THS4508: can you give me some suggestions about to amplify a pulse which bandwidth is 50MHz and amplitude is 0.7mV?

There is not enough detail here to go on - 

1. Amplify? what gain?

2. required input impedance?

3. settling info or noise or HD concerns?

4. ADC common mode and any filtering from the FDA to the ADC?  

Look at the THS4541 in the meantime, 

hello！

    I'm sorry I didn't describe it！

1 yes，I would like an type of amplify,and Try to gain between 30 and 100 times

2  The output impedance of the upper stage is 50 ohms.

3 I do not concern about the noise or HD

4 Sorry I do not know this question now

Directly, I used your company's OPA858 to amplify the current pulse signal to 0.7mv, which has a bandwidth of 50MHz. Now I want to amplify the output voltage signal of OPA858 by 30 times to 100 times, and sample the amplified signal with ADS4129. What device should I choose for secondary amplification

Well that is a lot of gain. 

1. When you say 50MHz is that the -3dB point or do you need something higher. 

2. Can you be AC coupled? that will make the DC offset much more manageable, 

3. With this much gain, you do want to work to lower noise - I know this stages input referred noise will be then input referred by the OPA858 gain, but lower will be better in the FDA stage. 

4. Is there any transmission line distance from the OPA858 to the next stage? If not, you may not need double termination and then we just need an input Z that the OPA858 can drive. 

The ADS4129 needs an input Vcm of 0.95V that is supported by some, but not all of the FDA's assuming you want single supply operation. If you were AC coupled, could run the FDA swing centered on its supply then reset the Vcm on the ADC side of output blocking caps. 

hello！

1. 50MHz is -3db point

2 I'm not going to do AC coupling, because the dc offset is not that big, so I can get rid of it later

3 Yes, the lower the noise, the better. Of course, the signal output of the upper level is already submerged in the noise. I will extract the signal during data processing.So as long as the noise is lower than other devices

4There is no transmission line from OPA858 to the next stage, but I want to drive the differential input ADC, so I chose FDA

Thanks