New Video: Parametric Oscillations

It's a bit off-topic from the channel, but hopefully fun.

formatting link

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott
Loading thread data ...

Fun! You might like Pippard's "the physics of vibration" vol I.

formatting link

He talks about parametric oscillators (PO). One thing I'd like to see is that there is a minimum level of oscillation amplitude that you need before the PO can grow. Did you see that with your pendulum? This does lead to the curious fact that you can make a swing that is too high to pump up. (I accidentally made such a swing in my back yard.)

Did you ever try standing on a swing and flexing your legs up and down to pump it? (Careful, I almost broke my ass that way. It was the only way I could get my "too tall" swing to work, without dad pushing you.)

Finally someone needs to make a PO using the voltage coefficient of a crappy ceramic cap. (There again there will be some threshold level to get it started.)

Reply to
George Herold

Oh as far as uses of parametric oscillators, There are optical PO's using non-linear xtals to make different colors of laser light. (Extremely complicated beasts, I used one as a post doc, but was not allowed to tweak it.)

George H.

Reply to
George Herold

George, I was (and am) allowed to tweek laser OPOs and OPAs. You missed early grey hairs, that's it.

Steve

Reply to
sroberts6328

Not experimentally, but it did fall out of the math. I didn't manage to beat the math into the ground (it, OTOH, did a number on me), but I did get as far as proving to my own satisfaction that bouncing the pivot up and down subtracts from the damping factor of the system in a way that's related in some complicated way with the pendulum's effective length, resonant frequency, period of the bouncing. The damping factor does go down in a way that's proportional to the absolute value of the bouncing, or the absolute value squared.

Makes sense.

Nope. But I could see it.

Get to it!!

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

My copy of Pippard is at home. I wanted to quote a bit out of parametric excitation. It's got my pencil scribble of "excellent!!" at the end, so I obviously have read it before, but forgot. (top of page 287, I'll scan and post the whole thing, tomorrow at work... He works out in a page that, a_c = 2/3 *l/Q where a_c is critcial amplitude modulation, l is the total length, and Q the Q.)

"There is a certain critical excitation which allow the oscillation to persist at constant amplitude. It is virtually impossible to exhibit this in practice, but the reader is encouraged to try, since more can be learned about the physical processes by being frustrated than by accepting defeat at the bidding of the printed page."

George H.

Reply to
George Herold

This was in the late 90's when I was at Vanderbilt. expensive stuff. (maybe $1M in 1999.) I was walked through the optics chain. (most is now a blur..) What sticks out in my mind where these big tilted curved gratings, where they expanded (in time) a pretty fast pulse, compressed it, (with something I've forgotten) and then put it back together with the other grating but much shorter. non-linear optics is close to magic, from my point of view. :^)

George H.

Reply to
George Herold

Actually I found it pretty easy to get an excitation value which allows the oscillation to persist at constant amplitude -- but the greatest source of damping in my system is wind resistance, which goes by the square of velocity, so the system as I implement it essentially has a varying damping ratio.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

I think what he meant was to find the excitation amplitude that just makes it grow. Here's a scan of the first 4 pages of chapter 10.

formatting link
(Sorry I don't know how to flip the pdf's over.)

George H.

Reply to
George Herold

(most is now a blur..)

tilted curved gratings,

and then put it back

from my point of view. :^)

A fibre-grating pulse compressor. The self-phase modulation in a holey fibr e broadens the spectrum and dispersion applies a linear chirp to the output pulse, with longer wavelengths arriving first. A zigzag path between two p arallelled gratings fixes the relative delays, because longer wavelengths g et diffracted through larger angles, resulting in longer path delays. If yo u get it right, you can make transform-limited pulses.

Steve used to work for the outfit where I got my fancy tunable OPG system ( Altos Photonics). Continuously tunable from 420 nm to 10 microns with a sma ll gap near 710 nm where the OPG became degenerate. (It was pumped with a t ripled YAG laser at 355 nm.)

The pump laser was much harder to keep working than the OPG! It was pretty cool, but I don't miss it.

Cheers

Phil Hobbs

Reply to
pcdhobbs

bre broadens the spectrum and dispersion applies a linear chirp to the outp ut pulse, with longer wavelengths arriving first. A zigzag path between two parallelled gratings fixes the relative delays, because longer wavelengths get diffracted through larger angles, resulting in longer path delays. If you get it right, you can make transform-limited pulses. Right, I don't remember any fiber.

(Altos Photonics). Continuously tunable from 420 nm to 10 microns with a s mall gap near 710 nm where the OPG became degenerate. (It was pumped with a tripled YAG laser at 355 nm.) Ahh, a fine skill, it's not a field I've at all kept up with. I assume they are doing more and more with fiber these days.

y cool, but I don't miss it. Grin, It was another staff guy at the FEL who was tasked with the OPA/OPO.. We would hang out after work/ weekends some, so I did hear some tales of woe. (I wasn't much of an optics guy then... not that I'm much of an optics guy now. :^)

George H.

Reply to
George Herold

Wonderful! Ed

Reply to
ehsjr

Thanks! I thought it was shitty.

That's not a comment on your opinion -- I often finish a talk or a book or whatever thinking "gawd, why am I not covered in rotten vegetables?", only to be accosted by people wanting to _thank_ me for my work.

OTOH, I can finish something up, think "hey, this is pretty good!", inflict it on an unsuspecting world, and find out that no, in fact, it was a steaming pile of crap (very powerful! Makes things grow!).

I've decided that I'm not a very good critic of my own work.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

Well, i thought that the editing done was implemented very nicely. The resulting "jumps" or "gas" were rather smooth and only a very professional system could improve it, along with multiple time-consuming re-enactments for more accurate body placement and hand-motion merging. In a word, this ain't Hollywood and we are not seasoned or professional actors. WELL DONE!

Reply to
Robert Baer

I read somewhere, on the blog of some YouTube biggie, that if you're doing a "talking head" video then the quality of the sound is far more important than getting the video perfect. I also noticed that quite a few of the "talking head" video channels that I watch will have even more sudden visual jumps than I use, and I just don't notice them unless I concentrate.

Everything is recorded on a Samsung Galaxy S5 cell phone, and edited using kdenlive open-source video software. I'm pretty amazing that I can do so well on stuff that I could get for free, or had lying around.

--
Tim Wescott 
Control systems, embedded software and circuit design 
 Click to see the full signature
Reply to
Tim Wescott

I have no work for you unfortunately, but I was trying to send a comment about "Implementing the PID Controller in Software" when Gmail complained about your email address. Probably a space in the address is too much for Gmail.

By the way, this was my comment or question. Hi

I read about your PID videos on the newsgroup and watched some. In "Implementing the PID Controller in Software" you talk about integrator wind up. I thought when I watched the video, that in one moment the error is 0/zero. So how about resetting the integrator to zero when the error is zero, what ever it means physically.

By the way, years ago opamp integrators had this problem because of opamp bias current and offset.

Reply to
LM

That would be bad, because you'd be resetting the integrator just when it'd finally achieved its goal.

A much more effective integrator anti-windup measure is to hold the integrator at zero when the error is _high_, the theory being that the integrator's job is to clean up residual messes, not to help you get to the target more quickly. Another method, which works when your system has discrete jumps in the command (i.e., when the machine is at point A and you want to command a move to point B) is to change the target point, then suppress integrator action for some fixed period of time.

I haven't used either of these methods, but I know they're out there.

Opamp offset is a different problem altogether -- in that case, if you're using an opamp as an integrator, the various offsets mean that you servo to somewhere close to, but not exactly on, the right target. You have the same problem with digital systems, because no matter how fancy you get you'll always measure the position wrong somehow, and then drive until your (unavoidably erroneous) position measurement is dead on to the target value.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

Yeah, the few I've done on my own had crap for sound, which did, indeed, make them horrid.

Agreed yours look quite good.

--
This email has been checked for viruses by Avast antivirus software. 
https://www.avast.com/antivirus
Reply to
eric.jacobsen

I actually had to discard about 1/3 of a video's worth of footage on this one and re-do it. I had the lapel mic contacting my chin, and even with a smooth shave every time I moved my head the damned thing scratched.

Fortunately I did a spot check that 1/3 of the way in, or I would have had to re-record everything.

--
Tim Wescott 
Control systems, embedded software and circuit design 
 Click to see the full signature
Reply to
Tim Wescott

Thanks.

Would working a yo-yo be an example of PO? As the yo-yo descends and unwinds, you pull up on the string to speed it up, adding energy.

Cheers

--
Clive
Reply to
Clive Arthur

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.