other brand, IET.

It was available for a long time under other names, like this for example: http://www.amazon.com/IET-DE-5000-Ha.../dp/B005EMT8PC

I'm sure someone in Europe bought it.

Dave @ eevblog even reviewed it a long time ago: http://www.eevblog.com/2011/12/31/ee...eter-teardown/

Shitty ac adapter (by the looks, don't actually know the insides), shitty method of soldering the wires of the battery to the pcb inside, shitty color on the front of the meter (puke yellow/green?)

[Kiriakos GR]

Re: Is any real need for a modern LCR to go more than 100K Hz?

There is opinions and opinions.

Either way DER EE is the ODM manufacturer.
And my first tests shown that this is faster than the Agilent U1733C.

Which probably Dave forgot to mention too.

[Kiriakos GR]

Re: Is any real need for a modern LCR to go more than 100K Hz?

The evaluation posted today.
It is a different approach on this matter, all that I did was write my thoughts in a web page.
http://ittsb.eu/de-5000.html

Regards and a Happy new Year to all.

[Behemot]

Re: Is any real need for a modern LCR to go more than 100K Hz?

I'll need some to go over that. No offense, but reading your english makes me troubles. Over so many years, I think I have more or less learned the proper english so this is not exactly my cup of coffee

Anyway, if I can ask. I think that the LCR I used and I talked about couple posts before actually may be the U1733C. Or at least something similar because I clearly remember the casing and layout to be the same. It may be some cheap copy, or not, I will try to ask what type exactly we have.

If it would actually be U1733C, do you have any idea why it could measure such nonsense while using higher frequencies?

[Kiriakos GR]

Re: Is any real need for a modern LCR to go more than 100K Hz?

I'll need some to go over that. No offense, but reading your english makes me troubles. Over so many years, I think I have more or less learned the proper english so this is not exactly my cup of coffee

Probably and the teachers which I had before 24 years ago they will agree with you, this is why my grades were at 13 to 15 ponts at the range of 20 as Max scale.

If it would actually be U1733C, do you have any idea why it could measure such nonsense while using higher frequencies?

I do have some thoughts on that, the LCR haves one basic digital circuitry which does the calculations, and also an analog part which called as divider and handles the scaling of the ranges and the test frequencies.
I am suspecting some sort of mismatch which loads the LCD with few counts even if the meter it is properly calibrated (internally) at 100K Hz.

In the real world if you like to verify such events which have to do with accuracy at the level of 0,3% you need equipment with 10 times better accuracy, which I do not have.

Edit:
At the electrical specifications in the Agilent requirements are that the user should add 50 counts at the 2.0 Ohm range for max accuracy in DCR and Rs.
When the DER EE LCR requirements only three counts.
0.7% + 50 VS 1.0%+3 (DCR)
Rs 100K Hz| 1.0% + 50 VS 2.0%+3

This is a good exercise to be solved only in the walls of a calibration lab with a reference resistor at hand.


.