Signal generator to oscilloscope connection advice.

[Rene]

I just got my new Siglent SDG1025 signal generator and I am eager to plug the generator to my oscilloscope and checkout some of those sexy waves. The problem right now is that I am new to all these and I am not 100% how to make the generator / oscilloscope connection (I am particularly troubled by the 50 ohms issue).

My guess is that the best and most practical way would be to connect the two devices together using a BNC cable with a 50 ohms terminator (close to the oscilloscope connector side) and do a direct BNC connection.

Another way I have seen this done is to first connect a BNC cable with alligator clips to the signal generator. Once that is done I can use the regular oscilloscope proves and connect the proves to the alligator clips. The problem here is that I am not sure if I have to add a 50 ohms resistor between the alligator clips.

Could someone help me sort this issue out?

Thanks.

[alex.forencich]

The generator will work fine without a 50 ohm load.  However, it is possible that the 'calibrated' output voltage will be specified for a 50 ohm load, so if you set it to 1 Vpp you will see 2 Vpp on the scope.  Usually there is a setting to change this so that the output voltage is specified for a 1 M ohm load so the voltage will be correct.  You won't damage a waveform generator like that by running it without a load. 

[Rene]

Thanks Alex,

The one thing I still don't understand is the 50 ohms load thing. For example, say that I built a circuit and the load on the circuit ends up being 75.7 ohms (or some other weird value). How would that affect the signal generator when I connect it to the circuit?

Why can't the signal generator just send a voltage like a DC power supply does? I have never seen a "50 ohms load" specified on a power supply. Why is this only an issue to a signal generator and not to a DC power supply?

Thanks.

[robrenz]

This post and the whole thread may shed some light.

[alex.forencich]

Thanks Alex,

The one thing I still don't understand is the 50 ohms load thing. For example, say that I built a circuit and the load on the circuit ends up being 75.7 ohms (or some other weird value). How would that affect the signal generator when I connect it to the circuit?

Why can't the signal generator just send a voltage like a DC power supply does? I have never seen a "50 ohms load" specified on a power supply. Why is this only an issue to a signal generator and not to a DC power supply?

Thanks.

The generator has a 50 ohm resistor in series with the output connector.  The idea is to make a controlled impedance environment where you won't have reflections going everywhere.  This becomes more important with shorter wavelengths (higher frequencies).  Power supplies do not do this because you would get a huge voltage drop with no real benefit. 

Basically, the output impedance of the generator forms a voltage divider with the input impedance of your circuit.  If both are 50 ohms, the voltage from the generator before the resistor is simply divided by 2.  However, if the load is 75 ohms, then the voltage on the load will be 75/(50+75) = 0.6 times the pre-resistor generator voltage. 

[LukeW]

You can use a BNC-BNC coaxial cable, a male BNC T-piece on the oscilloscope side connected to the scope, and a terminator on the other side of the T-piece.

(Sometimes I get a bit annoyed when I want a 50 ohm BNC patch lead, a tee or a terminator and I can't find one, but years ago when everyone replaced their 10-base coax LANs there must have been millions of the thing tossed in the trash.)

[Matje]

You can use a BNC-BNC coaxial cable, a male BNC T-piece on the oscilloscope side connected to the scope, and a terminator on the other side of the T-piece.

(Sometimes I get a bit annoyed when I want a 50 ohm BNC patch lead, a tee or a terminator and I can't find one, but years ago when everyone replaced their 10-base coax LANs there must have been millions of the thing tossed in the trash.)

Beware!

I think you still can get terminators and stuff for 10Base2 Ethernet (aka Thinnet, the thing using RG58 coax). These are definitely no good for higher frequencies, it was just not needed and so was not tested/specced.

You really don't want to use these (quite cheap) terminators, tees and so on for measurement purposes, the potential trouble isn't worth it.

[Rene]

Thanks guys, I really appreciate your help.

I am going to do a little some more research in this area but in the mean time, I have another potentially dumb question. Why 50 ohms for the signal generator output? Why not 25 ohms or 9.45 ohms or some other random value?

Thanks.

[Alex Eisenhut]

Thanks guys, I really appreciate your help.

I am going to do a little some more research in this area but in the mean time, I have another potentially dumb question. Why 50 ohms for the signal generator output? Why not 25 ohms or 9.45 ohms or some other random value?

Thanks.

http://www.microwaves101.com/encyclopedias/why-fifty-ohms

[w2aew]

You may find this video helpful regarding the output amplitude vs. the load impedance...

[JuiceKing]

For another explanation of what's going on, here's an oldie but a goodie:

[havefun]

You may find this video helpful regarding the output amplitude vs. the load impedance...

Hi, w2aew, thank you for your videos.   I have a question, why the amplitude of voltage will be double when I change the impedence on the generator from 50? to HighZ?   Waiting for your answer, thanks.

I have figure it out,  this is because the output impedance of the generator is always 50 ohm(both in the HighZ and 50 ohm mode) , so the voltage displayed on the FG's screen will double to ensure that the set voltage is the same as the measured voltage.

Anyway, thanks all of your videos put on youtube.

[David Hess]

The generator has a 50 ohm resistor in series with the output connector.

The generator's source termination does not have to be in series although this is the most common configuration.  Some designs use a parallel termination driven by a current output stage.  Both result in a controlled source impedance and with no load termination produce twice as high an output voltage.

I think you still can get terminators and stuff for 10Base2 Ethernet (aka Thinnet, the thing using RG58 coax). These are definitely no good for higher frequencies, it was just not needed and so was not tested/specced.

You really don't want to use these (quite cheap) terminators, tees and so on for measurement purposes, the potential trouble isn't worth it.

A good terminator used with a BNC-T in place of a feedthrough termination is good up to at least 100 MHz.  I do not have any ethernet terminators handy to test.

[w2aew]

For another explanation of what's going on, here's an oldie but a goodie:

This is my absolute favorite video for explaining wave propagation and reflections - totally brilliant!

[Marco]

A good terminator used with a BNC-T in place of a feedthrough termination is good up to at least 100 MHz.

And you're going to be down ~3dB at 100 MHz regardless of quality.

[macboy]

For another explanation of what's going on, here's an oldie but a goodie:

This is my absolute favorite video for explaining wave propagation and reflections - totally brilliant!

Imagine how much you could have learned if every one (or any single one!) of your university professors or high school teachers could convey concepts as effectively as Dr. Shive.  He is truly genius, not because he understands the concepts so well, but because he can help others understand them well despite his own excellent understanding. Only very talented teachers can do that.

[David Hess]

A good terminator used with a BNC-T in place of a feedthrough termination is good up to at least 100 MHz.

And you're going to be down ~3dB at 100 MHz regardless of quality.

Under what conditions?  Certainly that would be the case with a 100 MHz oscilloscope or other test instrument with a 100 MHz input bandwidth.

[Marco]

Meant from reflection, but I didn't think that through ... lots of energy reflected, but it's mostly just seen in phase shift at the scope.

[David Hess]

Meant from reflection, but I didn't think that through ... lots of energy reflected, but it's mostly just seen in phase shift at the scope.

Sometimes you can take advantage of this.  Some signal sources, like Tektronix vertical signal outputs which are 950 ohms, have a high source impedance so if you use a cable terminated into 50 ohms, you get full bandwidth with a reduced scale factor but if you use no termination, the capacitance of the coaxial cable acts as a low pass filter which will limit noise like with a x1 passive probe.

[Rene]

You may find this video helpful regarding the output amplitude vs. the load impedance...

Thanks. I really like your videos, so much so that you are a big part of the reason why I ended up purchasing my new signal generator! For a long time I have been wanting to reproduce the experiment on some your videos and now I am able to do it.... well, many time I can't reproduce them exactly (still trying to figure out why) but close enough!

For another explanation of what's going on, here's an oldie but a goodie:

This is a great find. Thanks for pointing it out!

[dwonga]

Thank you for publishing this very very good video explaining impedance. I felt amazed about watching that old tape and brilliant explanation as well.