imperfect exponential

You know the dramatically-poor capacitance vs voltage curves for typical MLCCs. Not to mention the DA.

This is John Larkin - he hasn't bothered to show us the actual rise-time of the pulse-generator, and he is comparing it with with a half-microsecond time-constant.

TC is 1/2 ms.

The subject line says 470ms - 47u - but the post-it sticker on the scope says 47p.

It's kind of difficult to get excited about somebody who can manage to introduce a six-order of magnitude discrepancy.

You seem to have decided that he meant 0.47uF which is plausible, but extracting what John Larkin may have been doing from what he has told us about what he thinks he has been doing is something of a waste of time.

I would have written 470n if I'd meant 0.47uF, but I worked in England and Europe. John Larkin may know what nanofarads are, but he doesn't use that notation.

It's obviously a 47 uF 6.3 volt capacitor. The low-voltage RC tau is

Sloman is flat wrong, maybe more than usual this time. Good thing he doesn't design electronics.

This board I'm designing has supplies

-6

-5

-3

-2.5

plus 11 variable power supplies. That's 27 supplies, and they all need chunky bypass caps.

Which is bad enough without having to factor in the actual capacitor values at every voltage!

I sort of miss 6.3 VAC and B+.

If he's wrong, then why did you write "p" on the scope?

Tim

The video starts out with a "Fundamentals Friday" intro, so if you're just now noticing this, um...

Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website:

Then why isn't that on the Post-it note?

And Winfield Hill's comment about the dire dielectric sounds a lot more plausible.

I was certainly misled, by a guy who claims he documents electronics ...

The note has 47 nF or 0.047 uF.

But you being your usual self just wants to demean John Larkin. Nothing new here.

The waveform for the Y5V cap looks a lot like John's waveform... why would he be using Y5V caps? I know I never do. So far I've never had a need for a cap that can change capacitance as dramatically as a Y5V type cap can.

On the copy I saw it said 47p 6.3v which is a very funny spec for such a tiny value capacitor so the implication is probably an electrolytic.

My money is on 47u being the intended value and post-it is wrong.

We used to have a lot more trouble with the non-ideal behaviour of ultrahigh value precision resistors in current detectors for Faraday cup. Careful choice of compensation capacitors and time constants will allow decent compensation to a desired response if you need that.

Not what John says it was, and not what I read.

It's his performance that is bit odd here, not mine.

The 20V for a 6.3V rated cap may certainly exacerbate nonlinearities. How does it perform for 5V?

Pere

You really need to use a fantastic vintage analogue storage scope such as the Tek 466 for this kind of measurement.

Good thing he doesn't do *anything*.

It's a mu, not a p. The file name of the image is Charging_47u.JPG.

The scope time base is 200 ms/div. Are you confusing Tek's "m" for "n" too?

Do you really think it's a 47 pF cap? Does Digikey stock a 47 pF, 6.3 volt cap?

Absurd.

Sloman doesn't think: he instantly, automatically emotes. Good thing he doesn't design electronics.

The high DA distorts the falling edge curve, some memory left over from the 20 volts. The rising edge, 0 to +5, is sort of representative; the slope changes at least 2:1.

I should measure the actual C:V curve. I'm testing TPS54302 switching regulators from 1.8 to 5 volts out, and the bottom line is how the breadboards work, but knowing the actual capacitance would be good for future reference. I'll need to hack up a way to do that.

Suggestions?

I stand corrected.