How to Build a 1 GHz Scope Probe for Next to Nothing

EDAboard.com Forum Index -> Other Design

Author

Message

hot_chmock

Joined: 22 Jun 2001
Posts: 171
Location: Where it is nice

27 Jul 2002 9:43 How to Build a 1 GHz Scope Probe for Next to Nothing

Instructions

Oscilloscope bandwidth is expensive, don't reduce it further with a slow probe. Expensive fast scope probes have high input impedance, which we do not need when looking at digital IC outputs. Oscilloscopes have more gain than we need for digital signals. So, let's sacrifice input impedance and voltage sensitivity, and save money.
Take a 3-ft piece of RG174 50-Ohm coax cable with a 50-ohm terminating BNC connector on one end, attach a 950 ohm (or 1 kilo ohm) small resistor to the other end and use a very short ground connection. Keep the ground-loop shorter than 0.5 inch.
Rise Time Budget
Ground loop (half-inch diameter) 0.17 ns
3-ft coax 0.05 Ns
BNC connector with 50 Ohm termination 0.01 Ns
Total rise time of 0.2 Ns means a bandwidth of 2 GHz. Input impedance 1 kilo ohm, input capacitance < 2 pF, attenuation 20:1.

hc

Wink

sick_man
Guest

27 Jul 2002 13:08 scope probes

realy lets be honest

we only buy scope probes so you get the accesories like

gnd crock clip insulated probes ... probe clip s ....
three adapters incase you loose bits and spare tips {various types for different jobs}

i pay out about 15 - 25 £ sterling per year
and have done for 15 years

thats lots of bootle....

but a good comprimise

becouse i
1} dont get a dangerious probe }
2} if for instance i measure the wrong point with a coax probe say 200 watt rf coil instead of 13.8 v supply rail
im in deeep shit with coax with a resistor becouse i just damaged the front end amps pumping 200 watts in o dear 200w at 10.1 attenuator {probe coupled} = about 2 watts so down the lossy road = lots less damage and no toasted expensive front end opamps

3} its ok having 2 g bandwidth if you have a 4 - 10 gb scope haha

last one i saw cost as much as a lotus espree {and made you tea on demand and had a 1 billion year no quibble warenty}

4} bandwidth is irrelivent if the 3 db point of the scope dosnt exceed 20 mhz {as per every scope i have ever seen used or own}
so no matter what freq above 20 mhz you look at the carrier of
you always get flat gain tailing of 20 mhz before the muf or double the max freq range of the amplifiers in the front end of the scope

but man something tells me this is a good idea

becouse its cheeper

but a bad idea becouse it is dangerious and not realy needed

a decent probe does exactly as it descibes its self

or buy one that does exacly what you want

i use many probes some made from pens for prescalling my counter and a sliding pot for attenuation

so ... coax is good for me up to 200 mhz

coax is a strange substance

but what i will add is this

any piece of coax has its own resonants so it isnt a good idea
to measure for instance 3 ft wavelengh = 1 meter ~ 300 mhz
so 600 mhz is also a resanant lengh and 1.2 ghz etc
as for the resistor if you cut the legs to 1 mm and body together
will resonate near telephone freq ranges anything from 1 to 10 ghz

this can be a real problem
rf genius spend 10 years designing probes

the point with a probe is it dosnt resonate ....

so carriers {digital or analogue } dont destort as flux of various remixed freq coherse and transverse the workbench.....

a much better solution is to use a data sample and hold
and built the front end into a probe that then eliminates the problem
and allows for scaling to be done in the probe

a tight scallar scalling design
means a 20 mhz lab scope can be as powerfull as a 2 - 20 ghz bandwidth morgage scope

and to the flaw in your thinking of 50 ohms

in idea for the probe above

any ANY and EVERY scope

uses a 75 ohm adapter {although a 50 ohm one will fit {the center pin is a little bigger diameter so the 75 ohm panel socket on your scope gets brocken nice 1 eh ???}

so your 50 ohm one will off by .75 - 1 ratio in impedance at the point of the plug and all the way down the coax
so square waves at 100 mhz will be scewed and destorted

USE 75 OHM coax and plugs

i use this method

a 75 ohm terminator and a three way 75 ohm splitter

i can then plug in a "dummy load" for very strong signals

bleading most via the load

echo47

Joined: 07 Apr 2002
Posts: 4205
Helped: 564

27 Jul 2002 14:14

Yes hot_chmock, excellent suggestion! For many years I have used similar probes with my 500 MHz scope.

I make x100 probes using a carefully shielded 4.99k RN55 resistor at the end of one meter of RG-174. I set my scope input to 50 ohms. For digital troubleshooting, these babies are far superior to my standard Tek P6139A probes. The capacitance is so low that sometimes I don't even need a ground clip!

hot_chmock

Joined: 22 Jun 2001
Posts: 171
Location: Where it is nice

27 Jul 2002 19:37

Simbox,

I have a 500MHz scope but only 250Mhz probes, those things are very expansive.

When I mesure digital signals and want the best mesurement possible my scope provide (500 pico second time base) then there is no risk to put my probe on any other point that the one I really want.

For all other mesurements I use my standard probes.

hc

Wink

sick_man
Guest

27 Jul 2002 20:09 HI

THE NOTE I PUT WASNT MENT TO DETER PEOPLE FROM MAKING A PROBE

LIKE ABOVE
IT WAS MEARLY TO POINT OUT THAT IT IS STILL A DANGERIOUS WAY TO USE A PIECE OF COAX

AND A GOOD POINT IS
A MAN WHO CAN AFFORD A 500 MHZ SCOPE CAN AFFORD A PROBE

DO YOU DRIVE ??? IF SO I THINK YOU MUST BE A FLINTSTONE....

hot_chmock

Joined: 22 Jun 2001
Posts: 171
Location: Where it is nice

27 Jul 2002 20:19

Simbox,

I bough this scope second hand, 2 year old, almost not used with not a scratch and perfect calibration for a fraction of the price.
No probes where supplyed with the scope.

Now I'm a flinstone that can watch with my scope up to 500 Mhz the heart beat from a my dynosore when I push him to pull my car up to 200 km/h!

hc

Wink

sick_man
Guest

27 Jul 2002 20:54

lol....

but he still needs a probe to get him moving and you must be expert by now

House_Cat

Joined: 21 Feb 2002
Posts: 1334
Helped: 274

28 Jul 2002 3:29

There is a web page that talks about do-it-yourself projects for signal integrity measurements. One of the projects is a 1Ghz probe.

The web page is: h**p://emcesd.com/1ghzprob.htm

A copy of the probe construction project in PDF format is attached.

If you go to his home page, there are other low budget projects that are very useful.

JNovak

Joined: 05 Jul 2002
Posts: 19

28 Jul 2002 4:54 re

Well, it may be a good idea with a coax and stuff. However, a probe that simple going to 2 GHz that's really suspicious. And by the way, if you are doing measurements with this kind of bandwidth, then you must be doing something serious and paid, why save on the commercial calibrated probes?

EDAboard.com Forum Index -> Other Design

Page 1 of 1

All times are GMT + 2 Hours

Abuse
Administrator
Moderators

sitemap