Reply-To: "Kevin Aylward"
From: "Kevin Aylward"
Subject: Re: connection between low currents and high phase shifts?
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
X-Inktomi-Trace: public1-pete2-5-cust19.pete.broadband.ntl.com 1033285870 17787 22.214.171.124 (29 Sep 2002 07:51:10 GMT)
Date: Sun, 29 Sep 2002 08:51:07 +0100
NNTP-Posting-Date: Sun, 29 Sep 2002 08:51:11 BST
"Ron J." wrote in message
> Hi everyone...hopefully someone can clear up this for me. I've
> seen in a lot of place frequent references to high phase shifts
> made in conjunction with low levels of current. For example,
> are supposedly prone to oscillations because in the input bjt of the
> pair, low emitter currents lead to high phase shifts (this confusing
> statement is taken out of the H&H book, btw).
>I also see frequent
> references to low currents/high impedances leading to low phase
> shifts when dealing with current feedback topologies. As usual, the
> sources where I see this never really get into any significant detail,
> so maybe someone here knows what the heck this is all about?
> Thanks in advance.
These statements are really referring to the relation between bandwidth
and current. The BW is increased (usually) with higher current. Lower BW
means more phase shift.
Higher current cases increased BW principle for two reasons.
For example, if a transistor stage requires say, 3V across its collector
resistor, then at 100ua this would be 30k. If the current was 1ma the
resistor would be 3k. Assuming that any capacitance on that node is
constant the BW at that collector would be 1/2.pi.RC, which is obviously
10time worse for he 30k resistor compared to the 3k resistor.
The second condition is the case where a transistor has a transistor as
a current source load. In this case the "resistance" at the collector
node is an active one is given by ro=VA/Ic, where VA is the early
voltage. The gain remains constant with Ic, i.e. VA/Vt, where Vt is
q/KT. However, it can be seen that increasing Ic reduces ro, hence its
ro.Cload will result in a higher BW as in the resister load case.
So, the summary is to get large BW and low phase shifts, one needs low
resistances, and low resistances mean high current.
The catch is that you cannot keep increasing current to get better BW.
At some point the transistor performance degrades with two much current.
This requires putting more devices in parallel to share the current.
This of course increases the capacitance, which gets you back to where
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.