Log Preamp?

[Teleno]

The two-stage preamp is a great improvement for those interested in early sampling.

Given the exponential nature of both transient and signals, earlier sampling runs into saturation problems. I recently replaced the IRF740 MOSFET (400V) for a 900V one hoping to sampe earlier. Instead, I got more preamp saturation (due to a stronger signal), so the high preamp gain defeated the purpose.

Adaptive gain could be achieved with a Log preamp (low gain for strong signals and high gain for weak ones). I would like to discuss the convenience of this approach.

Some IC options:

AD8307 500MHz, 1.5 nV noise - not expensive.


Another option is a Temperature-stable Log Amplifier based on a dual op-amp



The same NE5532 can be used or a pin-to-pin direct replacement LM4562 which has lower noise (2.7nV) and higher slew rate (20V/us).

Wating for your comments.

[Teleno]

Application note on Log amplifiers: http://www.ti.com/lit/an/slyt088/slyt088.pdf

[green]

The two-stage preamp is a great improvement for those interested in early sampling.

Given the exponential nature of both transient and signals, earlier sampling runs into saturation problems. I recently replaced the IRF740 MOSFET (400V) for a 900V one hoping to sampe earlier. Instead, I got more preamp saturation (due to a stronger signal), so the high preamp gain defeated the purpose.

Adaptive gain could be achieved with a Log preamp (low gain for strong signals and high gain for weak ones). I would like to discuss the convenience of this approach.

Some IC options:

AD8307 500MHz, 1.5 nV noise - not expensive.


Another option is a Temperature-stable Log Amplifier based on a dual op-amp



The same NE5532 can be used or a pin-to-pin direct replacement LM4562 which has lower noise (2.7nV) and higher slew rate (20V/us).

Wating for your comments.

[ATTACH]37149[/ATTACH]

No comments, just questions. Does the log amp make GEB difficult? With a linear amp a 20uv target signal causes the same change at zero input or with a ground signal offset. With a ground offset signal does the 20uv target signal cause a larger, smaller, or the same output change as it does with no ground signal?

[Teleno]

No comments, just questions. Does the log amp make GEB difficult? With a linear amp a 20uv target signal causes the same change at zero input or with a ground signal offset. With a ground offset signal does the 20uv target signal cause a larger, smaller, or the same output change as it does with no ground signal?

The MPP has no ground balance, so the issue is irrelevant in this case.

In the more general case the change gets smaller as the ground signal gets larger, making ground elimination more difficult. However, since a Log amplifier behaves lineraly at small signals, when the ground signal is not severe the change would be nearly the same.