| Electronics > Projects, Designs, and Technical Stuff |
| Improve bandwith of peak detector |
| << < (6/6) |
| Kleinstein:
In the 2 OP peak detector the 2 nd OP has to quite fast, as it's inside the feedback loop of the other OP. For a fast detector the memory cap should be way smaller. So not 1 µF, but more like 10-100 pF. If long time storage is needed, a second slower stage can be added. Especially when using 2 stages, the input bias for the first, fast stage is not critical - it is than more help full to to have a defined negative bias. Some years ago I used LF356 and LM318 for a fast peak detector. At first it look counter intuitive to have the high bias LM318 as the buffer, but the bias is more a good thing than a problem. |
| AE7OO:
I would take a look at the modern versions of the LMC660(I have a tube of original Nat Semi's I picked up over 15 years ago). When you combine them with a analog switch they are an extremely low loss sample and hold, which would be foundation for a peak detector. I doubt you could use the LMC660 itself, GBP of 1.4Mhz and a low slew rate. But the newer versions are faster. I've included a copy of the LMC660's datasheet so you can see what I'm talking about. |
| Navigation |
| Message Index |
| Previous page |