31 Jan
2009
31 Jan
'09
8:38 a.m.
On Fri, 30 Jan 2009 Chuck Guzis said:
Read about it:
http://www.smpstech.com/filter00.htm
Interestingly, the real-world example cited is the addition of an
input filter to a PDP 11/40...
The part of this that really amazes me is where it says something like "long
power supply input lines" can also cause problems or oscillation. I suppose
this includes for example all the wiring in my house or at the office, over
which I have no control and very little detailed knowledge. I wonder if the
effective length of the AC input line is all the way back to the transformer
on the pole or wherever? If my guesses are correct then it is sort of scary
to plug in any SMPS anywhere! I would hope that most SMPS these days have
an input circuit that is reasonably immune to most of these problems, but I
certainly don't know that.
BTW, I have been reading the digest form of this group for a couple of years
now but I think this is the first time I have posted a reply of any sort.
If I goofed it up, profuse apologies in advance!
Later,
Charlie C.
1 Feb
1 Feb
1:10 p.m.
The part of this that really amazes me is where it says something like "long
power supply input lines" can also cause problems or oscillation. I suppose
Well, it's farily common for long (or otherwise high-inductance) power
lines to cuase problems in some circumstances. That's why we use
decoupling capacitors on logic boards :-)
This could be a problem with a DC-DC converter or similar (in other
words, if you're using one of those, make sure you decouple the input
side well).
But it shouldn't be a problem with mains-input SMPSUs, at least not
normal ones./ In the normal small SMPSUs, the chopper drive is not
synchronised ot the mains frequency [1], so there's no way of the supply
knowing waht the mains input will actually be when the chopper turns on.
And thus the enrgy from the mains is stored in large (and low impedance)
capactiros, the chopper effecrtively runs off that. So the HF component
of the mains input current whould be prtty small, and thus the adition of
mains filters or input wiring shouldn't be a problem
[1[ There heve been supplies using SCRs triggered at the right point in
the mains cycle. They're not common, they put a lot of 'hash' back on the
mains and are geenrally frowned upon.
-tony
8:43 p.m.
In the normal small SMPSUs, the chopper drive is not
synchronised ot
the mains frequency [1], so there's no way of the supply knowing waht
the mains input will actually be when the chopper turns on.
Yes...but (assuming the chopper frequency is high compared to the mains
frequency) there will be a period near the top of the mains AC cycle
when there is a chopper-frequency component to the draw. If the filter
caps are decently large, it will be tiny, but it will be present -
well, possibly excepting the case where the filter caps are large
enough that the time current is drawn for is less than one chopper
cycle wide, in which case I'm not sure how fair it is to say that. In
the presence of resonance it doesn't take much, though; the saving
grace is probably that the resonance has to be both very close to exact
and very low loss for the small boost each peak to be enough to produce
an oscillation that grows in the long term - and neither of those is
likely to be true.
[1[ There heve been supplies using SCRs triggered at
the right point
in the mains cycle. They're not common,
Not in computers, maybe. But - well, incandescent lamp dimmers.
Variable-speed power hand drills.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
2 Feb
2 Feb
12:35 a.m.
On Sun, 2009-02-01 at 23:43 -0500, der Mouse wrote:
Most 1970s colour TVs...
Gordon
[1[ There heve
been supplies using SCRs triggered at the right point
in the mains cycle. They're not common,
Not in computers, maybe. But - well, incandescent lamp dimmers.
Variable-speed power hand drills.
5568
days inactive
5570
days old
3 comments
4 participants
participants (4)
-
ard＠p850ug1.demon.co.uk -
csquared3＠tx.rr.com -
gordonjcp＠gjcp.net -
mouse＠Rodents-Montreal.ORG