Thank you very much! Thank you all .
Sorry, I didn't exactly describe the problem. The problem is sometime no trigger and some time too many triggers. At first I think this feature can't work . So I did not carefully think about it.
Now, I draw the conclusion:
1 The delay trigger function can work in the 1000z scope including cracked 1054z;
2 Why my previous setting can't work?
(1).Most of time, the real signal is below pic 1 . The falling edge of channel 1 is before the falling edge of ch2 in the green oval.
At this moment , when use > setting , the trigger often can't happened , because the gap between two falling edge less than the setting.
(2) Another circumstance is pic 2. When triggered , the falling edge of channel 1 is the red arrow pointed not the green arrow pointed..
At this moment, the trigger happened too many times, because the gap between the two edge connected by the green line is bigger than the setting.
So, sometimes nothing happened ,and sometimes too many waves just like edge trigger effect.
3 How catch the signal I care
Use the setting "<>" , just like what @TurboTom did.
The lower limit is 50us , and the upper limit is 120us. So,I can catch the abnormal signal in the red oval in the pic 3.
.
The lower limit in the pic 3 is 80 us that is just for debug.
By the way ,this function is very rare in many scope. When I suspected the 1000z , I waned to buy a cheap second hand scope to test my idea. But after I checked old Agilent_54622D , HP 54645 and Agilent-54845A , they all do not have this feature ? despite the fact that 54622D and 54645 were both equipped with digital analyzer .