Fast rising time pulse (1ns) peak and hold detect circuit

Hi, I'm working on design the peak and hold detect circuit to catch the 1ns pulse peak. I used the traditional two amplifiers plus diode strategy. But the voltage feed back amplifier has the slew rate limitation here. The holding circuit has the hopping problem.

Does anyone design or know any circuit can detect the fast rising edge pulse?

Thanks.

If you do not want accurate amplitude information, just the yes or no answer to Is a pulse present? you can preceed the op amp detector by a Schottky diode in series and a few pf capacitor on the output in parallel with the signal path. The first passive detector will hold the pulse long enough for the second stage to integrate the signal in a form of "pulse stretcher"

We do need to know the amplitude.

I have a similar test to complete, however my pulses are approx. 300 ps wide. I thought about using a fast comparator, but it is still a long way for me, untill I'll be able to test this circuit.

The comparator I was intending to use is the following:
Analog Devices ADCMP572

It has a 8 GHz equivalent input raise time bandwidth (80ps min pulse width).

You should be able to get free samples (2 or 3) in 14 days time. There are at least 2 other (simmilar) comparators from AD - you can find them on their web page.