doubt square wave

[Adrian_Arg.]

The problem I see is that if I use the BNC-BNC the single block looks good, if I use the rigol probe x1, they look good, worry if I use them x10, the peaks appear, they only disappear if the amplitude is 6 volt or higher, can someone tell me what can it be?
  Rigol ds1054z and FY6900 20Mhz
- the first image, with BNC-BNC cable  (provided by feeltech), without attenuation of 20Mhz
- the second image, with BNC-BNC cable (provided by feeltech), with 20Mhz attenuation
- the third image, with probe of the rigol ds1054z, without attenuation attenuation of 20Mhz.
- the fourth image, with probe of rigol ds1054z, with attenuation attenuation of 20Mhz.
-the fifth image, with probe of the rigol ds1054z, without attenuation attenuation of 20Mhz. but the selected amplitude is 6V. The peaks are not exaggerated.

1- Subido en subir imagenes

2- Subido en subir imagenes

3- Subido en subir imagenes

4- Subido en subir imagenes.

5 - Subido en subir imagenes

[0culus]

Are you attaching the ground lead on the 10x probe to ground? That third image looks like it's just not grounded.

[Adrian_Arg.]

Are you attaching the ground lead on the 10x probe to ground? That third image looks like it's just not grounded.

In the third image, the only thing that changes is the deactivation of the 20 MHz attenuator of the oscilloscope. I think for the precise square wave 50 Ohms terminations

[w2aew]

I agree that it looks like you haven't connected the ground lead on the 10x probe, or you're using a very long ground lead. This video might help:

[Adrian_Arg.]

Thank you, I was watching the video, the oscilloscope brings some small additives to use with the probes in order to reduce the distance between + and ground, but the FY6900 of 60 Mhz, produced the same effect, otherwise it works correctly to be a generator of 69 dollars the FY6900 20Mhz, I do not have to forget that it is for hooby use  , I can measure inductors correctly.

[Adrian_Arg.]

 The technical service of FeelElec, answered me like this, how do I solve it?
"Dear Customers,
Overcharge is caused by impedance mismatch.
It can be matched by series resistance on the output signal terminal.
Yours sincerely,
Sarah
FeelElec

[rstofer]

Around here I just use probes set to x-10 as it creates the least load on the circuit.  Have you compensated the probes?  That will usually remove the overshoot at the leading and trailing edges.  In fact, if you overcompensate, you will see the corners round off.

If I'm using something like test leads with mini-grabbers, I don't expect to get clean edges.

The customer service response is correct.  Not really helpful, but correct.

[Adrian_Arg.]

Ok thanks for your response, if when I put the probes x1 or or connect BNC-BNC of the fy6900, it is almost perfect.

[rstofer]

When you compensate the probe, you are only tuning the x10 range.

https://www.electronics-notes.com/articles/test-methods/oscilloscope/scope-probe-compensation.php

Then too, maybe the reason you have overshoot is because, when the circuit is unloaded with a x10 probe, there actually are reflections in the wiring.  You can load the circuit with a resistor to ground while probing with the x10 probe.  See what happens with 1k to ground.

[Adrian_Arg.]

Thanks, I'm going to do that test.