Practically making a very low power 60 Hz transformer

[Ignoramus7428]

I am trying to find the easiest way of making a very low power
transformer. Its job will be to convert 240V to some voltage such as
100V (exact value does not matter). This voltage will have an only use
for voltage sensing, and several milliamps should be quite enough.

So... With that in mind... What is the easiest design for a
transformer. What is important is that

1) it does not get too hot and
2) it could produce a few milliamps.

How about:


1) using a steel rod such as a very thick nail, and two windings
2) using a ferrite toroid (I read unfavorable comments about doing that at
60 Hz).
3) using a few washers stacked together for a steel toroid
4) using a little toroidal cutout from a steel pipe

Anything else that is quick and dirty?

i

 

[Boris Mohar]

I am trying to find the easiest way of making a very low power
transformer. Its job will be to convert 240V to some voltage such as
100V (exact value does not matter). This voltage will have an only use
for voltage sensing, and several milliamps should be quite enough.

So... With that in mind... What is the easiest design for a
transformer. What is important is that

1) it does not get too hot and
2) it could produce a few milliamps.

How about:


1) using a steel rod such as a very thick nail, and two windings
2) using a ferrite toroid (I read unfavorable comments about doing that at
60 Hz).
3) using a few washers stacked together for a steel toroid
4) using a little toroidal cutout from a steel pipe

Anything else that is quick and dirty?

i

There must be a optocoupler that would do the job. How about
http://www.fairchildsemi.com/ds/HC/HCPL-3700.pdf




Regards,

Boris Mohar

Got Knock? - see:
Viatrack Printed Circuit Designs (among other things) http://www.viatrack.ca

void _-void-_ in the obvious place

 

[John Larkin]

I am trying to find the easiest way of making a very low power
transformer. Its job will be to convert 240V to some voltage such as
100V (exact value does not matter). This voltage will have an only use
for voltage sensing, and several milliamps should be quite enough.

So... With that in mind... What is the easiest design for a
transformer. What is important is that

1) it does not get too hot and
2) it could produce a few milliamps.

How about:


1) using a steel rod such as a very thick nail, and two windings
2) using a ferrite toroid (I read unfavorable comments about doing that at
60 Hz).
3) using a few washers stacked together for a steel toroid
4) using a little toroidal cutout from a steel pipe

1, 3, and 4 would be *very* lossy and would probably fry. 2 would
work, given high-mu material and *lots* of turns of wire. Ohmic losses
might be nasty, which is why 60 Hz transformers are usually laminated
silicon steel (or more exotic stuff), seldom ferrites.

Transformer design is complex and fabrication is difficult; try
another way, like a resistive or capacitive voltage divider, or buy a
standard transformer.

Got any details on the application?

John

 

[Ignoramus7428]

1, 3, and 4 would be *very* lossy and would probably fry. 2 would
work, given high-mu material and *lots* of turns of wire. Ohmic losses
might be nasty, which is why 60 Hz transformers are usually laminated
silicon steel (or more exotic stuff), seldom ferrites.

Transformer design is complex and fabrication is difficult; try
another way, like a resistive or capacitive voltage divider, or buy a
standard transformer.

Got any details on the application?

Sure.

I built this phase converter:

http://igor.chudov.com/projects/17.5-Phase-Converter/

it works fine. Actually quite nice, it is easy starting and quiet and
balanced.

I want to add some bells and whistles to it. So, I bought a K3NX-AD1C
meter on ebay.

http://oeiwcs.omron.com/webapp/comm.../itemdisp.d2w/report?prmenbr=316&prrfnbr=4127

I want to make it measure various things such as voltage between all 3
legs of 3 phase power, AC currents, etc.

One way suggested was with a differential amplifier like AD629. I have
some doubts about using them, perhaps out of ignorance, as I am trying
to measure with reference points of 240 VAC and that could be too much
for it. If I am wrong, I will be happy to use them.

So, I was wondering if I can convert AC to AC with a different
reference point (with a xfmr), rectify it, and use a little divider to
get a small DC potential to go into the meter.

i

 

[Jack W.]

I am trying to find the easiest way of making a very low power
transformer. Its job will be to convert 240V to some voltage such as
100V (exact value does not matter). This voltage will have an only use
for voltage sensing, and several milliamps should be quite enough.

So... With that in mind... What is the easiest design for a
transformer. What is important is that

1) it does not get too hot and
2) it could produce a few milliamps.

How about:


1) using a steel rod such as a very thick nail, and two windings
2) using a ferrite toroid (I read unfavorable comments about doing that at
60 Hz).
3) using a few washers stacked together for a steel toroid
4) using a little toroidal cutout from a steel pipe

Anything else that is quick and dirty?

I once used a small transformer for interstage audio tube operation. Quite
small actually, fit in my hand (not like the output transformers...). The
impedance was few kilo-ohms, DC resistance still quite high. Didn't complain
at all when connected to the main because of the so high impedance, it run
totally cool as if no power was connected to it. I used it for sensing
voltage with good galvanic isolation.
Just my idea...

Jack

 

[Adrian Jansen]

I am trying to find the easiest way of making a very low power
transformer. Its job will be to convert 240V to some voltage such as
100V (exact value does not matter). This voltage will have an only use
for voltage sensing, and several milliamps should be quite enough.

So... With that in mind... What is the easiest design for a
transformer. What is important is that

1) it does not get too hot and
2) it could produce a few milliamps.

How about:


1) using a steel rod such as a very thick nail, and two windings
2) using a ferrite toroid (I read unfavorable comments about doing that at
60 Hz).
3) using a few washers stacked together for a steel toroid
4) using a little toroidal cutout from a steel pipe

Anything else that is quick and dirty?

i

Just get any small wall-wart transformer, but make sure its AC output,
ie just a transformer and not a switch-mode type with DC out. You can
usually tell by the weight - transformers are heavier. Even the older
DC types can be used if you can rip them apart just for the transformer.
There must be heaps of these lying around in junk piles. At least
with a wall-wart the input side is mains rated and relatively safe, much
better than you can do by winding your own.

--
Regards,

Adrian Jansen adrianjansen at internode dot on dot net
Design Engineer J & K Micro Systems
Microcomputer solutions for industrial control
Note reply address is invalid, convert address above to machine form.

 

[Joerg]

I built this phase converter:

http://igor.chudov.com/projects/17.5-Phase-Converter/

it works fine. Actually quite nice, it is easy starting and quiet and
balanced.

I want to add some bells and whistles to it. So, I bought a K3NX-AD1C
meter on ebay.

With such big equipment why on earth does the transformer have to be so
tiny? Just get a small 240V to 120V transformer plus a fuse, drill four
holes into your chassis and mount it.

Regards, Joerg

 

[Rich Grise]

With such big equipment why on earth does the transformer have to be so
tiny? Just get a small 240V to 120V transformer plus a fuse, drill four
holes into your chassis and mount it.

Regards, Joerg

He needs three, but Mouser has little transformers that are quite
reasonable, albeit I don't immediately see a 240/120 unit.
http://www.mouser.com

Good Luck!
Rich

 

[Tim Wescott]

Sure.

I built this phase converter:

http://igor.chudov.com/projects/17.5-Phase-Converter/

it works fine. Actually quite nice, it is easy starting and quiet and
balanced.

I want to add some bells and whistles to it. So, I bought a K3NX-AD1C
meter on ebay.

http://oeiwcs.omron.com/webapp/comm.../itemdisp.d2w/report?prmenbr=316&prrfnbr=4127

I want to make it measure various things such as voltage between all 3
legs of 3 phase power, AC currents, etc.

One way suggested was with a differential amplifier like AD629. I have
some doubts about using them, perhaps out of ignorance, as I am trying
to measure with reference points of 240 VAC and that could be too much
for it. If I am wrong, I will be happy to use them.

You can resistively divide your 240 VAC down to something that you can
safely apply to the differential amplifiers, then you can do any signal
processing you want on the outputs of the diff. amps. You will want to
reference your differential amplifiers to ground, which means that you
won't be quite isolated, but you're able to use really high resistance
sense resistors so it's probably not a big deal.

Note that if I were doing this for $$$ I'd be paying _close_ attention
to electrical codes _and_ I'd probably want to subcontract some
consulting to an expert. In a production environment I'd probably want
to make something that's totally isolated, so I'd do something like
power a board off of the AC, measure the AC, convert the measurements to
pulse trains with V/F converters, optoisolate the results, then convert
the (isolated) pulse trains with F/V converters (or measure the pulse
trains with a counter built into a microcontroller). This would be
cheaper than transformers if you were making 1000 boards a year, but for
a 1-off it'd be a waste of engineering effort.

So, I was wondering if I can convert AC to AC with a different
reference point (with a xfmr), rectify it, and use a little divider to
get a small DC potential to go into the meter.

Yes, that should work. Do a web search for "precision rectifier" --
just a diode bridge will introduce nasty errors. Keep in mind that a
power-grade transformer will be optimized for cost and power transfer,
_not_ for accuracy. You can expect that the output of such a beast will
be phase shifted and probably show saturation effects; I wouldn't expect
more than 10% accuracy out of a power transformer and if I were under
contract I wouldn't guarantee _anything_.

In this case, however, 10% accuracy is probably all you really need, and
it's certainly all I'd look for in an economy sensing unit.

 

[Ignoramus21405]

You can resistively divide your 240 VAC down to something that you can
safely apply to the differential amplifiers,

Sure, reducing the difference is not a problem. However, the
difference would still referenced to phase voltage.

Think about trying to measure voltage between leg 1 and leg 3.

then you can do any signal processing you want on the outputs of the
diff. amps. You will want to reference your differential amplifiers
to ground, which means that you won't be quite isolated, but you're
able to use really high resistance sense resistors so it's probably
not a big deal.

Note that if I were doing this for $$$ I'd be paying _close_ attention
to electrical codes _and_ I'd probably want to subcontract some
consulting to an expert. In a production environment I'd probably want
to make something that's totally isolated, so I'd do something like
power a board off of the AC, measure the AC, convert the measurements to
pulse trains with V/F converters, optoisolate the results, then convert
the (isolated) pulse trains with F/V converters (or measure the pulse
trains with a counter built into a microcontroller). This would be
cheaper than transformers if you were making 1000 boards a year, but for
a 1-off it'd be a waste of engineering effort.

Yes, that should work. Do a web search for "precision rectifier" --
just a diode bridge will introduce nasty errors. Keep in mind that a
power-grade transformer will be optimized for cost and power transfer,
_not_ for accuracy. You can expect that the output of such a beast will
be phase shifted and probably show saturation effects; I wouldn't expect
more than 10% accuracy out of a power transformer and if I were under
contract I wouldn't guarantee _anything_.

In this case, however, 10% accuracy is probably all you really need, and
it's certainly all I'd look for in an economy sensing unit.

That makes good sense. I will think some more about it. I think that I
can get away with using a differential amplifier if I reference
voltages to legs 1 and 2 (two hots of the 220V electrical system),
which are only 1220V away from ground.

i

 

[[email protected]]

So... With that in mind... What is the easiest design for a

transformer. What is important is that

1) it does not get too hot and
2) it could produce a few milliamps.

Well, if you want it small, maybe use a little audio transformer from a
pocket radio and add a 0.1uF cap in series. You will get a few
milliamps from the secondary, but I don't know what the voltage would
be.

-Bill

 

[Joerg]

Hello Rich,

He needs three, but Mouser has little transformers that are quite
reasonable, albeit I don't immediately see a 240/120 unit.
http://www.mouser.com

You can buy little transformers that operate EU gear on US power and
vice versa. Usually cheap, <$30. The nice thing is they are enclosed
which can be a good thing for such workshop equipment.

Regards, Joerg

 

[Ignoramus21405]

Hello Rich,

You can buy little transformers that operate EU gear on US power and
vice versa. Usually cheap, <$30. The nice thing is they are enclosed
which can be a good thing for such workshop equipment.

I am eyeing some transformers that may cost me 60 cents apiece... But
I need to buy 50.

i

Regards, Joerg

--

 

[Joerg]

I am eyeing some transformers that may cost me 60 cents apiece... But
I need to buy 50.

So what if you'd sell the remaining 47 of them for 99c on ebay? Then
each of your three transformers would have cost you -$5.50 ... 8-D

Regards, Joerg

 

[Ignoramus24987]

So what if you'd sell the remaining 47 of them for 99c on ebay? Then
each of your three transformers would have cost you -$5.50 ... 8-D

Yes... You are right... I bought 27 Semikron 240A diodes for $4 apiece
on ebay and m selling them now in sets of 4 at $15 apiece. With these
transformers though, it is too much hassle.

i

 

[JosephKK]

Sure.

I built this phase converter:

http://igor.chudov.com/projects/17.5-Phase-Converter/

it works fine. Actually quite nice, it is easy starting and quiet and
balanced.

I want to add some bells and whistles to it. So, I bought a K3NX-AD1C
meter on ebay.

http://oeiwcs.omron.com/webapp/comm.../itemdisp.d2w/report?prmenbr=316&prrfnbr=4127

I want to make it measure various things such as voltage between all 3
legs of 3 phase power, AC currents, etc.

One way suggested was with a differential amplifier like AD629. I have
some doubts about using them, perhaps out of ignorance, as I am trying
to measure with reference points of 240 VAC and that could be too much
for it. If I am wrong, I will be happy to use them.

So, I was wondering if I can convert AC to AC with a different
reference point (with a xfmr), rectify it, and use a little divider to
get a small DC potential to go into the meter.

i

selecting a DC current process meter for measuring AC voltages and currents
is a rather poor choice. generally for that task the selection is an AC
voltage / current process meter. then the switching and scaling is
relatively easy.