Wideband Active high pass and low pass filter filter

[nix_arif]

Wideband Active filter

Hi,

I tried to design high pass and low pass filter using dual operational amplifier.
I am using LMH6715 from national semiconductor, my low pass filter cut off freq is 10MHz and high pass cut off freq is 20KHz.

I already use spice model and result is good in simulation. But fail in implementation circuit.

Is there anyone ever use this opamp for wideband filter. Please advise me on the schematic and layout design.

 

[keith1200rs]

Wideband Active filter

Post your circuit diagram, then we might be able to help.

Keith.

 

[nix_arif]

Re: Wideband Active filter

This is the image for pspice simulation. The simulation can work, but the PCB cannot work..i will attacted it

 

[nix_arif]

Re: Wideband Active filter

This is my PCB layout for the filter circuit.. Thanks
Any idea..

 

[keith1200rs]

Wideband Active filter

Looks OK. I assume you have built it with the two stages cascaded?

Are you getting nothing through it? If so, what is providing the DC bias for the first stage? You have a voltage source in the simulation but if the real circuit connects to something which is AC coupled then it won't work.

Check the DC voltages at all points - it may give a clue to the problem.

Keith.

Added after 4 minutes:

On the PCB you have power supplies V+ & V-, but I cannot see the ground - the junction betweent he two power supplies.

Keith.

Added after 51 seconds:

In fact there seem to be no ground connections.

 

[nix_arif]

Re: Wideband Active filter

There is dual amplifier inside the IC LMH6715, so I use them separately as LPF and HPF, ground is all the major red plane.

From the simulation I should get cutoff freq at 20KHz for HPF and 10MHz for LPF. But the result not seem like a filter after testing.

 

[LvW]

Re: Wideband Active filter

As KEITH has stated: Pay attention to the dc bias at the non-inv. input of the first opamp.

 

[FvM]

Wideband Active filter

LMH6715 is a current feedback OP that can't be stable operated without a feedback resistor.

 

[keith1200rs]

Wideband Active filter

Duh! I never noticed it was current feedback!

Keith.

 

[LvW]

Re: Wideband Active filter

FvM, congratulation! Do you have all these numbers and abbreviations in your mind?
LvW

I suppose that nix_arif did only an ac simulation and not a transient.
This is another example showing that ac analysis cannot reliably reveal instabilities.

 

[FvM]

Re: Wideband Active filter

Do you have all these numbers and abbreviations in your mind?

Not particularly. I suspected special properties of the part and checked the datasheet. Actually I'm often using CFB OPs in
fast circuits and really like them. If the SPICE model is correct, it should reveal the loop instability in AC analysis. But possibly,
the simulation frequency range didn't go up that high.

 

[nix_arif]

Re: Wideband Active filter

What do u mean by DC bias, is it at input?
current or voltage bias?
I have tried it but still not work..

Thanks

 

[FvM]

Re: Wideband Active filter

The bias problem has been mentioned for the low pass circuit, that needs to be fed by a DC coupled generator to work correctly.
A standard lab generator e.g. would work.

Unfortunately you didn't tell about the observed problems. "Cannot work" is somewhat unspecific, isn't it?

Did you understand the required "feedback resistor" point? When you operate a high frequency current feedback OP without it,
it will oscillate at a x00 MHz frequency with high amplitude, run hot and even may be damaged. High frequency OPs are required
to drive high output currents and are ususally not fully protected.

According to the datasheet, a 500 ohm resistor between output and inv. input would allow stable operation. Making it e.g. 1 kOhm
reduces the loop gain and can improve the circuit stability, if you don't need the full OP gain-bandwidth product.

You should add feedback resistors and then start with DC bias point measurements of the circuit. Just cut the connecting
traces between the two pins and solder a small (e.g. 0603) chip resistor to the pins, for both amplifiers. Alternatively, you can use
a voltage feedback dual OP with compatible pin assignment in the circuit.