signal generator output

quarter

the

unterminated

You do it to yourself.

John

Really. Assume a shielded black box with a coax connector coming out one face. We know that there is an ideal 50 ohm signal generator inside, with an ideal 50 ohm coax of unknown length inside, between the generator and the connector to the outside. The generator conveniently makes flat frequency sweeps, square wave sweeps, DC levels, and unit impulses from time to time. We know everything about the levels and waveforms except exactly when each is begun.

There no way to make measurements on that signal from outside the box that tells you whether there is a piece of coax inside the box, or how long it might be. The connector just looks like a 50 ohm source.

John

quarter

the

dipping

unterminated

--
I don't think so, and your snippage is telling...

I'm usually careful about what I post, and I count my P's and Q's, and
I mostly show my work so my thinking can be followed and any errors I
may make can be critiqued on that basis.

Kinda like peer review.

You, on the other hand, are dishonest in that you use subterfuge to
downplay your errors and you like to play fast and loose, so that when
anyone catches you in a technical error you use that looseness to
evade owning up to your errors and, instead, try to make it seem like
something was wrong with whoever caught your error.

generator,

quarter

the

dipping

unterminated

What error did I make here? I predicted the frequency response the OP would see, with the coax driving a hi-z scope input. I predicted he'd see no frequency peaks or dips at the scope, just flat with a corner around 200 MHz, and a 6 db/octave rolloff after that. Basic stuff, actually.

You and Mr Personality predicted mountain ranges of peaks and dips caused by standing waves.

You sim'd it and I was right. So why are you mad at me?

John

Good explanation. with a matched source the coax is immaterial, it's asif it wasn't there.

thanks.

--
?? 100% natural

--
I'm not mad at you, I just don't like you.

Why?  Basically, because you're nasty and sneaky when you don't get
your way or are found to be in error, and generally just
mean-spirited.

For instance, was that crack about Phil necessary?  Of course not, and
all it can do is increase the enmity around here, which you seem wont
to do in an effort to belittle everyone else and exalt yourself, just
like Sloman does.

As far as the SWR thing goes, I went to the trouble of simulating it
to determine what the truth of the matter was, and accepted the result
which proved you were right.

Had you been wrong though, you'd still be dancing around looking for a
"What I meant was..." loophole to crawl through and generally blowing
a lot of smoke around in order to try to create confusion.

We all make mistakes, but you seem to be on a crusade bent on having
yours be seen as trivial, no matter how serious, and ours serious, no
matter how trivial.

This is a public forum. Play the game or whine.

He shrieks and curses at everybody. Everybody. He deserves no respect, and has demonstrated that it doesn't help. And his expertise stops at

20 KHz. Best thing to do is laugh at him, which at least makes him useful.

Of course not, and

Don't be silly. I'm friendly and helpful to anybody who is reasonably sincere, and have made several personal in-the-flesh friends from s.e.d. I post schematics. I send people parts. I'm polite to anyone who is polite. Usenet does have an infinite supply of anonymous jerks, and there's no reason to be polite to them.

YOU escalate after you post stuff that's wrong, which you do fairly often. My earlier posts about the "standing wave" thing were factual. If, at that point, you had checked your claims, you might have avoided a dispute, and maybe learned something to boot. You and Phil jumped on me when you should have been thinking.

*AFTER* making silly and insulting statements.

I'm usually right because I post about stuff I know (I live transmission lines) and because I'm (usually) careful.

No. When I'm wrong I correct myself. It's no big deal top make a mistake occasionally. When we actually lay out boards in my shop, we design review and crosscheck exhaustively; I don't do that in newsgroups, so of course I make the occasional mistake. Electronics is complex, and newsgroups don't matter.

I pointed out, in this case, that the OP would not see the effects of standing waves at his scope input. No peaks, dips, mountain ranges. The OP was being misled, lied to almost, and that was not trivial.

And I thought you plonked me.

John

--
Telling you what you are is "whining"?

Thanks for the heads up, John...and to think my sister was about to go out with him...one shudders just thinking of what could have happened.

mike

I don't know what you mean by "read accurately." Attenuators add a calibrated loss in a specified environment.

Assume a 50 ohm generator driving some arbitrary load. Measure the voltage at the load. Now insert a 20 dB 50-ohm attenuator. The voltage at that same load drops by 10:1, namely 20 dB. The attenuator did what it was advertised to do, drop the signal level by 20 dB. That works for any linear load, any load impedance.

You probably need to Spice that one, too.

I don't really know what your point was. A 50 ohm attenuator, driving a 50 ohm cable, works just like the same attenuator with no cable. The cable only adds time delay.

Countercase, recent, in s.e.d.

John

skipping the rest

Ooh, you could have been uncle to a tiny Fieldslet.

John

--
Ah, a breath of fresh air in this stiflingly confrontational arena!

Thanks for the chuckle. :-)

--
Indeed, and if the load impedance is different from that specified,
then the environment will have changed and the loss will be different
from what was predicted for the specified environment.

I'm obviously going to need one. I'm not giving up on this. I love the project ( photomultiplier with scintillator) and have investeted in the ortec equipment , its just that the counter isn't registeing anything so I wanted a view of the pulse. I'll try Mr. Larkins suggestion of taking the pulse off a 50 term coax and feeding that to the pulse shaper. It coulld be the parasitic preamp. Yes , I'm certainly a babe in the woods, :) jk

"Arbitrary" means "any".

dB.

You can't afford a scope, or a voltmeter?

Maybe you got a bad attenuator. All of mine do what I described.

Hey, here's the guts of a VAT-20:

ftp://jjlarkin.lmi.net/VAT-20.zip

Thickfilms, only good to 6 GHz. I wonder what that gold dot in the middle is all about.

John

That could be true, but not the ones I've played with, although it's a pretty short list. 8116A and 33120A generators are calibrated at 50 ohm impedance.

On Tue, 14 Dec 2010 19:57:28 -0800, John Larkin wrote:

--- Yes, I know what "arbitrary" means, and I also know what "moving the goalpost" means, and it seems that's precisely what you're trying to do.

Here; let's see if we can't tighten thing up a little:

In order for a signal generator to operate as specified, the condtions under which it's supposed to work have to be specified, and those conditions are usually that the impedances of the generator, the transmission line, and the load all be equal and that all reactances cancel.

Let's say we have a variable frequency signal generator, with a 50 ohm resistive output impedance, which sports a variable attenuator.

Now, let's say that we connect one end of a coaxial cable with a characteristic impedance of 50 ohms to the generator, and connect the other end of the cable to a 50 ohm resistive load. OK so far?

Now we set the frequency to some arbitrary value and the attenuator to, say, 0dBm, all without measuring anything at either end of the cable.

To check the attenuator, though, we measure the voltage across the load and we get 632 millivolts, peak to peak, which is just what we expected.

Now, remove the 50 ohm load, replace it with a 15pF capacitor, and measure the voltage across the cap. Will it be 632 millivolts, peak to peak?

Only at one frequency, so it's not likely, which means that the attenuator's 0dBm setting is meaningless. In other words, it's not reading accurately, particularly because the load is mostly reactive and is dissipating only a miniscule amount of power.

Version 4 SHEET 1 1388 680 WIRE -784 128 -816 128 WIRE -640 128 -704 128 WIRE -464 128 -544 128 WIRE -224 128 -304 128 WIRE -64 128 -128 128 WIRE 48 128 16 128 WIRE -640 160 -672 160 WIRE -512 160 -544 160 WIRE -224 160 -256 160 WIRE -96 160 -128 160 WIRE -816 176 -816 128 WIRE -464 176 -464 128 WIRE -304 176 -304 128 WIRE 48 176 48 128 WIRE -816 288 -816 256 WIRE -672 288 -672 160 WIRE -672 288 -816 288 WIRE -512 288 -512 160 WIRE -512 288 -672 288 WIRE -464 288 -464 256 WIRE -464 288 -512 288 WIRE -304 288 -304 240 WIRE -304 288 -464 288 WIRE -256 288 -256 160 WIRE -256 288 -304 288 WIRE -96 288 -96 160 WIRE -96 288 -256 288 WIRE 48 288 48 256 WIRE 48 288 -96 288 WIRE -816 368 -816 288 FLAG -816 368 0 SYMBOL tline -592 144 R0 WINDOW 0 -7 -81 Left 0 WINDOW 3 -74 -50 Left 0 SYMATTR InstName T1 SYMATTR Value Td=10n Z0=50 SYMBOL res -688 112 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R1 SYMATTR Value 50 SYMBOL tline -176 144 R0 WINDOW 0 -7 -81 Left 0 WINDOW 3 -74 -50 Left 0 SYMATTR InstName T3 SYMATTR Value Td=10n Z0=50 SYMBOL res -448 272 R180 WINDOW 0 -35 77 Left 0 WINDOW 3 -32 41 Left 0 SYMATTR InstName R2 SYMATTR Value 50 SYMBOL voltage -816 160 R0 WINDOW 3 18 128 Invisible 0 WINDOW 39 0 0 Left 0 WINDOW 0 18 100 Left 0 SYMATTR Value SINE(0 .632 200e6) SYMATTR InstName V1 SYMATTR Value2 AC 1 SYMBOL res 32 112 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R3 SYMATTR Value 50 SYMBOL cap -320 176 R0 WINDOW 0 -23 8 Left 0 WINDOW 3 -69 60 Left 0 SYMATTR InstName C1 SYMATTR Value 15E-12 SYMATTR SpiceLine Rser=0 SYMBOL voltage 48 160 R0 WINDOW 3 18 128 Invisible 0 WINDOW 39 0 0 Left 0 WINDOW 0 18 100 Left 0 SYMATTR Value SINE(0 .632 200e6) SYMATTR InstName V2 SYMATTR Value2 AC 1 TEXT -792 344 Left 0 !;ac oct 255 10e6 1000e6 TEXT -784 312 Left 0 !.tran 1e-7

Snipped some buffoonery...

---

--- JF

Any interest in a 54522A? It's a two channel 500MHz 2GSa/s scope. Some screen burn, but display is still nice and bright. I'd sell it for $750 + shipping with a 30 day warranty.

You can see it on Ebay item 190384647489 I do the repair work for the guy, and this scope he's selling is mine. (In fact just about all his scopes for sale are mine.)

What I did was state that a 20 dB attenuator, after a 50 ohm source, introduces 20 dB of attenuation no matter what the load impedance.

Technically, an attenuator is set to some number of dBs, not dBm's. But ok...

It's a 50 ohm source. It makes a voltage divider with any impedance you connect to it. That's exactly the answer to the OPs original question. Why make it complex?

John