Test Equipment: Oscillocopes

[MarkE]

This thread is to discuss oscilloscopes and good measurement practices using oscilloscopes.

[Pirate88179]

Good topic Mark.  This will help me a lot I am sure, as well as many others.

Bill

PS  http://www.ebay.com/itm/New-Function-signal-generator-source-frequency-signal-generator-finished-board/131063212194?_trksid=p2054897.c100204.m3164&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D1%26asc%3D20140407115239%26meid%3
Ddee726ea5268418aaae964634ecd71a2%26pid%3D100204%26rk%3D7%26rkt%3D28%26mehot%3Dpp%26sd%3D361077176699

Is this FG worth getting if you don't have any as of yet?  (In other words, better than nothing?)

[John.K1]

Now we need some tutor

[MarkE]

The basics of your average passive oscilloscope probe:

Most passive probes come in x1/x10 switchable configurations.  The oscilloscope itself contains the amplifiers which  have a 1 megOhm resistance from the channel input to ground.  When set for x1, the probe is basically a coaxial cable connection from the probe to the oscilloscope.  The coaxial cable for a typical 1.5 meter scope cable has a total of about 50 - 100pF capacitance.  That is a parasitic that the circuit under test has to drive.

When the probe is in the x10 mode a 9Meg Ohm resistor gets switched in at the probe.  That's the good news.  The bad news is that 9Meg Ohms forms a low pass filter with the cable capacitance that kills the frequency response.  In order to restore the response:  the probe includes a small capacitance of 10 - 20pF in parallel with the 9 Meg Ohm resistor.  The circuit being measured now drives 10-20pF in parallel with 9Meg Ohms instead of driving 50 - 100pF in parallel with 1 Meg.  10-20pF may not seem like a lot, but that depends on how fast the signal is.  At 1MHz 16pF is ~10K Ohms, a far cry from the total 10M Ohms of the probe at DC.

The next problem is signal reflections.  Anytime a signal travels through a transmission like, such as the coaxial cable of an oscilloscope probe, whereever the impedance changes there is a reflection:  A ghost signal that travels in all directions from the point where the impedance changes.  When the cable is electrically short compared to the frequency content of the reflection, the reflection is not noticeable.  Where the cable is longer than about 1/10th the rise time of the signal, or about 1/3 the frequency of a sine wave, the reflections become a big issue.  Remember that the oscilloscope input is 1 MegOhm.  A typical coax probe cable has a transmission line impedance near 100 Ohms.  This is a very bad mismatch.  Without adjustment, the signal that hits the amplifier nearly doubles due to the reflection.  In order to get around that, the probe has a series resistor past the 10-20pF compensation capacitor that matches the transmission line (coax cable) impedance to try and cut the signal by half so that when the signal doubles at the scope input the result is correct when it first hits.  The ghost signal still needs to be absorbed by the time it gets back to the probe.  The series resistor would do that if the DUT circuit had zero impedance.  That means the DUT circuit sees that reflected ghost energy.  But it sees it through the probe.  When the probe is connected using a 4" or 6" ground clip, that introduces inductance that resists passing the fast changing waveform of the ghost reflection.  As a result, the ghost reflection is not completely absorbed and now what remains of it goes zooming back down the cable towards the oscilloscope where it will be seen as signal.

These reflections that are made much worse by the ground clip generates ringing that may not settle for 50ns or longer.  This makes it hard to measure narrow spikes accurately, especially using the oscilloscope probe ground clip, and where the circuit under test has a high impedance.  There are vaious techniques to mitigate these problems.

[MarkE]

Good topic Mark.  This will help me a lot I am sure, as well as many others.

Bill

PS  http://www.ebay.com/itm/New-Function-signal-generator-source-frequency-signal-generator-finished-board/131063212194?_trksid=p2054897.c100204.m3164&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D1%26asc%3D20140407115239%26meid%3Ddee726ea5268418aaae964634ecd71a2%26pid%3D100204%26rk%3D7%26rkt%3D28%26mehot%3Dpp%26sd%3D361077176699

Well it's better than nothing.  What works pretty well are the 3D25 USB generators from Hantek.  They go for about $160. and produce decent signals from 1Hz to ~10MHz ( they say 20MHz ) and are easy to control.  The signal output is limited to +/-3V into 50 Ohms which is so-so.  but one can add a voltage amplifier behind them if one is so inclined.  The next step up is to purchase a standaolne FG.  New, a decent digial one starts at about $350.

Is this FG worth getting if you don't have any as of yet?  (In other words, better than nothing?)