Re: [Electronics_101] Re: Question about a switch and a transformer...

Expand Messages
• ... you are close, but not spot on. If you look at the magnetisation of a transformer, it is the integral of the voltage. So it s not in phase with it and zero
Message 1 of 27 , Apr 1, 2007
On Sat, 31 Mar 2007 19:54:42 +0200, rtstofer <rstofer@...> wrote:

> The problem is inrush current. If the AC line is at its highest
> voltage at the time you turn on the switch, the fast fuse will tend to
> blow. The inrush can be, perhaps, 6 times full load (depends on
> transformer design).
> Richard
>

you are close, but not spot on. If you look at the magnetisation of a
transformer, it is the integral of the voltage. So it's not in phase with
it and zero voltage is actually the worst possible time to switch it on.
If you switch at zero voltage the flux will go up during all of the
halfwave, far into saturation. At the zero voltage moment the flux should
have been driven negative by the previous halfwave, which it hasn't,
because the switch was open. Peak voltage is best, assuming the residual
flux in the transformer is zero. As John already mentioned there can be
some residual flux, which shifts the ideal moment.

It seems too complicated to make a switch that chooses the moment, using a
slower acting fuse or an inrush current limiting circuit (which is usually
used with toroidal transformers) is better. I guess it was described
already, but a NTC works, or even resistors, you should put a relay in
parallel to short this element (to reduce losses and let it cool for
acting again during a power outage). Schematic can be seen on the web, the
relay coil is simply parallel with the transformer.

ST
• ... *snip* ... Thanks Stafan. This has to be the 10,000th time (or more) that you ve offered me very good advice. I put this together, used a 1.5A slow blow
Message 2 of 27 , Apr 1, 2007
--- "Stefan Trethan" <stefan_trethan@...> wrote:
*snip*
> It seems too complicated to make a switch that chooses the moment,
> using a slower acting fuse or an inrush current limiting circuit
> (which is usually used with toroidal transformers) is better. I
> guess it was described already, but a NTC works, or even resistors,
> you should put a relay in parallel to short this element (to reduce
> losses and let it cool for acting again during a power outage).
> Schematic can be seen on the web, the relay coil is simply parallel
> with the transformer.
>
> ST

Thanks Stafan. This has to be the 10,000th time (or more) that
you've offered me very good advice. I put this together, used a 1.5A
slow blow fuse, and all seems to work well. I will add the
Thermistor from N to Hot. I think the relay may be a bit of overkill
for this application. It started out as a VERY simple project.

However, while playing with it, I noticed something important I
overlooked in my planning process: There is a metal mounting bolt
that runs up the middle of the toroid; That sucker gets HOT!!! when
it has been on for a few minutes!

Um... Ok, I feel stupid for missing that typically insignificant
factor (I mean, who ever heard of accounting for heat while
engineering your project?). Solution: I need to cut a small hole in
the box and insert one of my surplus CPU cooling fans.

Questions:
* Where would be the best place to put the fan?
* Should the air be concentrated through the donut hole?
* Should the air go around the outside of the transformer?

The box does not have a lot of excess space above and below the
transformer. However, the corners have several inches of clearance
(the box top has a rectangle shape; it is shaped roughly like a
brick).

The box will be sitting under a table, attached mid way up the table
leg, situated such that the bolt that goes through the donut hole
will be attached to the leg. So, holes in the top should allow heat
out; however, this may not let all the heat escape from the donut
hole.

Final Question:
* Does it seem like this needs a forced air solution, or would a
convection configuration work?

Thanks,

Bob H.
• On Sun, 01 Apr 2007 21:08:19 +0200, Bob Hyland-PMP ... The thermistor goes in series with the primary, not parallel. The relay is required for good design
Message 3 of 27 , Apr 1, 2007
On Sun, 01 Apr 2007 21:08:19 +0200, Bob Hyland-PMP
<Electronics@...> wrote:

> Thermistor from N to Hot. I think the relay may be a bit of overkill
> for this application.

The thermistor goes in series with the primary, not parallel. The relay is
required for good design really, if you need the thermistor, you should
have it. But if the transformer doesn't need the thermistor (not too much
power) then a slower fuse alone is just fine.

> It started out as a VERY simple project.
> However, while playing with it, I noticed something important I
> overlooked in my planning process: There is a metal mounting bolt
> that runs up the middle of the toroid; That sucker gets HOT!!! when
> it has been on for a few minutes!

You haven't closed a conductive path around the transformer (through the
hole), right?

> Um... Ok, I feel stupid for missing that typically insignificant
> factor (I mean, who ever heard of accounting for heat while
> engineering your project?). Solution: I need to cut a small hole in
> the box and insert one of my surplus CPU cooling fans.
> Questions:
> * Where would be the best place to put the fan?
> * Should the air be concentrated through the donut hole?
> * Should the air go around the outside of the transformer?
> The box does not have a lot of excess space above and below the
> transformer. However, the corners have several inches of clearance
> (the box top has a rectangle shape; it is shaped roughly like a
> brick).
> The box will be sitting under a table, attached mid way up the table
> leg, situated such that the bolt that goes through the donut hole
> will be attached to the leg. So, holes in the top should allow heat
> out; however, this may not let all the heat escape from the donut
> hole.
> Final Question:
> * Does it seem like this needs a forced air solution, or would a
> convection configuration work?
> Thanks,
> Bob H.

Depends on the power levels, usually forced air is not needed for
transformers, if it is, there probably isn't enough thermal margin anyway.
Of course in some situations people cut corners (big corners for example
in micrwave ovens) and the transformer needs a fan, but those are
financial reasons not technical, at least for small transformers it should
be cheap enough to make it well large enough.

Convection is probably not unreasonable though. You could also mount the
thing to a large aluminum plate or something as a heatsink (although i'm
in general against attempting to cool transformers this way).

ST
• ... You have a metal brace on the top of the transformer, right? Remove it. What happens is that the bolt and metal top brace makes a closed loop around the
Message 4 of 27 , Apr 1, 2007
--- Bob Hyland-PMP <Electronics@...>
wrote:
>
> However, while playing with it, I noticed something
> important I overlooked in my planning process:
> There is a metal mounting bolt that runs up the
> middle of the toroid; That sucker gets HOT!!! when
> it has been on for a few minutes!

You have a metal brace on the top of the transformer,
right? Remove it.

What happens is that the bolt and metal top brace
makes a closed loop around the transformer, so it has
a lot of current flow and heats up.

Remove the top brace or replace it with something
non-conductive (i.e. plastic) and you won't have to

Jeff Engel

Happiness is - positive intake manifold pressure.

____________________________________________________________________________________
Now that's room service! Choose from over 150,000 hotels
in 45,000 destinations on Yahoo! Travel to find your fit.
http://farechase.yahoo.com/promo-generic-14795097
• On Sun, 01 Apr 2007 22:44:58 +0200, Jeffrey Engel ... The top brace alone is not a problem, it s normal for a toroid to be mounted with a
Message 5 of 27 , Apr 1, 2007
On Sun, 01 Apr 2007 22:44:58 +0200, Jeffrey Engel <jengeltx@...>
wrote:

>
> You have a metal brace on the top of the transformer,
> right? Remove it.
> What happens is that the bolt and metal top brace
> makes a closed loop around the transformer, so it has
> a lot of current flow and heats up.
> Remove the top brace or replace it with something
> non-conductive (i.e. plastic) and you won't have to
> Jeff Engel

The top brace alone is not a problem, it's normal for a toroid to be
mounted with a large washer like thing. What you can't do is close it
around the outside too.

ST
• ... The bolt will be part of a shorted secondary turn if there is conductive path through the hole and around the outside of the core. just having a bolt
Message 6 of 27 , Apr 5, 2007
Bob Hyland-PMP wrote:
> --- "Stefan Trethan" <stefan_trethan@...> wrote:
> *snip*
>> The top brace alone is not a problem, it's normal for a toroid to be
>> mounted with a large washer like thing. What you can't do is close it
>> around the outside too.
>>
>> ST
>
> Hmmm... This makes me think... The box I am using IS metal. Could that
> be closing the loop, so to speak? If so, how could I test to make sure?
> And, if indeed this is the case, what can I use to break the loop?
>
> Indeed, the only thing getting really hot is the bolt. It is too hot to
> touch after only a few seconds. Everything else runs very cool.

The bolt will be part of a shorted secondary turn if there
is conductive path through the hole and around the outside
of the core. just having a bolt through the hole does not
provide a path for current. But if the core is in a metal
box, and the bolt passes through the top and bottom of that
box, and also through the hole in the core, then the box and
bolt encircle the flux in the core, and the changes in that
flux will drive current through that short.
• ... *snip* ... Hmmm... This makes me think... The box I am using IS metal. Could that be closing the loop, so to speak? If so, how could I test to make sure?
Message 7 of 27 , Apr 5, 2007
--- "Stefan Trethan" <stefan_trethan@...> wrote:
*snip*
>
> The top brace alone is not a problem, it's normal for a toroid to be
> mounted with a large washer like thing. What you can't do is close it
> around the outside too.
>
> ST

Hmmm... This makes me think... The box I am using IS metal. Could that
be closing the loop, so to speak? If so, how could I test to make sure?
And, if indeed this is the case, what can I use to break the loop?

Indeed, the only thing getting really hot is the bolt. It is too hot to
touch after only a few seconds. Everything else runs very cool.

Bob H.
• Bob, How about using a nylon bolt? Regards, Ross McKenzie ValuSoft Melbourne Australia
Message 8 of 27 , Apr 5, 2007
Bob,

How about using a nylon bolt?

Regards,

Ross McKenzie
ValuSoft
Melbourne Australia
• ... *snip* ... John: Thanks for the comments. That makes sense to me. Is it enough to break the current flow? In other words, if I put a rubber stopper at the
Message 9 of 27 , Apr 5, 2007
--- John Popelish <jpopelish@...> wrote:
*snip*
> The bolt will be part of a shorted secondary turn if there
> is conductive path through the hole and around the outside
> of the core. just having a bolt through the hole does not
> provide a path for current. But if the core is in a metal
> box, and the bolt passes through the top and bottom of that
> box, and also through the hole in the core, then the box and
> bolt encircle the flux in the core, and the changes in that
> flux will drive current through that short.

John: Thanks for the comments. That makes sense to me.

Is it enough to break the current flow? In other words, if I put a
rubber stopper at the top and bottom of the bolt (both places it makes
contact with the box), would that be sufficient?

Thanks again!

Bob H.
• ... From: Bob Hyland-PMP To: Sent: Friday, April 06, 2007 3:26 AM Subject: [Electronics_101]
Message 10 of 27 , Apr 5, 2007
----- Original Message -----
From: "Bob Hyland-PMP" <Electronics@...>
To: <Electronics_101@yahoogroups.com>
Sent: Friday, April 06, 2007 3:26 AM
Subject: [Electronics_101] Re: ANOTHER Question about a switch and a
transformer...

> --- "Stefan Trethan" <stefan_trethan@...> wrote:
> *snip*
>>
>> The top brace alone is not a problem, it's normal for a toroid to be
>> mounted with a large washer like thing. What you can't do is close it
>> around the outside too.
>>
>> ST
>
> Hmmm... This makes me think... The box I am using IS metal. Could that
> be closing the loop, so to speak? If so, how could I test to make sure?
> And, if indeed this is the case, what can I use to break the loop?
>
> Indeed, the only thing getting really hot is the bolt. It is too hot to
> touch after only a few seconds. Everything else runs very cool.

Use a nylon screw.

Leon
• ... From: Bob Hyland-PMP To: Sent: Friday, April 06, 2007 4:09 AM Subject: [Electronics_101]
Message 11 of 27 , Apr 5, 2007
----- Original Message -----
From: "Bob Hyland-PMP" <Electronics@...>
To: <Electronics_101@yahoogroups.com>
Sent: Friday, April 06, 2007 4:09 AM
Subject: [Electronics_101] Re: ANOTHER Question about a switch and a
transformer...

> --- John Popelish <jpopelish@...> wrote:
> *snip*
>> The bolt will be part of a shorted secondary turn if there
>> is conductive path through the hole and around the outside
>> of the core. just having a bolt through the hole does not
>> provide a path for current. But if the core is in a metal
>> box, and the bolt passes through the top and bottom of that
>> box, and also through the hole in the core, then the box and
>> bolt encircle the flux in the core, and the changes in that
>> flux will drive current through that short.
>
> John: Thanks for the comments. That makes sense to me.
>
> Is it enough to break the current flow? In other words, if I put a
> rubber stopper at the top and bottom of the bolt (both places it makes
> contact with the box), would that be sufficient?

You only need one, to avoid a shorted turn.

Leon
• Bob Hyland-PMP wrote: *snip* ... For kicks, I drilled out a bigger hole at the exit point, and inserted a plug of non-conducting material
Message 12 of 27 , Apr 5, 2007
"Bob Hyland-PMP" <Electronics@...> wrote:
*snip*
> Is it enough to break the current flow? In other words, if I put a
> rubber stopper at the top and bottom of the bolt (both places it
> makes contact with the box), would that be sufficient?
>
> Thanks again!
>
> Bob H.

For kicks, I drilled out a bigger hole at the exit point, and
inserted a plug of non-conducting material that had a predrilled hole
that fit the bolt nicely. It was a spacer of some sort that I had in
my collection. It is not rubber (could not finda rubber stopper);
rather it is more like a hard plastic material.

Anyway, inserted this in one end, turned on the unit, ran it for 30
minutes or so, and... the bolt was cool as a cucumber!

Thanks everyone! Couldn't have done it without you!

Bob H.
• ... Any break in the loop, anywhere, is sufficient. Just remember that the here will be some voltage across the insulated gap, so you will probably want to
Message 13 of 27 , Apr 6, 2007
Bob Hyland-PMP wrote:

> John: Thanks for the comments. That makes sense to me.
>
> Is it enough to break the current flow? In other words, if I put a
> rubber stopper at the top and bottom of the bolt (both places it makes
> contact with the box), would that be sufficient?

Any break in the loop, anywhere, is sufficient. Just
remember that the here will be some voltage across the
insulated gap, so you will probably want to protect the
metal on both sides of the gap from an accidental short by
any metal coming in contact with both sides. Replacing the
bolt with a nylon one is an easy solution.
Your message has been successfully submitted and would be delivered to recipients shortly.