The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: "John Jardine"
Subject: Re: 2.5kW-3kW-SMPS-(cheap)-suggestions?
Date: Fri, 20 Dec 2002 01:45:35 -0000
NNTP-Posting-Date: 20 Dec 2002 01:32:41 GMT
X-Newsreader: Microsoft Outlook Express 5.00.2615.200
John Woodgate wrote in message
> I read in sci.electronics.design that John Jardine
> wrote (in >) about '2.5kW-3kW-SMPS-(cheap)-suggestions?', on Wed, 18 Dec 2002:
> >Did you buy this one? ;-)
> Not really. It's too easy for someone outside the problem to:
> - think impracticably;
> - think there's a better way.
> The problem is knowing which applies!
> Put it like this:
> Take a *known satisfactory* 50 V design. Halve the cross-section of the
> wire and run two wires in parallel. All still OK? Now, instead of
> running them in parallel, run them in series from 100 V. Yes, you have
> to look out that the wires can't short together at the feed end. But
> that's not rocket science, surely?
(The power connections are on widely seperated terminals)
> Regards, John Woodgate, OOO - Own Opinions Only.
> Interested in professional sound reinforcement and distribution? Then go
> PLEASE do NOT copy news posts to me by E-MAIL!
I'm rather confused by the wire slicing and dicing as all I'm seeing is an
illustration of maintaining constant power. I may be missing some
interesting physical layout here in which case I'm talking 'across' you. If
I am, could you illustrate it differently?.
Say existing manufactured item is 50V at 1A.
That's 50ohms of a particular length of resistance wire. 50W. (The 50W of
heat is of course a key factor.
Task ... redesign to a much more usable 100V.
So, simply change R to 200ohms to maintain the same 50W dissipation. Select
a suitable resistance wire and et voila' Sorted!.
No it isn't ...
1) We cannot change the length of the wire used.
(Imagine a flat or circular inductor coil). Extra wire would change the
heating surface area or wiring pitch. These are rigidly defined by the
particular heat flow dynamics and final joint strength (1mm wire, 1mm pitch
results in 0% joint, as melting material must flow -through- seperated
Increasing this wire length also increases the linear wire
expansion.v.temperature. As the wire can be 'moulded in' it is restrained at
low temperatures by solid plastic. Put power in and wire temp' ramps up, it
can't expand, so internal stresses increase. At some temp' the plastic will
soften sufficient for all the built up stresses to start releasing. Too long
a piece of wire (or too intense power input) means the internal stresses can
release in an extremely fast and localised area such that all the extra
expansion length can suddenly appear as as a few cm's of 'chicane' in one of
the windings sufficient to short across many of the others. Or if pitch
spacing is too low due to extra length then the windings can expand sideways
(quite normal) and clash into each other
2) So as our wire length is immutably fixed we simply need to replace it by
a different diameter (x0.5) or use a different rho factor and diameter.
Not that easy ...
(For demo I'm showing some numbers here from an old prog I wrote.)
Assume existing 50V design uses 15mtrs of wire at a 2mm pitch winding (or
spiral) rate. Resistance wire type is .52mm dia stainless steel. 50ohm at
23degC. 60.5ohm at 200degC. Max linear expansion 35mm. Cost say 22pence.
First option ... go for a lesser diameter of existing wire so ...
0.26mm SS gives 200ohms 50Watt, same material, same length, same tempco, and
available at 1/2 the price!.
Useless! ... The machines can only handle wire down to 0.4mm . £15k to alter
and then problems start turning up with wire stressing, requiring a further
expensive stress relief operation .
Second option ... go for a different material that will work with the
machines .4mm min, yet still give the 50Watts at 100V using same length.
Useless! ... Out of 18 resistance wires none can get down to 50W at 0.4mm
We're buggered, so lets pretend we made a mistake and the example problem is
really 50V, 2ohm, 25amps, 1250watts. Stainless would have been impossible to
use for this design (>£5) and the >2.5mm dia is too big for the machines. so
copper (or a copper mix) *would* have been used for size and cost.
So assume a particular copper mix (K3) heating wire ... gives 0.54mm dia,
2.8ohm at 200degC, cost would be 27pence. 38mm max expansion.
So at 100V can't use K3 again as dia would be too small. So look for
another. Most are u/s but a cheap copper (K10) mix looks promising ...
0.5mm, 9.1ohms hot, cost 25pence, a saving!.
Useless!. ... This time we suffer from a temco problem. The tempco is only
9.1/8= +13% compared to previous tempco of 2.8/2=+40%. This will mean that
too much power is being dumped into the plastics at the tail end of the
cycle. Heat flows and material degradation will suffer badly. (too many
total Joules per mm^2).
Stainless would work nicely in this 50V to 100V upgrade as it's tempco is
useable but the 1.31mm wire costs £1.42 a go, pitch spacing problems turn up
and the accountants would (and should) veto the idea simply on cost
The devil is always in the detail (and there's more) but this is just
normal, straightforward, production engineering. The clever stuff is in
getting a handle on the internal heat flows.
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup