Eeyore said:
The strain gauge thread reminded me of a product with an 'issue' I was once
asked to take a look at.
It was the receiver section of a marine radar.
The detector was followed by several variable gain stages. These were wideband
IC amplifiers made by Ferranti ( now Zetex ) with a control current / voltage
that varied each IC's gain over about something like a 12-15dB range. The
control ports were driven in parallel so that gain was swept equally between all
the devices.
They were at least 6 of these cascaded to provide the necessary gain before the
signal went off for further processing.
After the pulse was transmitted ( and the receiver input clamped ) the gain was
ramped up with a triangle waveform to provide the necessary low gain for near
targets and higher gain for distant targets.
I took one look at the schematic, grinned and said I could fix it.
What problems ( not less than 2 of them ) were they encountering ?
Graham
You have to be demented to think this so-called brain teaser has any
informational content whatsoever: a crummy little marine radar, ppi
display, detector of unknown specification, a gain cascade of unknown
specification, and a "triangle waveform"- don't you mean sawtooth?- the
phone call in the middle of the night, your toothless grin, and a bunch
of other troll trash. No one is going to use a wideband video amp that
does not have logarithmic gain control characteristic, so that has
little do with causing any intrinsic problems. The compromise here is
how to go about maintaining a constant return S/N independent of range
when the attenuation is due to two distinct phenomena: the usual
spherical spreading loss everyone is familiar with by way of the radar
equation varying as LOG(R), and the atmospheric absorption loss varying
directly as R, loss understood to be in dB. You did say this is a marine
radar which I take to mean surface object detection out to the horizon,
making it short range, so the spherical spreading is predominant.