Qa470 input impedance

I have a QA401 and QA470 connected with 30 dB amplifier. I would like to know why in the specs of QA470 the input coupling is AC and Fc= 2 Hz for the 30dB and 10dB amplifiers.
Making some test about voltage references and RC Filters the voltage change when I connect the input of + 30 dB amplifier. I can see that the coupling is DC and not AC because with a multimeter I measure 100 Kohm all the time in every amplifiers inputs.
This condition affects all the measures for cause of the loading the circuit with 100 Kohm and changing the workind DC point of the device under test creating and unexpected behaviour and of course a bad result. I use this amplifier because the limit of DC voltage of QA401 is 5v and QA470 is 50v. Normally I’m working with 15v supply. The solution is a capacitor before the BNC input but i dislike it.
Have you got any solution, for this,
It seems a mistake, the QA470 is internally connected diferent of your specifications.

Question 2:
Is possible to be more length in the Title field? Sometimes this space is limited for explanation, perhaps one solution can do that this field accepts two lines of text.

Question 3:
Imagine this serial circuit, DC-DC converter (Out A) to a LM317 regulator (Out B) to a Sziklai noise reductor (Out C) to a load of 100 Ohm.
I would like to view in the same graph the plots of the Noise Floor with input shorted and Out A, B and C with different colors.
Now only I have one snapshot of the Noise Floor, overlay in green color and the plot of DUT in blue color, only two.

Could you implement these features for new software versions?, now I’m using the v1.917.

Any suggestion will be apreciated. Thanks in advance.

Hi Marconi, the QA470 10 dB and 30 dB are indeed AC coupled. The input circuit is BNC->100K shunt->22uF/100V series->100K shunt->49.9 ohm series to opamp + terminal.

The 100K you measure is the bleed resistor. This is because if on Monday you measure a 48V DC power supply’s noise with the +30 dB amp, and then on Tuesday you decide to connect your phone and measure it’s output noise, if you did not have the bleed resistor, the 48V would still be be impressed on the input capacitor, and you’d subject your phone to ~48V and likely kill it. So, the resistor is present to bleed down whatever charge might be remaining on the input cap after measurement. The value could be larger or smaller, of course. The 100K takes around 10 seconds (5 times constants) to bleed a large voltage down to a small voltage.

This topology means your DC input resistance is 100K, your AC input resistance is ~50K

Is possible to be more length in the Title field

There’s no limit to the length right now, the issue is the text will flow off screen. I don’t think there’s any easy way to support multi-line right now.

I would like to view in the same graph the plots of the Noise Floor with input shorted and Out A, B and C with different colors.

Yes, agree, more is needed here. I think for the 2.0 release there will be a new graphing engine that can expose more flexibility here. But the time frame will be further out (probably 2H 2021). Ideally it would be nice to add annotations freely in many places.

Hello Matt,
I’m evaluating your answer, one solution is an external non polarized capacitor larger than 22uF at every input amplifier, like attenuator style tube, BNC female-BNC male. Another solution will be, if is easy, to desolder and extract the 100K bleed resistor with the benefit of less load and less noise.
Don’t worry about to kill a phone the next day :), I’m sure it never happened and will be only my fault.
Normally, the majority of measuring instruments, except frequency meter, oscilloscope etc… with added an Input Impedance Selector 1M/50 ohm, are true AC coupled.
The QA401 seems don’t have this resistor, there is a high value with an ohmmeter, and perhaps don’t needed because the highest voltage is 5V.

About the new graphics engine that you expose will be very helpful and better of my thought. Congratulations.

Thanks a lot.

Hi @Marconi, it should be straightforward to remove the bleed resistors if they are interfering with your measurements and if you are handy with a soldering iron. They are R25 and R31, located on near the dual BNC connectors.

Just remember if you measure a higher voltage DC supply that you should briefly short the inputs before moving to the next measurements to avoid impressing that voltage on the next DUT.

Thanks, Matt

Hello Matt,
This will be the best solution for my measurements.
In brief I’ll proceed to open the Red Magic Box and remove these resistors, without any problem.

Thanks for your quick and detailed explanation.
I wish you a good year