BUF802: Clamping with output shifted to 2.5V

This should not be the case , i request you to look @ Figure 7-3 in datasheet.

Also please refer to the below post to ensure no misses as mentioned in the below post.

https://e2e.ti.com/support/amplifiers-group/amplifiers/f/amplifiers-forum/1559027/buf802-clamping-circuitry?tisearch=e2e-sitesearch&keymatch=Buf802#

https://e2e.ti.com/support/amplifiers-group/amplifiers-internal/f/amplifiers-internal-forum/1557402/buf802-clamping-circuitry?tisearch=e2e-sitesearch&keymatch=Buf802#

To start the testing with simplified set up can you use voltage sources in the lab to set CLL and CLH at least 1.5V away from your +ve peak and -ve peak of the signal.

So in your case 200mVpp signal the +ve peak and -ve peak is +100mV and -100mV. So add clamps atleast at 4.1V ( 2.5 + 0.1 + 1.5) and at  0.9V (2.5-0.1-1.5) .

Also ensure your BUF802 output load is terminated to 2.5V and not terminated to GND.

Thanks for the replay. 

A few notes, firstly Figure 7-3 shows the THD when getting closer to the clamp voltages, I understand that distortion will increase as you get closer to the clamp voltages but this is not what I'm worried about. The problem I am seeing is that the bandwidth has a drastic drop with my setup. 

I did some additional testing and narrowed down the problem. Firstly, I ran a test where I set the clamping voltages with a power supply as you suggested. When using 0.9V(CLL) and 4.1V(CLH) as you suggested the bandwidth is fine:

Now I tightened the clamps to my desired voltages of 1.5V(CLL) and 3.5V(CLH) and the bandwidth remains fine and virtually the same as above:

In fact I can tighten it up a lot more without a reduction of bandwidth (I understand that the THD will degrade as in Figure 7-3 but not the bandwidth)

Here is the interesting part, when I install the zener diode that sets my 3.5V (CLH) the bandwidth is still good and there are no problems. However, as soon as I install a diode to set the 1.5V (CLL) that is when the bandwidth starts to have issues:

So the problem is with using a zener diode to set the CLL clamp voltage. When it's being set by a power supply there is no issue with the 1.5V but when trying to set it with a diode, the problem appears. 

Do you have any more detailed schematics or diagrams of how the clamps are implemented? I'm trying to figure out why the diode is not setting the clamping voltage as needed? What are the requirements for setting these voltages? Is a low impedance source required? Or for example could a simple voltage divider be used? Do you have any additional details regarding the clamping circuitry?

Thanks     

Can you check to which potential is your load on the BUF802 terminated to . can you terminate the load to 2.5V and try.

i also believe the way you have placed zener it has ability to only pull current from CLL as a low impedance source. It does not have ability to push current via  a low impedance source.

Can you try the below configuration

I am also looping in our designer of BUF802 to also provide his views.

Hi, I will give that a try. Thanks. As well yes it is terminated to 2.5V. 

Did this idea work ?