26 Apr
2011
26 Apr
'11
5:21 p.m.
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-
bounces at classiccmp.org] On Behalf Of Dave McGuire
Sent: Tuesday, April 26, 2011 5:05 PM
To: On-Topic and Off-Topic Posts
Subject: Re: new here
On 4/26/11 7:59 PM, Bob Bradlee wrote:
can not be 0
internal and 0 external load.
OK, it sounds to me that in your designs the difference between load and no-load
conditions are sufficiently small that you are designing your control loop to no-load
conditions and loading is not substantial enough to push your design out of that window.
As I consider it, I guess I should not say that *any* switcher will fail to run without a
load - I have a little wall-wart for my Android tablet that puts out 5V just fine
(although I have to wonder if there isn't an internal load that consumes a small
parasitic amount of energy). But for the designs we see in vintage systems - such as the
H7100 - they require a significant load for the control loop to function. In such systems
the load is an indispensable component in the control loop. -- Ian
> I have quite a few of them that run just
fine without a load.
I've
also
designed quite a few that also run fine without a load.
There needs to be a small internal load than to maintain bias, there something about efficency going to infinity.
I'd like to see the math on that. I have a pretty good grasp of
switching regulators; I've done a lot with them in various designs.
Admittedly only very small ones for portable equipment though, in the
sub-1A range, but the concepts are the same. PWM a transistor between
saturation and cutoff, smooth it out with a capacitor, sample the
output
into an error amplifier to control the PWM duty cycle. No load there,
unless you're talking about leakage across the PCB and such.