Oscilloskope 50Ohm Input

[tautech]

I cannot see on the first (2nd, 3th..) glance what is the frequency scale, however. Would be good to add it such we, novices, can get the information.

With the right menu it's very clear what the Hz/div setting is yes this is always visible in a proper spectrum analyser.

The real art to is to have the appropriate menu showing when doing FFT screenshots with a DSO. 

[Apochrom]

PS: I've spent 5 minutes staring at the picture n. SDS00014.png trying to get an idea where the large peaks are positioned in the spectra, with little success..
@Siglent: could you provide a hint how to get the large peak freqs from the data shown on the display?

hello imo,

I don't know if I understood you right. The Mhz scale is always a bit dark, maybe you have to set your monitor a bit brighter. On picture 14 the center is 98.2Mhz and the right edge 108.2Mhz, the left edge 88.2Mhz. So one box is 2Mhz. I chose this because I was able to display the whole FM Radio Freuquenzband in Germany.

regards, Jürgen

[iMo]

The spectrum analyzer with the Siglents is a nice feature (and a nice working one).
Adding 2 cursors with an information on the freqs, amplitude and deltas would make a great tool from it. No need for an expensive SA then

[tautech]

PS: I've spent 5 minutes staring at the picture n. SDS00014.png trying to get an idea where the large peaks are positioned in the spectra, with little success..
@Siglent: could you provide a hint how to get the large peak freqs from the data shown on the display?

hello imo,

I don't know if I understood you right. The Mhz scale is always a bit dark, maybe you have to set your monitor a bit brighter. On picture 14 the center is 98.2Mhz and the right edge 108.2Mhz, the left edge 88.2Mhz. So one box is 2Mhz. I chose this because I was able to display the whole FM Radio Freuquenzband in Germany.

regards, Jürgen

Display/P2/Intensity and Graticule brightness. 

[Apochrom]

The real art to is to have the appropriate menu showing when doing FFT screenshots with a DSO. 

 

Okay, picture 7 might be better... ;-)
The middle is exactly at 100Mhz and the pitch is 2Mhz/box.

regards, Jürgen

[iMo]

@Juergen - sorry - my bad, I've started to play with my display's viewing angles and 've taken my glasses too - now I can see there is freq scale on the TOP.
Thanks!

[iMo]

The spectrum analyzer with the Siglents is a nice feature (and a nice working one).
Adding 2 cursors with an information on the freqs, amplitude and deltas would make a great tool from it. No need for an expensive SA then

Again, that would be a great feature.. (if not implemented already)

[tautech]

@Juergen - sorry - my bad, I've started to play with my display's viewing angles and 've taken my glasses too - now I can see there is freq scale on the TOP.
Thanks!

The graticules and scales are always darker in Siglent screenshots than on the live display.
A moment in these menus can fix that:

Display/P2/Intensity and Graticule brightness.

[Apochrom]

@Juergen - sorry - my bad, I've started to play with my display's viewing angles and 've taken my glasses too - now I can see there is freq scale on the TOP.
Thanks!

The graticules and scales are always darker in Siglent screenshots than on the live display.
A moment in these menus can fix that:

Display/P2/Intensity and Graticule brightness.

I've toyed a lot with the brightness settings, but they don't do much. And with the raster setting nothing at all...
Here are two examples with 1% raster and 99% raster in the appendix. The brightness of the screens only increases at the edge, neither inside nor at the inscription.
I like the default color scheme of Siglent, but I didn't manage to change anything, the colors seem to be fixed?

regards

[iMo]

Back to your original question on 50ohm termination at the oscope's input.
Below are 3 simulations with the LTspice.

The cable is 1.7meters of RG58 50ohm coax.

The first picture is 1V square wave of 10-300ohm source impedance (R+j0) into the 1Meg||22pF oscope input.

The second picture is 1V square wave of 10-300ohm source impedance (R+j0) into 1Meg||22pF oscope input in parallel with a 50ohm termination.

The third picture is 1V square wave of 50ohm source impedance (50+j0) into 50ohm termination.

[rhb]

What's the frequency of the square wave?  The figures I showed are measurements with a 100 KHz square wave.  It would be very helpful if you did a simulation to match my measurements.  Changing the square wave frequency is not possible.

Also can you do anything about the display?  Blue on black is especially hard to see.

[iMo]

The frequency was 50MHz, 20ns period, with 1ns edges.
I will do with different colors.

[rhb]

You need to use a much lower frequency, <= 1 MHz to see what is going on.  I'm going to get a custom 1 Mhz pulser from Leo Bodnar for exactly this reason.  Even 10 MHz is too fast.

[rhb]

FYI  These just popped up in an Amazon ad

https://www.amazon.com/Copper-Adapter-38-5mm-Through-Terminator/dp/B07G566JC7

[Apochrom]

Thank you very much for the link. Unfortunately they do not deliver to Germany.

I had now ordered these two from my favourite dealer. (Straight adapters are currently not available in Germany...     ):

https://www.reichelt.de/adapter-bnc-bnc-90-50-ohm-rnd-205-00479-p194674.html?
https://www.reichelt.de/adapter-bnc-bnc-y-stueck-90-50-ohm-rnd-205-00482-p194677.html?

Are those the kind you recommended?
Both are indicated with 50ohm. In addition I still have a totally stupid beginner question.
If I understood it correctly, the 50ohm only appear at high frequencies as impedance. If I measure the resistance normally with a Mulitmeter the Mulitmeter shows 0 ohm. Since the resistance of 50ohm only occurs at high frequencies, I can't measure it with a Mulitmeter, can I? How do I measure the impedance and at which frequency?

regards, Jürgen

[rhb]

Those are just adapters, not thru terminations.  But essential items to have.  I don't have a tee of that style.  It looks as if it could be handy.

With regard to measuring the impedance, stay tuned.  I'm trying to work out a cheap way to do that by TDR.

The impedance of a line in the limit as the frequency goes to infinity is sqrt(L/C) where L and C are the inductance and capacitance per unit length.

[Apochrom]

Thank you very much,


what is the meaning of "TDR"?

[tautech]

Thank you very much,


what is the meaning of "TDR"?

Time Domain Reflectometry

https://en.wikipedia.org/wiki/Time-domain_reflectometry

[iMo]

If I understood it correctly, the 50ohm only appear at high frequencies as impedance. If I measure the resistance normally with a Mulitmeter the Mulitmeter shows 0 ohm. Since the resistance of 50ohm only occurs at high frequencies, I can't measure it with a Mulitmeter, can I? How do I measure the impedance and at which frequency?

regards, Jürgen

The 50ohm (50+j0) appears at all frequencies, it is frequency independent.
Your multimeter shall show 50ohm if the impedance is resistive (thus your termination will show 50ohm when measuring the mid pin against the gnd).

Generally an "impedance" is a "complex value" R+jX or R-jX, based on the sign it could be capacitive or inductive (X is the reactance).

When there is the reactance X the impedance is frequency dependent.

For example your antenna impedance could be 180-j200 at 90MHz, or 340+j12 at 105MHz, etc.


Your 50ohm BNC termination is theoretically 50+j0 (pure resistive, not dependent on frequency), in reality there is a capacitive or inductive portion of reactance based on parasitic capacitancies and/or inductancies. You may see the influence of the parasitic reactancies usually at higher frequencies, based on the manufacturing quality of the termination.

Measuring impedance is not an easy exercise for a beginner, unless you have got the equipment and you know well what you are doing.

PS: https://en.wikipedia.org/wiki/Electrical_impedance

[iMo]

Imagine an ideal signal Generator (50+j0 output impedance), lossy 2m long RG58 coax, and a cheapo 50ohm BNC Termination (an example only) put on Oscope's input.
Below a simulation how the Generator "sees" the Oscope's impedance (the Generator "looks into the input connector" of the RG58 cable).
The impedance is in "polar form" (impedance magnitude and phase, could be converted to R+jX at given frequency).

[rhb]

Here is an example of measuring impedance by TDR. 

This is just an initial installment.  I will have a *lot* more to say on the subject  once I have written the software to do vector network analysis (VNA) by means of time domain reflectometery (TDR).  I just ordered a couple of square wave pulsers from Leo Bodnar with 1 MHz instead of 10 MHz outputs so I can move the reflection event away from the ringing of the DSO AFE.

The setup I'm using here is a square wave feeding a tee at the scope input with a ~35 cm SMA-M to BNC-F cable.

167 shows the reflection with a 50 ohm terminator on the BNC-F connection

168 shows the same setup with a 75 ohm terminator

169 shows a 75 ohm cable in place of the terminator

170 shows a 50 ohm cable in place of the terminator.

Neither cable is terminated, but you can't see that as it is off the screen.

171 shows the full step with the 75 ohm terminator  The initial step, measured using the cursors is 168 mV.  The 2nd step is 18 mV.  That *appears* to be a reflection coefficient of |0.11| however, there is a factor of ~2 error as it is |75-50|/(50+75) = 0.20.  This is almost certainly because I am using a tee rather than a two port splitter/combiner and the cable  is not exactly the specified impedance.  According to my 34401A using a banana plug to BNC-F, the 75 is 74.998 ohms. And the 11801 really isn't accurate enough to give  2 digits of accuracy even if I adjust the step to full scale.  If I assume a correction factor of 2 then the measurements give the impedance of the terminator as 77 ohms

The main point here is that TDR can give you a means of measuring impedance using a DSO.  Accurate results require transferring the data to a PC and doing a bunch of windowing, FFTs,  phase shifts and summing.  But a 100 dB dynamic range using an 8 bit DSO is not an unreasonable expectation.

[iMo]

I do assume your 35cm long coax is the RG174. What is your LeoB_pulser's amplitude (mine is 1V)?
My first step is 3.6ns, btw.

1st step 334mV
2nd step 378mV
diff      44mV

[rhb]

I'm using the <35 ps rise time 250 mV 100 KHz square wave calibrator output on the 11801.  I didn't measure the rise time, but it's showing at the moment around  20 ps with the scope not warmed up.  Generally I get a measurement of around 25-26 ps with it warmed up.  The heads are rated at about 13 ps.  So obviously the calibrator is faster than spec by a good bit.

It took some looking, but it's Rohm RG174.  The line from the calibrator to the tee is ~ 30 cm of RG402.

[iMo]

I cannot find the lossy model for the RG402, using params of RG174. That is not good as the 402 is pretty different beast, but ok for this measurement.
Below with 250mV generator/calibrator pulse amplitude. Also I've changed your oscope imped to 1Meg.
Update: your sampling's head is 50ohm.

First step is 124.9mV, the second 150mV.
First step is 83.34mV, the second 94.48mV.

PS: added sim with RG402's lossless td=4.8ns/m. Almost the same result (diffs of 0.1mV).

[rhb]

The SD-26 is a 20 GHz, 13 ps rise time  50 ohm input sampling head.  It *only* has 50 ohm inputs.  The 11801 mainframe has a 50 GHz BW using an SD-32 7 ps rise time sampling head.  This is *not* a DSO.  Sampling scopes are completely different instruments.

One can do similar measurements to the BW of a DSO with one of Leo's square wave generators, though obviously not to the extreme which the 20 GHz SD-26 will reach.  The calibrator step is actually about 17 ps.

The reflection from the terminator sees 25 ohms at the tee from the parallel combination of the 50 ohm sampling head input and the 50 ohm calibrator output.  I've not done the algebra, but I'm pretty sure that is the source of the 2x factor.

I'm using the 11801 because it happens to be sitting in the middle of my bench at the moment.  Getting to another scope is a tad difficult as it's a very small bench area.