LMH6586: LMH6586

Bela Geczy

Part Number: LMH6586
Other Parts Discussed in Thread: TMUX1134,

I would like to use this part to switch non-video analog signals around +/-400mV. Will this device work in that application to give the same output? If so, what would the setting be?

over 2 years ago

0 Bryan Marshall over 2 years ago

TI__Genius 10974 points

Hi Bela,

Can you give more details on the application? Do you really need a crosspoint switch with 16 channels?

Let me know so I can best recommend the correct part for your application.

Thanks!

Bryan

0 Bela Geczy over 2 years ago in reply to Bryan Marshall

Prodigy 40 points

Hello Bryan,

Yes, I need a 32 to 16 analog switch. It does not need to be non-blocking however. Do you have any other suggestions?

Thanks,

Bela

0 Bryan Marshall over 2 years ago in reply to Bela Geczy

TI__Genius 10974 points

Hi Bela,

Thanks for confirming. So the LMH6586 cannot support -400mV signals as it violates the absolute maximum specification in this datasheet. For the channel count you need, I would advise using several muxes together to form this 32:16 mux. I would recommend using 8x of our TMUX1134 devices which is a 2:1x4 channel that can support +/-2.75V signalling.

Let me know if you have further questions on this.

Thanks!

Bryan

0 Bela Geczy over 2 years ago in reply to Bryan Marshall

Prodigy 40 points

Hi,

The LMH6586 is AC coupled and it internally adjusts it upwards for processing. Figure 3 in the datasheet shows a signal that is about 1.0V peak to peak so a +/-400mV in would only be 800mV peak to peak. What am I missing?

Bela

0 Bryan Marshall over 2 years ago in reply to Bela Geczy

TI__Genius 10974 points

Hi Bela,

So this is not about peak to peak voltage, it is in reference to what your negative voltage is. When your input voltage is too low (less than -300mV as specified in this datasheet), then you risk the internal ESD diode beginning to conduct and shorting out the device:

So, during the condition when your input voltage is -400mV and your VSS is 0V, that creates a forward bias on that internal diode and will cause a significant current to flow that can damage the device if not limited.

Let me know if this is clear.

Thanks!

Bryan

0 Bela Geczy over 2 years ago in reply to Bryan Marshall

Prodigy 40 points

Hi,

Yes, ok, this makes sense. Therefore I should limit my input signal to +/-250mV. If I did, what would the output range be?

Bela

0 Bryan Marshall over 2 years ago in reply to Bela Geczy

TI__Genius 10974 points

Hi Bela,

This is a rail to rail device, so you can output up to the VDD rail of the LMH6586. There is an output gain buffer than can be set to 1x or 2x, so as long as that is below the VDD rail, then there should not be any issues.

Thanks!

Bryan

0 Bela Geczy over 2 years ago in reply to Bryan Marshall

Prodigy 40 points

If I apply an input signal of +/-250mV, what are the DC output values? Also +/-250mV (gain = 1) or something else?

0 Bryan Marshall over 2 years ago in reply to Bela Geczy

TI__Genius 10974 points

Hi Bela,

That is correct. You can see this in the datasheet where it portrays the input vs. the output and appropriate gain setting:

Thanks,

Bryan

0 Bela Geczy over 2 years ago in reply to Bryan Marshall

Prodigy 40 points

Thank you, you have been very helpful. My last question, which the figures do not show, is what are the DC values of the output. If the input voltage is +/-200mV, at a gain of two, does the output go from -400mV to +400mV? I don't think so as there is no negative voltage applied to the switch. But what is it?

0 Bryan Marshall over 2 years ago in reply to Bela Geczy

TI__Genius 10974 points

Hi Bela,

You will not be able to amplify the negative signal as it will clamp to the lowest rail (in this case GND). I would advise not to use this part if you actually want to pass the negative voltage signals as it really is not optimized to do so and has a lot of signal conditioning. If you simply want to pass a negative signal to the output with little to no signal attenuation/modification, I would suggest using the TMUX1134 as previously suggested.

Thanks,

Bryan

0 Bela Geczy over 2 years ago in reply to Bryan Marshall

Prodigy 40 points

What do you mean? The input to the switch is AC coupled and level shifted internally to be used. Let me ask this again in a different way. If I apply a signal that is -200mV to 200mV BEFORE the coupling capacitor, with a gain of 1, what is the DC output of the switch? If it's a gain of 2, what is the output of the switch?

0 Bryan Marshall over 2 years ago in reply to Bela Geczy

TI__Genius 10974 points

Hi Bela,

In that case, it will depend on what the V_CLAMP pin is set at. The recommended level is 300mV. So given that, it will set 300mV as the baseline voltage (so it would shift say the -200mV signal to 300mV) and then adjust the positive side by the same amount (so 200mV becomes 700mV). So peak to peak would still be 400mV (300mV to 700mV). The output would then depend on if you have your gain at 1x or 2x (2x would be 600mV and 1400mV). You can see an example of this in the figure below:

Thanks,

Bryan

0 Bela Geczy over 2 years ago in reply to Bryan Marshall

Prodigy 40 points

Ok, this is helpful. Let me put this in my own words and see if it is the same meaning as yours.

The figure you show has a 1.0V peak-to-peak input signal prior to the coupling capacitor. At a gain of 1, the output is also a 1.0Vpp signal and is level shifted up by 600mV.

If have a +/-200mV (400mV peak-to-peak) signal prior to the coupling capacitor and I set V_Clamp to 300mV, then at a gain of 1, the +/-200mV signal gets level shifted up by 600mV. Therefore, the DC output is from 400mV to 800mV.

At a gain of 2, the 400mVpp becomes 800mVpp and is level shifted up by 600mV. Therefore the output DC is from 400mV to1.200V.

Is my understanding correct?

0 Bryan Marshall over 2 years ago in reply to Bela Geczy

TI__Genius 10974 points

Hi Bela,

Please look at my previous response again.

The V_CLAMP voltage is what the negative signal will be clamped to. So if you set it to 300mV, the -200mV will now be clamped to 300mV (not 400mV). The positive signal will then be shifted up by the same amount to be 700mV (peak to peak would be still be 400mV, 300mV-700mV).

Thanks,

Bryan

0 Bela Geczy over 2 years ago in reply to Bryan Marshall

Prodigy 40 points

I see. That's the input V to the part. At gain = 1, the output would be 300mV to 700mV. Then the gain of 2 output would be 300m to 1.1V?

+1 Bryan Marshall over 2 years ago in reply to Bela Geczy

TI__Genius 10974 points

Hi Bela,

For 2x gain it would be twice the VCLAMP level for the lower end:

So that would be 600mV and then the postivie side would be 1400mV.

Thanks,

Bryan

Switches & multiplexers

Switches & multiplexers forum

LMH6586: LMH6586