JIMMIE wrote:
On Sep 16, 8:46 pm, brian whatcott wrote:
Andy GD1MIP wrote:
Hi I am building a (240volt in) 13.8 volt (out) PSU ...
the transformer must have a secondary of 16.5 volt.

I have a Talema branded toroid I want to use, its spec is as follows
(from maker)...
73 Andy
Now that the ideas are cooking.... consider this.
1100 watt defunct microwaves can be had for the asking.
They have a particular transformer core design that makes it easy to
saw off the 1 KV secondary, and with the generous core window, you could
wind 18 to 20 turns of wire good for 10 amps. The turns per volt is not
much over 1 tpv.... And you have a decent rectifier diode for the
spares box too....

Admitted - a bandsaw in the workshop makes a 15 minute job out of this
transformer hack.(if you don't nick the primary)

Brian W
KA5WOG

Brian, I have done this with old uwave transformers, the ones I could
knock out the magnetic shunt with a hammer and punch, but on the newer
ovens the shunt is a weld. Will these still work?

Jimmie

I left the magnetic bypass leg in the core on the one I used for a spot
weld source. Its a security blanket against core saturation due to the
magnetron,I reckon.

Brian W

thanks for all the comments folks. I considered all and decided to take the learning curve. I was lucky the transformer was only held togethert by a non sticky clear plastic wrap.
Having unwrapped the windings I took the 55v secondary off completely. Re the 20v secs, I worked out the value in volts of each turn, then removed 8 turns from each of the 4 secs. I now have 16.48 volt (near enough to 16.5 volt for me) I had to re-space the secondary windings (i did it to look neat - there may be a sound physics reason too).
Lets hope it works. Before the 'surgery' the 20 volt secondaries were good for 18amp in parallel (according to manufacturer). Tim stated that lowering the volts to 16 will give 25% more current. so I should be good for at least 20 amps. I only wanted 15 so the margin is ok.


thanks again folks Andy

On Thu, 17 Sep 2009 11:55:59 -0500, brian whatcott wrote:

JIMMIE wrote:
On Sep 16, 8:46 pm, brian whatcott wrote:
Andy GD1MIP wrote:
Hi I am building a (240volt in) 13.8 volt (out) PSU ...
the transformer must have a secondary of 16.5 volt.

I have a Talema branded toroid I want to use, its spec is as follows
(from maker)...
73 Andy
Now that the ideas are cooking.... consider this. 1100 watt defunct
microwaves can be had for the asking.
They have a particular transformer core design that makes it easy
to
saw off the 1 KV secondary, and with the generous core window, you
could wind 18 to 20 turns of wire good for 10 amps. The turns per volt
is not much over 1 tpv.... And you have a decent rectifier diode for
the spares box too....

Admitted - a bandsaw in the workshop makes a 15 minute job out of this
transformer hack.(if you don't nick the primary)

Brian W
KA5WOG

Brian, I have done this with old uwave transformers, the ones I could
knock out the magnetic shunt with a hammer and punch, but on the newer
ovens the shunt is a weld. Will these still work?

Jimmie

I left the magnetic bypass leg in the core on the one I used for a spot
weld source. Its a security blanket against core saturation due to the
magnetron,I reckon.

Brian W

I _think_ that shunt serves to limit the secondary current and therefore
protect the magnetron.

I should go through the math on that...

--
www.wescottdesign.com

On Thu, 17 Sep 2009 23:08:23 +0100, Andy GD1MIP wrote:

thanks for all the comments folks. I considered all and decided to take
the learning curve. I was lucky the transformer was only held togethert
by a non sticky clear plastic wrap.
Having unwrapped the windings I took the 55v secondary off completely.
Re the 20v secs, I worked out the value in volts of each turn, then
removed 8 turns from each of the 4 secs. I now have 16.48 volt (near
enough to 16.5 volt for me) I had to re-space the secondary windings (i
did it to look neat - there may be a sound physics reason too). Lets
hope it works. Before the 'surgery' the 20 volt secondaries were good
for 18amp in parallel (according to manufacturer). Tim stated that
lowering the volts to 16 will give 25% more current. so I should be good
for at least 20 amps. I only wanted 15 so the margin is ok.

No no. Tim stated that _after_ you _completely_ rewind the secondaries
with wire that's got 25% more cross-sectional area you can run 25% more
current. if you increased the secondary current to 20A the per-inch
dissipation in the secondary would go up by about 56%; you can probably
run more than 16A in the secondaries now because they're more spread out
and because the primary will be cooler, but I very much doubt that you
can go as high as 20A and maintain the same temperature.

--
www.wescottdesign.com

Tim Wescott wrote:
.... 1100 watt defunct microwaves can be had for the asking.
They have a particular transformer core design that makes it easy
to saw off the 1 KV secondary, and with the
generous core window, you
could wind 18 to 20 turns of wire good for 10 amps. The turns per volt
is not much over 1 tpv.... And you have a decent rectifier diode for
the spares box too....

Admitted - a bandsaw in the workshop makes a 15 minute job out of this
transformer hack.(if you don't nick the primary)

Brian W
KA5WOG

Brian, I have done this with old uwave transformers, the ones I could
knock out the magnetic shunt with a hammer and punch, but on the newer
ovens the shunt is a weld. Will these still work?

Jimmie
I left the magnetic bypass leg in the core on the one I used for a spot
weld source. Its a security blanket against core saturation due to the
magnetron,I reckon.

Brian W

I _think_ that shunt serves to limit the secondary current and therefore
protect the magnetron.

I agree....

Brian W

Andy GD1MIP wrote:
Hi I am building a (240volt in) 13.8 volt (out) PSU along the lines of
this plan..
http://www.warc.org.uk/proj_psu.php it is very similar to the old
Marchwood practical wireless unit from the early 1980's using 2n3055
transistors.

My question. The spec says the transformer must have a secondary of
16.5 volt. Now when I last built a PSU using a similar plan the
transformer output could be much higher, in the region of 15 to 20
volts.

I have a Talema branded toroid I want to use, its spec is as follows
(from maker)
+++++
Toroidal transformer type 5028-P3S03:
Prim: 0-110V-230V-240V orange-yellow-violet-blue
Sec1: 20V/6A green-red
Sec2: 20V/6A black-white
Sec3: 55V/2,5A braun-braun
If you use only the 2x20V you can get out 9A of each winding (no load
on the 55V)
+++++


as you can see the secondary output is 20 volt, now folks do you think
this transformer will suffice at 20 volts, or should I removed a little
winding to bring it down?

73 Andy

PS the variac is just there for scale.




Is this transformer potted in the middle?
If it isn't you could add some turns and connect them in series with the
primary to reduce the secondary voltages. This would be a lot easier
than removing turns from the secondaries, especially if the secondaries
are NOT wound on top!

On Sep 20, 7:43 am, Kenneth Scharf wrote:
Andy GD1MIP wrote:
Hi I am building a (240volt in) 13.8 volt (out) PSU along the lines of
this plan..

OK, we are all being technical here - can I ask if the crowbar IC is
still available? Is the specified relay still available?

At an overvoltage condition,there will be over 200 watts being
dissipated in the 2r2 resistor in series with the SCR - hopefully it
wont instantaneously go open circuit....if the overvoltage protection
works, wont the thing go into cyclic oscillation as it attempts to
restart?....(not sure here, perhaps the gate never goes low enough for
this)
LM723 regulators tend to be a bit unstable - VHF ground plane type
construction is recommended.
The transformer ratings - the manufacturer rates them at a specified
current for a reason. For low duty cycle use, eg SSB, you can get away
with pulling more current. But a solid state Class C FM amplifier? -
also, as mentioned, there is very little headroom at 16v - 18v would
be better.....and thats assuming the mains input voltage is reference
quality - here in OZ, its a design variable, up to 20% fluctuations
are not uncommon, especially at the end of a long run.
Motorola 2N3055 transistors are fine - BUT- when the original Motorola
patents expired, and Motorola stopped making them, a lot of
substandard crappy ones (typically Eastern European) came onto the
market. Be warned, they can be unpredictable. TO3 is a pretty well
obsolete case style anyway. Good luck. And the Motorola symbol was
widely forged as well.

The above, of course, are my Neanderthal rantings on a newsgroup and
will no doubt start an impassioned debate on points of theory , or
maths, or how it was done in 1935. Remember, free advice is worth what
you pay for it.

But I would advise putting a 5w 15V zener across the output terminals,
and before that put a 30amp fuse. Minimal cost for added insurance.

Go for it - build it, experiment. Blowing something up is always a
fantastic spur to learn the correct theory for the application..if it
works first up, then its just a mechanical exercise.

Andrew VK3BFA.