Signal generator power level too low

[petert]

Hello,

I made a simple circuit, out of a 1kOhm resistor and a diode in series. Using a 5V power supply with the positive lead on the resistor and the negative/GND on the diode's cathode. A signal generator as power supply / signal source seems to have a very limited power output (see below). So I wonder how it can be used to test circuits, which do not have a pretty high impedance.

Details:
I measure the voltage across the positive and negative power supply rails:

• when the voltage source is a µC-board, I have about 5.056V; voltage across the resistor is 4.403V
• when the voltage source is my signal generator (SDG2042X), I have about 4.786V; voltage across the resistor is 4.135V

When I measure the sig gen output directly at the end of the 50 ohm coax cable, I get the expected 5V (almost perfectly, sometimes 4.999V). But when I hook the end of the coax cable to the breadboard and across the resistor-diode-series-circuit, I only get 4.786V.

I assume this is not due to wrong termination on the breadboard side, since the output of the sig gen is just a DC voltage, no AC. I notice that the voltage is more close to the real value when I use a higher valued resistor.

But given the apparently low power output of the sig gen, it seems I would only be able to test very high impedance circuits properly, which seems quite limiting. Or is there some way around that/am I missing something?

Is there a preferred method to hook a sig gen to a breadboard? Special cables? I noticed using scope probes (on 1x setting) would give even worse results / lower voltage (4.207V across power rails, and 3.563V across the resistor).

[stafil]

The datasheet says "Output current: min: -200mA, max: 200mA", so I think what you describe should work fine. Probably your problem is somewhere else?

[petert]

Here is a picture of the breadboard setup, maybe the coax-to-breadboard-connection is not ideal?

The big probes go to the multimeter, the jumper wires go to the coax, which goes to the sig gen.
(It looks like the resistor is in the wrong place, but that's just because how it is bent.)

[stafil]

I am a complete newbie myself, so probably I am talking rubbish. But the voltage you measure(4.78) sounds around what you would get if you had a 1kOhm in series with 50 Ohm.

Is the output set to 50Ohn or HighZ on the signal generator?

[bdunham7]

The sig gen has an output impedance of 50 ohms.  It is equivalent to (and actually is, really) a voltage source in series with a 50 ohm resistor.  Your results are pretty much exactly what you would expect.  And changing the output from 50R to HiZ on that model does not change the actual output in any way, it just causes the display to read 1/2 as high in the 50R mode.

[petert]

Thanks, that indeed explains most of the results.

As summary:

When adding a 50Ohm termination, i.e., a 50Ohm resistor (or rather a combination of resistors, since there is no 50Ohm one) in series with the coax output, then in series with ground, it works. Measuring across that 50ohm termination yields 4.955V. Still a slight variation, but pretty close.

And indeed, the signal generator outputs always the same voltage, but when adding a 50ohm termination (instead of a very high impedance), displaying the output signal as meant for a 50ohm load, is more practical.

So essentially you form a voltage divider with the signal generator and the 50ohm load.

When the load has a very high impedance, you still have a voltage divider, but the the signal generator's 50Ohms become negligible compared to the very high impedance, and the output is almost unaffected. There, HighZ is the right choice, as most of the voltage will drop across the high impedance load, and be very close to that of the voltage output by the signal generator.

Normal circuit rules apply: if there is another circuit in parallel/series to this 50ohm termination, the overall load may not be 50ohms anymore.

So every load circuit, DC or AC, has to be 50ohm overall, such that this voltage divider keeps working as expected. If the load circuit does not already have an impedance of 50 ohm, you need to add impedances in parallel or series to adjust.

Results depend on the cables:
Scope probes are not suitable for hooking up signals from the signal generator to a breadboard, even for DC voltage (as opposed to what was suggested in another thread, when asking about how to connect to a signal generator). Even in 1X setting! I measured mine, for the Rigol DS1054Z, and they have a resistance of 173-180 Ohms on the signal/center wire, and 0.5-1 Ohms on the Gnd wire.

A coax cable with an adapter plug and jumper wires works, but adds about another 2 Ohms. That could explain the slight variations, but still gives the best results.

Probably a coax with integrated crocodile clips or other types of clips works better. Any suggestions are welcome.

[CDaniel]

Every signal generator ( or 99%  ) have an output attenuator ( resistors ) after the final stage  amplifier , this explains why it is only calibrated to show correctly the voltage for an 50ohm resistor load . For other impedances you have to measure the voltage yourself.