1 Mohm coaxial cables
Arthur Lee Kam Hung
like the ones used in oscilloscope probes.
Could some one help to let me know of any sources or makers ?
The Hong Kong Crystals Company,
Technical Director,
Arthur Lee, artl...@hk.super.net
Tom Bruhns
: I am looking for coaxial cables with 1 M ohm nominal impedance, just
: like the ones used in oscilloscope probes.
: Could some one help to let me know of any sources or makers ?
I believe you are under a misunderstanding. The cables used in
oscilloscope probes are coaxial, and high impedance for coaxial, but by no
means are they 1 megohm. Without coiling the inner conductor, you would be
obliged to use an inner conductor of diameter smaller than 10^-700 times as
large as the outer conductor inside diameter--or far smaller than
fundamental particles, much less atoms.
Check with manufacturers of coaxial cables. I suspect most of the scope
probe cables are custom-made for (or by) the probe manufacturer.
Frank Miles
Steve Rickman <ste...@quiknet.com> wrote:
>Sorry, but oscilloscope probe cables do not have 1 megohm nominal
>impedance. The input impedance of general purpose oscilloscopes is
>typically 1 megohm. The cable impedance is in the range 30 - 75 ohms,
>but 10X probes contain an internal RC voltage divider which produces a
>probe tip impedance of 10 megohm. Cables with characteristic impedance
>higher than about 300 ohms are not practical.
Quite so. To expand on this, note that the impedance of a coaxial
transmission line is 60*SQRT(LOG(k^2)), where k is the ratio of the
shield to inner conductor diameters. Varying k from 100 to 10000 changes
the line's impedance from 276 ohms to 552 ohms.
To get to 1Mohm, you would need a k of e^16666.6, which is approximately
10^7238!!!!
This is considerably too big for the known universe, even making the conductor
out of a single line of electrons.
Hey, this stuff is *neat*!
-frank
Frank Miles
Arthur Lee Kam Hung <artl...@hk.super.net> wrote:
>
>Thanks for your comment.
>
>However, I am still confused how does a X1 passive oscilloscope probe
>work.
>
>The input impedance at the probe tip is 1 M. Yes, but if the cable is
>50 Ohm, then there must be some wide band ( DC to some hundreds MHz )
>matching network to match 1 MOhm to 50 Ohm, and then at the same time,
>there must be some wide band DC amplifier in the oscilloscope to
>amplify the signal, of which the gain could be in the range of
>1,000,000 / 50 = 20,000.
>
>Is it at all possible to design such network and amplifier at such low
>cost ?
>
Confusion is understandable, since probes and cables and attenuators
are fairly messy things (at high impedances and frequencies).
One thing that might allieviate some of this is to realize that the
probe's input impedance is not purely resistive. It has a significant
input capacitance. One might imagine, for example, that the input impedance
for a mythical probe cable of 50 ohms might be 500 ohms at some very
high frequency; the probe body would have a 450 ohm resistor to provide
the necessary 10x attenuation (this would be coupled through some
capacitor). At very low frequencies, it would be 1Mohm.
The probe body would contain a 9Mohm resistor for the 10x attenuation.
Additional components might be necessary to "make it all come out right".
If you look at the impedance plot for your scope probes (you did buy good
ones, didn't you, with full specs?), you will find that their impedance
is pretty messy. I have glossed over a lot.
Similar games go on inside high-impedance attenuators that you actuate when
you rotate the Volts/Div selector switch.
Hope this clarifies more than it obscures.
-frank
(former designer for Tek)
Tom Bruhns
: However, I am still confused how does a X1 passive oscilloscope probe
: work.
: The input impedance at the probe tip is 1 M.
This is true at DC; or more correctly, the input impedance of the scope
itself is echoed to the probe tip, because there is only a wire between
them. But at higher frequencies, the impedance drops. Expect that the
cable capacitance will be on the order of 10 to 15 pF per foot, and if the
probe cable is 3 feet long, that might well be 45pF of capacitance. Up to
the point where the cable is a significant fraction of a wavelength long,
you can fairly accurately view it just as a capacitor. You should add the
capacitance of the scope itself in parallel. At 1MHz, for example, the
impedance might be about 3k ohms, capacitive. In fact, this much
capacitance is enough to seriously detune most tuned circuits even at such
low frequencies. x1 "high impedance" passive probes are NOT good,
generally, for RF work. That's why active probes are made, and why fast
scopes are made with 50 ohm inputs: with them, you use 50 ohm cable and
drive it from 50 ohm sources--well, that's the short version of the story
anyway.
At some frequency, you should start considering the probe cable as a
transmission line, but by the time you reach that frequency, you will have
dropped to a low enough impedance that it's probably not worth worrying
about. An estimate of a frequency at which to start worrying about
transmission line effects would be the frequency at which the line is about
20 electrical degrees long. For 3 feet of line with .8 velocity factor,
this would be about 4MHz. At some length, the capacitance and very high
resistance of the scope input will reflect back as nearly a short circuit!
It may well be that there is some inductance or resistance at the probe tip
to reduce this very severe loading problem, at the expense of response at
that frequency.
: Yes, but if the cable is
: 50 Ohm, then there must be some wide band ( DC to some hundreds MHz )
: matching network to match 1 MOhm to 50 Ohm, and then at the same time,
: there must be some wide band DC amplifier in the oscilloscope to
: amplify the signal, of which the gain could be in the range of
: 1,000,000 / 50 = 20,000.
Well, if it worked that way, you would need a voltage gain of only
SQRT(20,000), since the impedance is the voltage/current ratio, and if no
power is lost in a passive impedance conversion, the current goes up as the
voltage goes down. But as explained above, that's NOT the way it works.
: Is it at all possible to design such network and amplifier at such low
: cost ?
Not today, as far as I know. At least not in the volumes used. But the
processing power of a Pentium PC would have cost quite a bit more back in
1950, too.
On 10x probes: Be careful that you do not assume these will have very high
probe tip impedance out to very high frequencies either! Typically a good
first-order model for them is the scope resistance and capacitance in
parallel with the cable capacitance, and that in series with a parallel
resistor (9x the scope input resistance) and capacitor (1/9 of the scope
plus cable capacitance). So the input capacitance at the probe tip is
still going to be 1/10 of the scope+cable capacitance, or around 5pF.
That's a LOT lower than 10 megohms, for any frequencies above audio.
(High speed 10x scope probes are more complicated than this, but it's a
reasonable model to start with.)
--
Cheers,
Tom
to...@lsid.hp.com
Richard A. Schumacher
Take a 50 ohm coax. Put a 1M ohm resistor in series with the signal conductor
at the input end. Suture and close.
John Lundgren
: Thanks for your comment.
: However, I am still confused how does a X1 passive oscilloscope probe
: work.
: The input impedance at the probe tip is 1 M. Yes, but if the cable is
: 50 Ohm, then there must be some wide band ( DC to some hundreds MHz )
: matching network to match 1 MOhm to 50 Ohm, and then at the same time,
: there must be some wide band DC amplifier in the oscilloscope to
: amplify the signal, of which the gain could be in the range of
: 1,000,000 / 50 = 20,000.
: Is it at all possible to design such network and amplifier at such low
: cost ?
: The Hong Kong Crystals Company,
: Technical Director,
: Arthur Lee, artl...@hk.super.net
Just because the cable has an impedance of 50 ohms does not mean that it
has to be terminated at each end with that impedance. The input
impedance of the amp is 1 megohm. There is a resistive and capacitive
divider in the probe tip.
--
#===================================================================#
| John Lundgren - Elec Tech - Info Tech Svcs. | jlundgre@delta1 |
| Rancho Santiago Community College | .deltanet.com |
| 17th St @ Bristol \ Santa Ana, CA 92706 | |
| My opinions are my own, and not my employer's. | |
| "You can flame your brains out -- it won't take long." |
#======P=G=P==k=e=y==a=v=a=i=l=a=b=l=e==u=p=o=n==r=e=q=u=e=s=t======#
John Lundgren
: artl...@HK.Super.Net (Arthur Lee Kam Hung) wrote:
: >I am looking for coaxial cables with 1 M ohm nominal impedance, just
: >like the ones used in oscilloscope probes.
: >
: Sorry, but oscilloscope probe cables do not have 1 megohm nominal
: impedance. The input impedance of general purpose oscilloscopes is
: typically 1 megohm. The cable impedance is in the range 30 - 75 ohms,
: but 10X probes contain an internal RC voltage divider which produces a
: probe tip impedance of 10 megohm. Cables with characteristic impedance
: higher than about 300 ohms are not practical.
: Steve
I have seen scope probe cables that are made of a series of layers. The
first innermost is an insulator. Over this is a coil of wire that is the
"center conductor" so to speak. Then comes the dielectric. Over that is
a shield, and of course the insulating jacket. The coiled center
conductor has a resistance that's a lot higher than 300 ohms.
I'm not sure what the principle behind this is, tho. There is a lot of
attenuation. And there is a lot of expense making a custom cable.
Stan Zyra
(John Lundgren) wrote:
> I have seen scope probe cables that are made of a series of layers. The
> first innermost is an insulator. Over this is a coil of wire that is the
> "center conductor" so to speak. Then comes the dielectric. Over that is
> a shield, and of course the insulating jacket. The coiled center
> conductor has a resistance that's a lot higher than 300 ohms.
>
> I'm not sure what the principle behind this is, tho. There is a lot of
> attenuation. And there is a lot of expense making a custom cable.
You were just describing a delay-cable. Such a cable consists of a
distributed capacity and inductance. The center conductor is a coil acting
as a capacitor at the same time. The impedance which is the squareroot(
inductance/capacity ), can be made quite high. Most known is the RG65/A
cable. It has a nominal impedance of 950 Ohms and the velocity factor is
0.024. The resistance is 23 Ohms per meter.
Stan
Steve Rickman
>I have seen scope probe cables that are made of a series of layers. The
>first innermost is an insulator. Over this is a coil of wire that is the
>"center conductor" so to speak. Then comes the dielectric. Over that is
>a shield, and of course the insulating jacket. The coiled center
>conductor has a resistance that's a lot higher than 300 ohms.
>
>I'm not sure what the principle behind this is, tho. There is a lot of
>attenuation. And there is a lot of expense making a custom cable.
>
>--
You're right. I believe many Tektronix probes have high distributed
resistance in the cable center conductor. They may have a patent on
the process. I was thinking of low-bandwidth commodity probes with
ordinary coax cables.
And in answer to Mr. Lee, if this is what he seeks, then perhaps
Tektronix can help in identifying a supplier. Otherwise, alternative
methods of probing high-frequency signals are to use a hi-Z active
probe, or a low-Z passive probe. A typical 10X low-Z probe has a tip
impedance of 500 ohms and is terminated in the cable characteristic
impedance at the oscilloscope input. Either type can have a bandwidth
of several GHz.
Steve