I have the Keysight DSOX1102G oscilloscope that includes a function generator, and the oscilloscope part works just fine as far as I can tell. It defaults to the Hi-Z output mode. I've measured things like PWM generated by an Arduino just fine. I am careful that nothing is earth grounded except the scope. (Checked my power supply: Negative is not tied to earth.)

When I measure the generated function with the oscilloscope, I get strange results, with the voltages being too low. The signal shapes and frequencies are fine though: Only the voltages are wrong. Changing to 50ohm mode doesn't do anything measurable in this situation, as I'd expect. Furthermore, if I use a breadboard to connect the generator to the scope it works, but if I connect anything in parallel with the scope (tried a 10k potentiometer all the way at 10k and 0), no matter what, the oscilloscope goes flat.

I even tried the generator in pure DC mode, and it gives me exactly half the voltage: Generator at 5V DC reads as 2.5V horizontal line on the scope, 10V generator reads as 5V.

I hooked up an N-JFET to try to get the generator to drive an LED. External power source -> 1k resistor -> blue LED -> drain -> source -> ground, then function generator positive -> source -> gate -> ground. I tried a 3Hz 5V-0V 50% duty cycle square wave, the jfet never turned off. I put the generator in DC mode, turned it up to 10V, still didn't turn off jfet. If I hook up a 9V battery from source to gate instead, the jfet turns off just fine.

More strange behaviour: If I have the generator at 10V DC hooked up across the jfet and the scope measuring in parallel, with both lines connected, the scope is at 0V; If I remove the positive generator line, it drifts up to 5V over several seconds, and jumps instantly back to 0V when I plug it back in, and when I remove the negative generator line it jumps instantly to 5V then back to 0V when I put it back.

What am I doing wrong? The oscilloscope itself seems to be fine (it's brand new), but I can't seem to get it to drive a jfet or to even properly generate a signal. The signals generated are exactly as they should be except the voltage according to the oscilloscope, but they won't seem to actually do anything in a real circuit and putting them in a real circuit drops the voltage (according to the scope) to zero.

 

It sounds like you may have the scope input set for 50 ohms. This is something normally used for special probes and direct connection with 50 ohm coaxial cable. Most modern higher end scopes have 50 ohm termination internally as a selectable option. Older scopes and lower end scopes don't, so you have to use an in-line terminator (which is a surprisingly expensive item, if it is well made). [EDIT]: scopes with internal termination will show that it has been enabled on the screen, often adjacent to the vertical scale factor.

A signal source that has 50 ohm output normally employs series termination - that is there is a 50 ohm resistor (or slightly less to account for the source impedance of the output amplifier) between the output of the amplifier/driver and the output connector. When used with a cable of matching impedance, the signal will "correct" at the other end of the cable, even if it isn't terminated. I can be horrible at points between the ends, but that doesn't matter. Typically the receiving device will terminate the cable with 50 ohms in parallel. This makes for very good signal integrity, however it also means that the signal amplitude at the receiving end is one half of that at the "inboard" end of the source series termination resistor. Thus, 5 V at the source becomes 2.5 V at the receiving end.

If you have a high impedance source and run it into a 50 ohm termination, it will attenuate it according to the impedance ratio. A normal scope input is 1 megohm (in parallel with a small capacitance).

I haven't thought carefully about each scenario you have described, so I may be missing something.

 

By the way, welcome to AAC. I didn't notice it was your first post.

Spend some time becoming very familiar with your new oscilloscope. Modern scopes are pretty spectacular in what they can do and it takes a good deal of time just to learn about the capabilities. Learning how to use an oscilloscope (or any other instrument) well is key to making measurements that are meaningful, Understanding the limitations of the instrument and how connecting the instrument to a circuit changes the way the circuit behaves is important. Judging by the scope displays I've seen at AAC in the year I've been around, I'd say about 10% of scope users are good at setting their scope well.

The late Jim Williams, most recently with Linear Technology, wrote some of the best applications notes in the industry, in my opinion. His primary oscilloscope was an ancient Tektronix analog scope, but he did many very impressive things with it because he was an absolute top notch expert in using oscilloscopes for the type of work he did. In some of his ap notes he included quite lengthy and detailed sections on how to use oscilloscopes and probes well for the circuits he was discussing. I advise anyone interested in analog electronics to at least skim all of his work. They can be found on the web, both at Analog Devices (which acquired Linear Tech) and elsewhere. Some of them are out of date in terms of the specific products he wrote about, but he covers a lot of fundamental material that is very instructive and broadly applicable.

 

It sounds like you may have the scope input set for 50 ohms. This is something normally used for special probes and direct connection with 50 ohm coaxial cable. Most modern higher end scopes have 50 ohm termination internally as a selectable option. Older scopes and lower end scopes don't, so you have to use an in-line terminator (which is a surprisingly expensive item, if it is well made).

Unfortunately, the Keysight DSOX1102G, which is Keysight's latest entry-level scope, doesn't have switchable 1M/50ohms inputs.

I have the Keysight DSOX1102G oscilloscope that includes a function generator, and the oscilloscope part works just fine as far as I can tell. It defaults to the Hi-Z output mode. I've measured things like PWM generated by an Arduino just fine. I am careful that nothing is earth grounded except the scope. (Checked my power supply: Negative is not tied to earth.)

When I measure the generated function with the oscilloscope, I get strange results, with the voltages being too low. The signal shapes and frequencies are fine though: Only the voltages are wrong. Changing to 50ohm mode doesn't do anything measurable in this situation, as I'd expect. Furthermore, if I use a breadboard to connect the generator to the scope it works, but if I connect anything in parallel with the scope (tried a 10k potentiometer all the way at 10k and 0), no matter what, the oscilloscope goes flat.

The big question is *how* did you perform these measurements? Because it sounds like you're poking with your probe into the BNC output, which isn't a proper probing technique. And it would explain why your voltages are off.

The signal generator output on the DSO-X1102G is terminated with 50 ohms (the 50ohms/hi-z only changes the indicated voltage, it doesn't change the actual impedance). The scope inputs are 1 Mohms. To get proper results, the transmission line (i.e. the connection from generator to scope input) must be correctly terminated. The generator is 50 ohms, so you need to bring the scope inputs to 50 ohms, too, which you do as ebp has described in his post: use a 50ohms inline terminator. The last bit is the line in between. Breadboards and wires are amongst the worst way to transfer HF. What you need is a 50ohms coax cable with BNC connectors on both sides, something like RG-58 which is cheap to buy. Then set the generator to 50 ohms mode, and connect the BNC cable to the generator output and the scope input.

This is a great lesson to demonstrate that probing technique matters - a lot actually, and that passive probes aren't always the right tool for the job.

The late Jim Williams, most recently with Linear Technology, wrote some of the best applications notes in the industry, in my opinion. His primary oscilloscope was an ancient Tektronix analog scope, but he did many very impressive things with it because he was an absolute top notch expert in using oscilloscopes for the type of work he did. In some of his ap notes he included quite lengthy and detailed sections on how to use oscilloscopes and probes well for the circuits he was discussing.

While I agree that Jim Williams wrote some great app notes on (analog) scope usage, a lot of it isn't relevant or even counter-productive on a digital scope. Williams hated digital stuff, and when he wrote his notes technology was notably different than it is today. For a beginner, I think it makes more sense to stick with app notes and tutorials for modern scopes, which can be found with the big manufacturers.

LeCroy has a good Probing Tutorial which is worth reading:
http://cdn.teledynelecroy.com/files/appnotes/lecroy_probing_tutorial_appnote016.pdf
and a 4-part blog series 'Probing Techniques and Tradeoffs':
http://blog.teledynelecroy.com/2017/11/probing-tradeoffs-what-to-use-and-why.html

There's also a presentation with tips for accurate probing from Keysight:
https://www.keysight.com/upload/cmc_upload/All/Probing-6-Hints-Scopes-Dec12-2007-webcast.pdf

 

Hello,

If you want to do 50 Ohms measurements on a scope that can be switched between 50 Ohms and 1 MOhms, you could use a feed through 50 Ohms terminator like the one in the link:
https://www.ebay.com/itm/50-Ohm-Fee...h=item2854f7991f:g:XHYAAOSwKp1arzT1:rk:9:pf:0

Bertus

 

I had not even considered the impedance of the probes, since I was just using the second probe that came with the scope as the output of the function generator. I tried to measure the current, and I was getting barely a trickle, and then I switched the attenuation of the probe to 1x (which yes, because I am so new to this, I forgot was a feature) with the button on the probe, and then I started getting measurable current out of the probe. By my calculations, if the function generator has exactly 50 ohms impedance internally, the probe in 1x mode has 517 ohms.

Thank you for your excellent insights and quick responses. Clearly I have a lot more to learn, and I need to get a different kind of cable to connect to the function generator's output as well.

 

Thank you for your excellent insights and quick responses. Clearly I have a lot more to learn, and I need to get a different kind of cable to connect to the function generator's output as well.

A nice thing to have is a coax to banana plugs cable, plus some aligator clips or hooks with that.