General > General Technical Chat

Review: Hantek DDS 3X25. Anyone own one?

<< < (92/108) > >>

robrenz:

--- Quote from: jahonen on January 08, 2013, 09:00:04 pm ---Why not just get a coaxial attenuator, fixed ones should not be too expensive? You can also cascade them to get desired attenuation if one is not enough.

Regards,
Janne

--- End quote ---

Is there something in the internal design of a coaxial attenuator other than the shielding effect of the outer shell being grounded that makes it better than an equivalent non inductive resistive divider with proper shielding?

alm:
A well designed attenuator will maintain a 50 ohm impedance throughout, a loose resistor will most likely cause an impedance mismatch, with the usual signal degradation and loss.

torch:
I did some testing of resistor voltage dividers, with a 50 ohm audio transformer standing in for the antenna input transformer. As you guys warned, the signal was badly distorted.

Two pieces of wire placed side by side resulted in a fairly decent attenuated coupling. Reversing one piece of wire inverted the signal, but that shouldn't matter.

Mechatrommer:

--- Quote from: robrenz on January 08, 2013, 09:16:25 pm ---Is there something in the internal design of a coaxial attenuator other than the shielding effect of the outer shell being grounded that makes it better than an equivalent non inductive resistive divider with proper shielding?

--- End quote ---
i believe you missed the attenuator construction i teared down recently and some usefull link(s). this is china cheapo one. if you can lay better and uniform carbon layer and then verify its profile then thats a good engineering ;) here the teardown https://www.eevblog.com/forum/beginners/replace-carbon-layer-with-passive-resistors-components-in-attenuator with 1 concise and reasonable answer why not to use normal resistors.

torch:
I am aligning a short wave radio and need an audible tone AM modulated on a carrier -- from around 9MHz to 17MHz. So I figure a 1000:1 ratio should be fine, giving me a 17KHz tone at the upper end.

In ArbExpress, I generated an arbitrary waveform using the expression

Sin(1000*w)*(1+0.75*Cos(1*w))

to generate 1000 sine waves, the peaks of which form one overall sine wave. Zooming in on the waveform within ArbExpress confirms the waveform is as expected:



When I load it into the DDS3x25 via Goltek, the overall shape appears correct (goltek set to 20 hz):



However, zooming in a bit, something doesn't look right -- instead of sine waves, it appears to be an underlying modulated wave:



Zooming in further confirms this. I do have the right number of pulses, but they ain't sine waves, They are groups of square waves that gradually average out into the overall shape:



I'm not entirely sure what is going on here:



I suspect those random drop-out glitches are something to do with the Rigol and timing. If I zoom out still further the waveforms appear fine again. It only happens at that one setting.

Is this some sort of glitch with Goltek or with the Hantek? Or am I doing something completely wrong?

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod