why the signal from a 50-ohm output impedance device should be checked by a scope channel with a 50-

[for_idea]

As I understand, if the scope channel has 1 meg ohm input impedance,
the received signal would be stronger since most of the voltage is
taken by the input channel. Why people always say 50 ohm is a proper
input impedance for the scope channel?

thanks,

Zhi

 

[Tim Shoppa]

If the source is designed to be operated into a 50-ohm load, that's the
way it should be tested. Most scopes these days have a switchable
50-ohm termination built-in to make this easy.

Nothing wrong with doing it with a scope with 1Mohm input impedance as
long as you've got a 50-ohm terminator Teed off at the input too
(that's the way it used to be done before scopes had built-in 50-ohm
terminations.)

Tim.

 

[Meindert Sprang]

If the source is designed to be operated into a 50-ohm load, that's the
way it should be tested. Most scopes these days have a switchable
50-ohm termination built-in to make this easy.

Oh? Wich ones?
My Tek doesn't and it seems a dangerous feature to me...

Meindert

 

[for_idea]

Thanks. but what would be wrong if I don't use a 50 ohm terminator and
connect the signal directly to a 1meg ohm channel? I guess I will get
as much as twice of the signal observed by a 50 ohm channel since most
of the voltage will be dropped on the 1meg ohm.

 

[for_idea]

I have a Lecroy 9341 which has several coupling modes including 1 meg
and 50 ohm.

 

[Tim Shoppa]

Thanks. but what would be wrong if I don't use

a 50 ohm terminator and connect the signal directly
to a 1meg ohm channel? I guess I will get
as much as twice of the signal observed by a 50 ohm
channel since most of the voltage will be dropped on
the 1meg ohm.

For DC and low frequencies (<< 1MHz without sharp edges) you're mostly
right.

At higher frequencies (and even low frequencies with sharp edges) not
terminating the 50 ohm line means you will see lots of reflections.

Don't they teach this transmission line stuff in like
Freshman/Sophomore physics anymore?

Tim.

 

[Tim Wescott]

Thanks. but what would be wrong if I don't use a 50 ohm terminator and
connect the signal directly to a 1meg ohm channel? I guess I will get
as much as twice of the signal observed by a 50 ohm channel since most
of the voltage will be dropped on the 1meg ohm.

What happens will depend on the source. If your source has a 50 ohm
source impedance then your prediction will be true. But if your source
is something like an RF power amplifier then it will be to work _into_ a
50 ohm impedance but it won't _have_ a 50 ohm impedance. Some RF
amplifiers aren't even _stable_ working into such a severely mismatched
load, so you'd get totally meaningless results. Even assuming the thing
stays stable, depending on the output coupling from the amp you could
get anything from almost no voltage to 10 times the loaded voltage.

So if the source wants to see 50 ohms -- give it what it wants.

 

[for_idea]

"if the source wants to see 50 ohms -- give it what it wants"... sounds
reasonable

 

[for_idea]

as I understand, with 1MHz, the wavelength is 300meters, if the
interconnecting length is short (such as 1 meter or less), the
transmission line effect is insignificant and the reflection is not a
major problem to the detecting signal. Even with the signal I'm using,
3MHz, I guess I still don't have to really count the transimission line
effect in. In this case, we can treat the signal as if it is "DC". I
still don't understand why the advantage of connecting a 50ohm output
to a 50 ohm input.

Thanks,

Zhi

 

[Tim Shoppa]

and the reflection is not a major problem to the

detecting signal. Even with the signal I'm using,
3MHz, I guess I still don't have to really count the transimission
line effect in.

If it's a 3MHz square wave with 10ns rise/fall times, then you really
need to consider the edges at least as having 300MHz components.

Tim.

 

[for_idea]

what if it is sinusoidal wave? thanks.

 

[Mac]

Oh? Wich ones?
My Tek doesn't and it seems a dangerous feature to me...

Meindert

Dangerous how?

--Mac

 

[Phil Hobbs]

what if it is sinusoidal wave? thanks.

If you know that already, what do you need the scope for?

Cheers,

Phil Hobbs

 

[Mac]

Thanks. but what would be wrong if I don't use a 50 ohm terminator and
connect the signal directly to a 1meg ohm channel? I guess I will get
as much as twice of the signal observed by a 50 ohm channel since most
of the voltage will be dropped on the 1meg ohm.

Well, if the source really has a 50-Ohm source impedance, then you will
see twice the Voltage signal, yes.

But you may often find sources which are designed to drive 50-Ohms but
which do not actually have a 50-Ohm source impedance. Such a source may
become squirrelly if it sees a high-impedance load.

Even if the source doesn't mind driving a high-impedance load, probing it
that way will not tell you what the signal looks like with a 50-Ohm load,
obviously.

Anyway, why don't you ask the person who told you this to explain his/her
reasoning?

--Mac

 

[Paul Burridge]

What happens will depend on the source. If your source has a 50 ohm
source impedance then your prediction will be true. But if your source
is something like an RF power amplifier then it will be to work _into_ a
50 ohm impedance but it won't _have_ a 50 ohm impedance. Some RF
amplifiers aren't even _stable_ working into such a severely mismatched
load, so you'd get totally meaningless results. Even assuming the thing
stays stable, depending on the output coupling from the amp you could
get anything from almost no voltage to 10 times the loaded voltage.

So if the source wants to see 50 ohms -- give it what it wants.

Nice one, Tim. That's by far the most outstandingly clear answer to
this (entirely creditable) question yet!

 

[redbelly]

what if it is sinusoidal wave? thanks.

Sinusoids are no problem if the wavelength is a lot longer than your
cable (use a velocity of 2/3 speed of light to calculate wavelength
from frequency).

In grad school I was using rf signals where the wavelength was
comparable to the cable length. With no 50 ohm terminator, the scope
would show resonances as the frequency was tuned due to standing waves
building up in the cable at the resonances. Even though it was a
"constant amplitude" source, the amplitude varied wildly as the
frequency was tuned.

Putting a 50 ohm at the scope put an end to that problem. As others
have said, the 50 ohm prevents the signal from reflecting back and
forth within the cable. This is important for short pulses, and square
waves with short transitions between steps, and high frequency sine
waves.

Mark

 

[Phil Allison]

"Mac"

Dangerous how?

** It would be very easy to forget the switch was set to the 50 ohm position
and connect the input to a source of high voltage. Phut 50 ohm resistor.

Very easy to "phut" some circuit you are probing too.



............ Phil

 

[Pooh Bear]

Thanks. but what would be wrong if I don't use a 50 ohm terminator and
connect the signal directly to a 1meg ohm channel? I guess I will get
as much as twice of the signal observed by a 50 ohm channel since most
of the voltage will be dropped on the 1meg ohm.

If you're scoping a signal then it's somewhat unlikely that the scope is
the intended load in actual use.

The signal should be terminated at its normal destination.

Graham

 

[Mac]

"Mac"



** It would be very easy to forget the switch was set to the 50 ohm position
and connect the input to a source of high voltage. Phut 50 ohm resistor.

Very easy to "phut" some circuit you are probing too.



........... Phil

This has never happened to me, and in practice, I don't think it is as
easy to do as you suggest.

The scope input can be set to 50-Ohms. The PROBE input cannot. Also, the
probes usually have some extra pins on them that force the scope into the
correct mode for that probe. If it is an active probe, then the input will
be set to 50 Ohms, and if it is a passive, it will be set to 1M. If you
connect your own coax, then you can set it to either one.

So to phut something, you would probably have to jury rig a probe onto
some coax. Besides, most things don't phut on 50 Ohms.

As for phutting the input termination, I don't know how much power it
takes to do that, but I'm sure it is possible. Like I said, I've never
done it.

--Mac

 

[Phil Allison]

"Mac"

This has never happened to me, and in practice, I don't think it is as
easy to do as you suggest.

** OK - we have ourselves a " live one " here folks !!!!

The scope input can be set to 50-Ohms. The PROBE input cannot.

** Probes that are 1:1 or use clips or plugs on a length of co-ax
certainly can.

Millions of them in use.

So to phut something, you would probably have to jury rig a probe onto
some coax. Besides, most things don't phut on 50 Ohms.

** Do you have any experience with electronics at all ???

Connecting arbitrary internal circuit points to ground via 50 ohms is
lethal.

As for phutting the input termination, I don't know how much power it
takes to do that, but I'm sure it is possible.

** So your grasp of Ohms Law is hazy too ???

What do you expect the 120/240 AC supply might do to a 50 ohm 1/2 watt load
resistor ??

Like I said, I've never done it.

** I suspect that is a universal truism in your case - Mac.



........... Phil