Hi All, I'm hoping someone can point me in the right direction troubleshooting a high amplitude issue on my new, unmodified SDG2042x.
I've set the generator to 1khz sine wave, 0.0 dBm. On my oscilloscope, I'm reading roughly 1.26Vpp, which translates to 6 dBm @ 50 ohms. That number is backed up by my Keithley 2000, which is reading right around 6 dBm @ 50 ohms reference.
If I go to the "HighLevel" menu, there is a value of 316.3 mV, and if I zero that, my meter goes to 0.0 dBm, but switching back to "Amplitude" shows the generator also entered a value of -6.021 dBm into the "Amplitude", with an "Offset" of -158.1 mV.
What am I missing here? Why is there always a 6.0 dBm offset being added to my output at all times? Am I missing a parameter somewhere?
Photo #3, hit the attachment limit on my first post.
Output based on 50 ohm load?
Output based on 50 ohm load?
Yes, the SDG2042X will not allow you to set output by dBm until an output impedance is specified.
Is the scope giving a 50 ohm load?
Scope is 1M input, but all I'm measuring there is voltage, which reads about 1.26 Vpp. This correlates to 6.0 dBm based on a 50 ohm output from the generator.
I'm not measuring on the Keithley and oscilloscope at the same time, so no chance the scope is changing the load. Keithley is set to 50 ohm reference.
The AWG "assumes" the load is an ideal 50 ohms when in dBm amplitude mode. As such the AWG produces an "open terminal voltage" equivalent to twice the load voltage when the load is an ideal 50 ohms as assumed, and the AWG always presents an internal 50 effective source impedance regardless of output settings.
So when one specifies 0dBm for example for a sinewave the output voltage will measure as 2*[223.6mVrms] or 2*[316.2mVp] or 1.265Vpp with an open or high impedance load. Of course with a proper load it will measure as 223.6mVrms (316.2mVp, 632.5mVpp) as expected.
Edit: Does the Keithley actually present a 50 ohm input impedance, or does it "assume" a 50 ohm source and exhibit a high input impedance?
Best
Thank you, that is very informative, and the math checks out to what I was seeing in my generator.
On the Keithley, when in ACV, you select the "dBm" math function and the meter asks for a reference impedance. I'm assuming the meter still presents a high impedance load, as most meters do, and would calculate the offset.
I have a 30dB attenuator arriving this Friday. Once that arrives I'll get brave enough to plug it into my spectrum analyzer and see what I read there, as it's a 50 ohm load.
Thank you, that is very informative, and the math checks out to what I was seeing in my generator.
On the Keithley, when in ACV, you select the "dBm" math function and the meter asks for a reference impedance. I'm assuming the meter still presents a high impedance load, as most meters do, and would calculate the offset.
I have a 30dB attenuator arriving this Friday. Once that arrives I'll get brave enough to plug it into my spectrum analyzer and see what I read there, as it's a 50 ohm load.
I don't know if it works in that mode, but there is a Hi-Z option on the output of the SDG.
Unfortunately, the option to select output in dBm is not available unless the output impedance is specified, which I guess makes mathematical sense.
Hi All, I'm hoping someone can point me in the right direction troubleshooting a high amplitude issue on my new, unmodified SDG2042x.
I've set the generator to 1khz sine wave, 0.0 dBm. On my oscilloscope, I'm reading roughly 1.26Vpp, which translates to 6 dBm @ 50 ohms. That number is backed up by my Keithley 2000, which is reading right around 6 dBm @ 50 ohms reference.
If I go to the "HighLevel" menu, there is a value of 316.3 mV, and if I zero that, my meter goes to 0.0 dBm, but switching back to "Amplitude" shows the generator also entered a value of -6.021 dBm into the "Amplitude", with an "Offset" of -158.1 mV.
What am I missing here? Why is there always a 6.0 dBm offset being added to my output at all times? Am I missing a parameter somewhere?
Yup.
In 50 Ohm output mode
any waveform generator expects a 50 Ohm balanced transmission line, 50 Ohm coax
and a some sort of 50 Ohm feedthrough termination.
Ideally a 50 Ohm scope input or a 50 Ohm feedthrough however a BNC Tee with a 50 Ohm termination (load) can be used as a makeshift feedthrough termination.
Thank you! I need a few adapters to properly splice in a dummy load, but I got creative last night and was able to cobble together a stack of banana adapters to put the 50 ohm load from my Nano VNA into the circuit, and, viola!
It's not a perfect 0 dBm but then again, nothing I am using is calibrated, so this is to be expected. It's way close enough for the hack work that I'm doing. Also, it proves the theory, and the more I think about it, it makes sense that the load would have to be actual, and not calculated. Learning every day!
Hi All, I'm hoping someone can point me in the right direction troubleshooting a high amplitude issue on my new, unmodified SDG2042x.
I've set the generator to 1khz sine wave, 0.0 dBm. On my oscilloscope, I'm reading roughly 1.26Vpp, which translates to 6 dBm @ 50 ohms. That number is backed up by my Keithley 2000, which is reading right around 6 dBm @ 50 ohms reference.
If I go to the "HighLevel" menu, there is a value of 316.3 mV, and if I zero that, my meter goes to 0.0 dBm, but switching back to "Amplitude" shows the generator also entered a value of -6.021 dBm into the "Amplitude", with an "Offset" of -158.1 mV.
What am I missing here? Why is there always a 6.0 dBm offset being added to my output at all times? Am I missing a parameter somewhere?
Yup.
In 50 Ohm output mode any waveform generator expects a 50 Ohm balanced transmission line, 50 Ohm coax and a some sort of 50 Ohm feedthrough termination.
Ideally a 50 Ohm scope input or a 50 Ohm feedthrough however a BNC Tee with a 50 Ohm termination (load) can be used as a makeshift feedthrough termination.
50 Ohm load is not mandatory.
You can set the function generator output to Hi-Z or set the reference impedance to 50 Ohms on your multimeter:
The whole reason why specifying impedance is important is because dBm stands for "dB miliwatt"
Both the AWG and DMM operate using voltage and so they must do math conversions to translate between volts and watts. So in order to do that the impedance must be known. The other thing is that AWGs always have 50 Ohm output impedance even if you you are not turning on 50 Ohm mode. So when presented with 10Meg impedance they assume the voltage drop over its own 50 Ohm impedance is zero and so show a 2x larger voltage. If you want a signal source that looks like 0 Ohm impedance you typically need to add a high power buffer amplifier on the output.
In general dBm only makes sense when used in the context of RF circuits where everything has matched impedance (so it is known and constant). This is why all RF gear (synthesizers,spectrum analyzers, network analyzers..etc) works by default in dBm. When you try using dBm for anything else it just makes life harder and introduces extra math for conversions, hence why most DMMs don't even have the ability to display dBm.
If you really want to use dB while operating with voltage you actually want to be using dBµ that stands for "dB microvolt" a voltmeter can measure that without needing to worry about impedance.
50 Ohm load is not mandatory.
You can set the function generator output to Hi-Z or set the reference impedance to 50 Ohms on your multimeter:
(Attachment Link)
If you set AWG to HI-Z and load is 10MΩ multimeter than voltage on meter and AWG will mostly agree. Multimeter will have large input capacitance, and as your frequency goes up there might be errors, provided your meter can measure high enough.
But if you have multimeter with dBm function, if set your multimeter to dBm and 50Ω reference, and connect meter to AWG without 50 Ω terminator, you will have wrong measurement compared to what was set on AWG. Meter can measure dBm/50Ω but has no terminator inside. You still need external 50Ω terminator.
The whole reason why specifying impedance is important is because dBm stands for "dB miliwatt"
Both the AWG and DMM operate using voltage and so they must do math conversions to translate between volts and watts. So in order to do that the impedance must be known. The other thing is that AWGs always have 50 Ohm output impedance even if you you are not turning on 50 Ohm mode. So when presented with 10Meg impedance they assume the voltage drop over its own 50 Ohm impedance is zero and so show a 2x larger voltage. If you want a signal source that looks like 0 Ohm impedance you typically need to add a high power buffer amplifier on the output.
In general dBm only makes sense when used in the context of RF circuits where everything has matched impedance (so it is known and constant). This is why all RF gear (synthesizers,spectrum analyzers, network analyzers..etc) works by default in dBm. When you try using dBm for anything else it just makes life harder and introduces extra math for conversions, hence why most DMMs don't even have the ability to display dBm.
If you really want to use dB while operating with voltage you actually want to be using dBµ that stands for "dB microvolt" a voltmeter can measure that without needing to worry about impedance.
50 Ohm load is not mandatory.
You can set the function generator output to Hi-Z or set the reference impedance to 50 Ohms on your multimeter:
(Attachment Link)
If you set AWG to HI-Z and load is 10MΩ multimeter than voltage on meter and AWG will mostly agree. Multimeter will have large input capacitance, and as your frequency goes up there might be errors, provided your meter can measure high enough.
But if you have multimeter with dBm function, if set your multimeter to dBm and 50Ω reference, and connect meter to AWG without 50 Ω terminator, you will have wrong measurement compared to what was set on AWG. Meter can measure dBm/50Ω but has no terminator inside. You still need external 50Ω terminator.
Thank you both! My assumption was since the meter should know it's own impedance, if tell the meter it's 50 ohms, it should be able to calculate the difference between it's own impedance, and the desired load impedance. But on further thought, the generator, also not being natively 50 ohms, is doing it's own math, and also expecting to see a 50 ohm load to get where it needs to go. So neither instrument really know what's in the circuit. Now the need for a dummy load makes perfect sense.
Just needed to get it in the right context. Thanks!