MUX506: MUX506

Hello Parker,

Thanks.

The load of the MUX is an input to LMH6570 terminated with 10K to ground. (see attached drawing)

Regards,

Giora

Hi Giora,

So for this set up the following eq. circuit is used:

Yellow = R_on which has a worst case spec at 12V single supply = 430 Ohms - this is the closest spec we have to +/-5V in the datasheet and it may be a bit higher at +/-5V.

Green = C_on - the on capacitance this should be right around the typical value of 18pF

Blue = Is the Input capacitance of the LMH6570 ~ 0.8pF 

Red = This is the load resistance - it is a parallel combination of the Input resistance of the LMH6570 (5k) and the 10K Ohm Termination resistor  which is ~3.33k Ohms.

This leads to the approximate -3dB point for worst On resistance is:

This value is an approximation of the worst case R_ON value at a 12V single supply - which is the closest spec'd supply to +/- 5V - so worst case R_on may be a bit worse. However this is assuming you are operating at max temp conditions and using the full signal swing. These factors will also affect bandwidth so the effective bandwidth may be slightly higher than above depending on the nuances of the application. 

When I simulate this in spice using passive elements I get the following AC response:

Which has its corner frequency at ~16.71MHz. This is more accurate as this isn't the linear

If this is a concern for your application - can you please share what your anticipated required signal bandwidth is going to be? Also if I could get the I/O signal range and expected operating temperature (ambient) -  That way I can see if we have a better fit that can suit the higher frequency signal or if the actual application may be able to handle higher frequencies than a worst case scenario.

Please let me know!

Best,

Parker S. Dodson

Hello Parker,

Thanks for the answer.

1. In the simulation (25 C) I got  ~ 30MHz using +/- 7V for the supplies, Surely it is not the worst case. 

2, Ambient Temp. will be ~ 50 Celsius degree.

3. I need at least 40Mhz BW.

So I was mislead by the "500Mhz BW".

Regards,

Giora 

Hi Giora,

Do you have any flexibility in your design for higher supply rails to the Multiplexer? After looking through our portfolio I am worried you will have similar problems with our other 16:1 options. The higher on resistance at lower supply voltages are killing the bandwidth.

If you can get the +/- 10V supplies you are at ~45MHz and possibly higher bandwidth if the signal swing is less than the +/- 10V. 16:1 multiplexers have a tendency to be lower bandwidth because the 16:1 configuration introduces a lot of parasitic capacitance which degrade bandwidth. Also our 16:1 multiplexers that can handle a bi-polar supply have higher on resistance at +/-5V. 

I apologize for the 500MHz confusion - there was a mistake ; under optimal conditions the MUX507 (2 Channel 8:1 Device related to the MUX506) can hit 500MHz under optimal conditions (where this device can hit 240MHz under optimal conditions) and that is what is being displayed there. At +/-5V supplies the parts we have that operate at that voltage will not be working in optimal conditions.  

If you need any more help please let me know!

Best,

Parker Dodson

Hi Parker,

Thanks for your detailed answer.

We are going to do as you suggested and bias the MUX with +24V and -5V.

First measure indicates ~ 46Mhz.

Is it OK to work with asymmetric voltages? 

Regards,

Giora