Hi folks,

The other day, I was trying to troubleshoot a 100MHz amaplifier in a
piece of equipment ( actually, a Cushman CE-3 Spectrum Monitor ). I
was using a 150MHz tektronix scope with a 10X probe.
I was getting different results with each measurement, it was driving
me crazy.

My friend Bob, the RF guru, said "Your trouble is that you're
using the HI-Z input to the scope. For VHF, Hi-Z is crap. It
just doesn't work. Especially if you have a pigtail for the
ground." Guilty as charged...

He then asked if the scope had a 50-ohm input ( which it does ).
"Then just make up a probe with a 450-ohm series resistor, and
some little ground clip. That's all you need. It will be stable, and
it will work up to about a gig. The impedances in
this RF stuff tend to be low anyway, so you don't really need Hi-Z."

Makes sense. Any way to buy a probe like this off he shelf?

- Jerry Kaidor ( )

wrote:
Hi folks,

The other day, I was trying to troubleshoot a 100MHz amaplifier in a
piece of equipment ( actually, a Cushman CE-3 Spectrum Monitor ). I
was using a 150MHz tektronix scope with a 10X probe.
I was getting different results with each measurement, it was driving
me crazy.

My friend Bob, the RF guru, said "Your trouble is that you're
using the HI-Z input to the scope. For VHF, Hi-Z is crap. It
just doesn't work. Especially if you have a pigtail for the
ground." Guilty as charged...

He then asked if the scope had a 50-ohm input ( which it does ).
"Then just make up a probe with a 450-ohm series resistor, and
some little ground clip. That's all you need. It will be stable, and
it will work up to about a gig. The impedances in
this RF stuff tend to be low anyway, so you don't really need Hi-Z."

Makes sense. Any way to buy a probe like this off he shelf?

I'll bet Tektronix and HP/Agilent will sell you one, but you don't want
to know what it costs.
Problem with commercial probes is that they have a rod resistor with a
probe point soldered into the end. Lots of stress, easy to break,
expensive to fix.

The resistor on the end of a coax is a pretty good solution.
Pack it into a ball-point pen. When you break it, replace the resistor.
1/4W carbon composition works well. If you try to use a metal film, you
may have to worry about the inductance. Don't know any way to tell
without measuring it.

If you want accurate/repeatable measurements at 100MHz. you need to
build the measurement point into your circuit. ANYTHING with a ground
clip will require an inch or more of wire just to let the probe reach
the circuit. Ground inductance is your enemy.

Think of it as a series resonant circuit. The tip capacitance and the
ground + tip inductance resonate at some frequency and make a short
at your circuit. Obviously, the accuracy decreases rapidly as you
approach this frequency. Doesn't take much ground wire to mess things up.
mike



- Jerry Kaidor ( )




--
Return address is VALID but some sites block emails
with links. Delete this sig when replying.
..
Wanted, PCMCIA SCSI Card for HP m820 CDRW.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Wanted 12" LCD for Compaq Armada 7770MT.
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
htremovethistp://www.geocities.com/SiliconValley/Monitor/4710/

mike wrote:
wrote:
Hi folks,

...snippety..

Makes sense. Any way to buy a probe like this off he shelf?

I'll bet Tektronix and HP/Agilent will sell you one, but you don't want
to know what it costs.
Problem with commercial probes is that they have a rod resistor with a
probe point soldered into the end. Lots of stress, easy to break,
expensive to fix.

The resistor on the end of a coax is a pretty good solution.
Pack it into a ball-point pen. When you break it, replace the resistor.
1/4W carbon composition works well. If you try to use a metal film, you
may have to worry about the inductance. Don't know any way to tell
without measuring it.

If you want accurate/repeatable measurements at 100MHz. you need to
build the measurement point into your circuit. ANYTHING with a ground
clip will require an inch or more of wire just to let the probe reach
the circuit. Ground inductance is your enemy.

Think of it as a series resonant circuit. The tip capacitance and the
ground + tip inductance resonate at some frequency and make a short
at your circuit. Obviously, the accuracy decreases rapidly as you
approach this frequency. Doesn't take much ground wire to mess things up.
mike
tp://www.geocities.com/SiliconValley/Monitor/4710/

A good hint is to downlod the manual from Tektronix website, called
"ABC of probes". Gives a lot of basic info.

The "DIY" approoach is nice and will work well.

good Tektronix Probes in the low-Z Area are on2nd hand market like
ebay. I am not shure about the correct probe number, AFAIK it is P3056
(10:1 low Z, made in the 1970+). I recommend you look in Tek Catalog
for the description , under "obsolete probes" reference page. Probe is
subminiature and works well, price was about 15 USD. Specified for up
to 3 GHz use, which is more than adeaquate for my Tek 7104/485/2465B
scopes ;-)

hth,
Andreas