Signals & Noise: September 2, 1996
to send a letter to signals & noise via email: |
"Capacitor amplifier reduces ripple without dc loss" by Martin Kanner (June 6, 1996, pg 137) is interesting and obviously works when you consider ac signals. With a PSpice model, it is easy enough to show that for transient signals the performance depends on the ability of the op amp to source or sink current. This action, in turn, depends on the supply voltage and energy stored on the op amp supply rail. So, the projected application to switch-mode power-supply energy storage for large signal-load step changes doesn't work unless the op amp can source the required transient current and the supply for the op amp has the required energy stored on its supply rail.
The only energy stored is that of the capacitors. The op amp can't store energy, and it simulates this performance only while operating within its linear range. If the "high-voltage dc supply" (Figure 3) is large compared with the output voltage, then you can store more energy in the capacitor.
Jim Booth
via the Internet
I agree with Steven H Leibson that when most large companies try something radical, people view the innovation as a conspiracy to swat the smaller competitors ("Conspiracy theories," June 6, 1996, pg 11). However, I think that there is one true "conspiracy," and some of its players may not be aware of their role.
Microsoft has been growing Windows since its inception. Windows 95 requires so much memory and hard-disk space that I won't be able to run it at home until I discard my 486DX4-100 with 8-Mbyte memory and 300-Mbyte hard drive and get myself a Pentium.
I had an Ensemble, and, for a long time, I ran it on a 386. The system had a special version of Quattro, had all the goodies that came with Windows (even a Wordlike word processor that provided WYSIWIG output on a dot matrix), and was Motif-based. Ensemble kicked Windows' butt! The system took maybe 2 Mbytes of hard-disk space, used 1 Mbyte of memory, gave better performance, and cost $90. The only problem was that the company didn't provide a software developers kit soon enough, and, consequently, few people wrote for it.
Meanwhile, Windows pushed a memory/hard-disk hog, and the producers of memory/hard-disk space responded by allocating more room. But, as more memory was available, Microsoft increased the requirements (without really improving the product). Now, Windows doesn't run decently without 16 Mbytes of memory and what used to be all the hard-disk space available to the PC.
I realize that more features require more memory, but Microsoft has been dominating the market on a substandard product. I understand that our jobs depend on the volume of our sales, but it's depressing that we have to push so many products to support such a mediocre-yet-pervasive product. A conspiracy between memory makers and Microsoft? Maybe.
Alfredo Ortiz
Barco Chromatics
Tucker, GA,
| Sound off |