[John L Stewart]

Ever wonder how you could make accurate measurements of small resistance’s common in audio or power transformer windings? An ordinary Volt-Ohm-Meter (VOM) or Digital Multimeter (DMM) does a poor job. Your meters leads & contact resistance’s become a large part of the result. And the one Ohm mark on your analogue meter doesn’t leave much room to guesstimate the result.

For these kinds of measurements the 4-terminal method provides good results. But you will need some kind of Constant Current Source (CCS) with which to drive the Device Under Test (DUT). The test current from the CCS is applied to the DUT leads while a voltmeter reads the result on these same DUT leads. That way meter lead & contact resistances are eliminated. Accuracy of the result is insured.

A single VOM or DMM are all the measurement equipment required. The meter should have a low voltage scale, say one volt FS or less. That is now commonly available on low cost DMM’s.

I used a 12 volt automotive battery as my source. But any 12 volt battery of sufficient capacity would work OK. If the battery has been on charge it should be left for a few hours to stabilize.

The CCS is nothing more than the widely available automotive 1157 bulb. The bulb has two filaments, so two measurement ranges are possible. A 5 Amp fuse in the battery lead is also a good idea.

The meter is first set to its 10 amp scale. Then connect the meter, 12 volt battery, fuse, 1157 bulb & DUT all in series. The meter will indicate the total current thru the circuit. Under normal conditions the 1157 bulb will dissipate more than 20 watts. That is one reason I chose the bulb rather then looking for an equivalent 50 watt resister.

The meter is then disconnected & set to measure volts on a low voltage range. The circuit above is then reconnected as before without the ammeter in place. However the current will be altered only slightly from that measured since the ammeter voltage drop including its leads is only about 200 mv out of the 12 volts, less than 2%. Now connect the meter leads (set to volts) to the DUT leads. Measure the resulting voltage on the DUT leads & apply Ohm’s Law.

The 1157 bulbs large heater will apply about 2 Amps to the DUT. The small heater of about 0.5 Amp can be used on DUT windings of higher resistance. The other reason I chose an incandescent bulb rather than a power resister is for its dynamic resistance at its operating point in this application. While the large heater has a dynamic resistance of more than 10 ohms at 12 volts the resistor would still be about 5.5 ohms for these conditions. That helps to keep the test current constant while changing meter leads.

I built a much more elegant CCS in 1969 for doing these 4-terminal tests. But that is another story. You can read about it in AudioXpress July 2001.

Cheers, John

[Phil Allison[_3_]]

"John Lunatic Stewart"

Ever wonder how you could make accurate measurements of small
resistance's common in audio or power transformer windings? An ordinary
Volt-Ohm-Meter (VOM) or Digital Multimeter (DMM) does a poor job.


** All you need is a 3.5 digit DMM and a 2.2 ohm WW 5 watt resistor PLUS a
regulated, adjustable voltage PSU capable of 1 amp DC. Connect the 2.2 ohm
resistor and winding in series and energise the pair with the bench PSU -
set the voltage to about 2 volts.

The ratio of voltages across the winding and the resistor will be the same
as the ratio of their resistances.

5th grade math will suffice - and a pocket calculator.


..... Phil



..... Phil

[John L Stewart]

[quote='Phil Allison[_3_];927190']"John Lunatic Stewart"

Ever wonder how you could make accurate measurements of small
resistance's common in audio or power transformer windings? An ordinary
Volt-Ohm-Meter (VOM) or Digital Multimeter (DMM) does a poor job.


** All you need is a 3.5 digit DMM and a 2.2 ohm WW 5 watt resistor PLUS a
regulated, adjustable voltage PSU capable of 1 amp DC. Connect the 2.2 ohm
resistor and winding in series and energise the pair with the bench PSU -
set the voltage to about 2 volts.

The ratio of voltages across the winding and the resistor will be the same
as the ratio of their resistances.

5th grade math will suffice - and a pocket calculator.


..... Phil


What fun it is to have Phil checking my posts again! I thought for a while the Law of Gravity had been repealed south of the equator & Phil had dropped off. Be careful Phil, there are high tides tonight at 20:46 GMT. The Astrologers will be out looking for you.

But thankyou Phil for your carefully considered response. You may go to the head of the class now.

Happy to note you are still on top of your game, whatever that might be. And you have hit the nail squarely on the head as usual. Perhaps the solution I offered was somewhat off track for RAT.

However, I did this for a guy who is not involved with electronics. Was looking for a way to check stator resistance in a 3-phase alternator on a motorcycle. He doesn’t have a box of 5 ohm resistors of sufficient power, a regulated PS, Etc. But he does have a DMM, a 12 volt battery & lots of 1157 bulbs.

All looked interesting so it is posted for the curious. Others may leave the room, this is not required reading. There will be no test on this post.

Cheers to all, John

[Patrick Turner]

On Mar 4, 2:29*pm, John L Stewart John.L.Stewart.
wrote:
Ever wonder how you could make accurate measurements of small
resistance s common in audio or power transformer windings? An ordinary
Volt-Ohm-Meter (VOM) or Digital Multimeter (DMM) does a poor job. Your
meters leads & contact resistance s become a large part of the result.
And the one Ohm mark on your analogue meter doesn t leave much room to
guesstimate the result.

For these kinds of measurements the 4-terminal method provides good
results. But you will need some kind of Constant Current Source (CCS)
with which to drive the Device Under Test (DUT). The test current from
the CCS is applied to the DUT leads while a voltmeter reads the result
on these same DUT leads. That way meter lead & contact resistances are
eliminated. Accuracy of the result is insured.

A single VOM or DMM are all the measurement equipment required. The
meter should have a low voltage scale, say one volt FS or less. That is
now commonly available on low cost DMM s.

I used a 12 volt automotive battery as my source. But any 12 volt
battery of sufficient capacity would work OK. If the battery has been on
charge it should be left for a few hours to stabilize.

The CCS is nothing more than the widely available automotive 1157 bulb.
The bulb has two filaments, so two measurement ranges are possible. A 5
Amp fuse in the battery lead is also a good idea.

The meter is first set to its 10 amp scale. Then connect the meter, 12
volt battery, fuse, 1157 bulb & DUT all in series. The meter will
indicate the total current thru the circuit. Under normal conditions the
1157 bulb will dissipate more than 20 watts. That is one reason I chose
the bulb rather then looking for an equivalent 50 watt resister.

The meter is then disconnected & set to measure volts on a low voltage
range. The circuit above is then reconnected as before without the
ammeter in place. However the current will be altered only slightly from
that measured since the ammeter voltage drop including its leads is only
about 200 mv out of the 12 volts, less than 2%. Now connect the meter
leads (set to volts) to the DUT leads. Measure the resulting voltage on
the DUT leads & apply Ohm s Law.

The 1157 bulbs large heater will apply about 2 Amps to the DUT. The
small heater of about 0.5 Amp can be used on DUT windings of higher
resistance. The other reason I chose an incandescent bulb rather than a
power resister is for its dynamic resistance at its operating point in
this application. While the large heater has a dynamic resistance of
more than 10 ohms at 12 volts the resistor would still be about 5.5 ohms
for these conditions. That helps to keep the test current constant while
changing meter leads.

I built a much more elegant CCS in 1969 for doing these 4-terminal
tests. But that is another story. You can read about it in AudioXpress
July 2001.

Cheers, John

--
John L Stewart

I would have thought it wise to never subject anything which is a low
resistance to any serious current, certainly not 2Amps.

2 amps would be enough to easily fry or fuse many things such as tiny
RF coils with low resistance windings but very fine wire.

So IMHO, a 10Vdc source with a 100 ohm series resistor will generate
100mAdc, and be fairly safe to apply to anything with R 10 ohms.

One can also use 100Vdc and a 1k0 x 10W R to get a more constant
current of 0.1Amps.


If the low resistance was say 0.01 ohms, then the Vdc across the 0.01r
= 0.001V, or 1mV,

My hand held Fluke has a low range for mV dc, and could measure 1mV
well enough.

But most low resistances I might need to measure accurately would be
above 0.05 ohms and with 0.1A the voltage across the R would read
0.005Vdc, very easy to measure.
One could even make a bridge with the CCS & unknown R and the two
higher more easily measurable but variable resistances, and then
measure where the voltage nulls across the bridge and then make one's
calculations as one might.

For most coils of wire on transformers I will measure the wire dia,
and count the turns and estimate the turn length, apply a formula and
calculate the R.
Usually there are many other things I may want to know about a given
transformer to establish just what is in it if I have no data for it.
I went through a pile of over 100 transformers recently, and one gets
faster at working out what's in them.

Patrick Turner.

[Phil Allison[_3_]]

"John Lunatic Stewart"

** All you need is a 3.5 digit DMM and a 2.2 ohm WW 5 watt resistor PLUS
a
regulated, adjustable voltage PSU capable of 1 amp DC. Connect the 2.2
ohm
resistor and winding in series and energise the pair with the bench PSU
-
set the voltage to about 2 volts.

The ratio of voltages across the winding and the resistor will be the
same
as the ratio of their resistances.

5th grade math will suffice - and a pocket calculator.

** Snip smartarse ****e.


But thankyou Phil for your carefully considered response. You may go to
the head of the class now.

** Been that for nearly all my school career.


Happy to note you are still on top of your game, whatever that might be.
And you have hit the nail squarely on the head as usual. Perhaps the
solution I offered was somewhat off track for RAT.

** It's the method I regularly use for measuring the primary ohms of large
power transformers.

Not just at room temp, but also when they are at full operating temp too -
cos the ratio of the two resistances allows the internal temp of the tranny
to be calculated with good accuracy.

The fractional increase in resistance is simply multiplied by 245 to get the
actual temp rise of the copper - which is the thing that matters most.

Eg:

The resistance increases by 25% = 0.25

So the rise in temp is 245 x 0.25 = 61 degrees C


.... Phil

[Patrick Turner]

Snip....


** It's the method I regularly use for measuring the primary ohms of large
power transformers.

Not just at room temp, but also when they are at full operating temp too *-
cos the ratio of the two resistances allows the internal temp of the tranny
to be calculated with good accuracy.

The fractional increase in resistance is simply multiplied by 245 to get the
actual temp rise of the copper - which is the thing that matters most.

Eg:

The resistance increases by 25% *= 0.25

So the rise in temp is 245 x 0.25 * = 61 degrees C

Very interesting.

Seen this site folks?

http://www.cirris.com/testing/temperature/copper.html

They say....

The Temperature Coefficient of Copper (near room temperature) is
+0.393 percent per degree C. This means if the temperature increases
1°C the resistance will increase 0.393%.

Let us suppose we had a large transformer at room temp. Suppose the
input VA power was 240, so if input voltage = 240V, input current = 1
amp.

This means also the load looking into the tranny primary = 240/1 = 240
ohms.

Let us suppose the primary Rw was 2.5% of the load = 6 ohms.

Power in heat in wire = I squared x R = 6 watts.

This may cause a rise in temp of 20C above room temp depending on how
big the tranny is and how well it gets ris of its winding heat and the
varnishng and if it a toroidal or not or if potted or not etc. Temp is
also affected by core losses.
There is a positive FB effect here, because as the copper warms the I
squared x R calculation increases the heating effect

If there was a rise in temp of +61 degrees, the resistance would have
risen by a factor of 61 x 0.393 percent = 23.973%, or to very close to
what Phil calculated.
So resistance at +61C rise above room temp means R = 7.5 ohms.
So Pd at 1 amp = 7.5Watts. The amount of heating power increase isn't
huge although a rise of 61C in any transformer would be of great
concern because the transfrmer would be stinking hot.

But who could predict from calculations so far if there is any easy
mathematical relationship between copper temp and actual transformer
temperature after say 4 hours?

If a transformer has been designed about right and which uses low loss
GOSS core than the T rise might only be 10C or say to 35C if room temp
was 25C.
Methinks one would have to seriously over load the transformer to get
it to heat by an additional 51C, and a fuse would blow first.

One must just use thick enough wire to get Rw low enough and one rule
of thumb is that current density be no more tham 3 amps per aquare
millimeter of copper wire sectional area.

I like to see single ended OPT primaries with no more than 2A/sq.mm at
the idle condition. Secondaries may be 3A/sq mm.

Patrick Turner

[John L Stewart]

------------------------------------------------------------

Having read your responses carefully, as a former instructor of electronics, physics & many aspects of electronics, I can advise both have failed, for the following reasons.

1) Did not read or understand the original application. I thought the description was pretty good.

2) Failed to notice the 1157 bulb has two heaters, all noted in the original post. So operation at 2A or 0.5A is possible.

3) Patrick, where did RF Chokes come up in the original application?

4) The application as first noted is for low resistance windings on transformers, ie heater, 2A & up in my limited experience. And OPT secondaries.

5) Power still equals I (squared) * R. A relatively small Hammond 125E at 15 watts sees 1.37 A into 8R, so no problem with this solution.

6) Later I made the point of the 3-phase stator in an alternator as being the users primary application.

7) Patrick has a DMM that wil read one mv, but at what resolution? BTW, mine has a resolution of one mivrovolt, but no good to the guy who needs to make the measurement. I have two of them plus several other analogue & digital instruments.

http://www.gossenmetrawatt.com/engli...etrahit29s.htm

8) Patrick launched into a discussion of his transformer winding preferences. What has that got to do with the applications noted?

9) A bridge is always nice. But how does that solve the users problem? Do you recommend Kelvin or Wheatstone & why?

FYI, I did work in the test department at Ferranti for a couple of years. Most everything was 3-phase, the kind you take away on a railway flatcar. Temp rise is a serious problem on anything that size. There was oil cooling & in those days PVC (fireproof)....{should have said PCB}.... And lots of fans on the oil radiators. Also glass & porcelain insulations on some air cored monsters built for use inside buildings.

Here is something I'm sure Phil is aware of if he has done any heavy duty wiring- 1000 ft of #10 Copper wire at 20C is close on one ohm.

Go up 3 sizes to #13 & we have 2R.

Go down 3 sizes & we have 0.5R.

The 110/220 running into our house is 500 ft of 3-0 copper. Cost like hell even 40 years ago.

In the real world people look for short, concise answers, in particular in sales. If my customer asks if the radio test set I'm trying to sell can do a cross band repeater test he needs an answer now, not a 4 page story. That sure works for me. And it made me a lot of money after I shutup!

Cheers to all, John

[Big Bad Bob]

On 03/03/11 19:29, John L Stewart wrote:
Ever wonder how you could make accurate measurements of small
resistance’s common in audio or power transformer windings?

I would build a "half bridge" of sorts, maybe a 10 ohm resistor in series
with the winding, with a very low voltage DC power supply (1v) going through
the two resistors. Then you'd measure the voltage across the 10 ohm
resistor directly on the component, and the voltage across the test
resistance directly on the component, using a decent digital multimeter that
can measure very low voltages (down to 1 millivolt, for example). Then you
make sure your 10 ohm resistor is accurate (maybe purchase a 1% or 0.5% or
measure it with a multimeter first). Then you can calculate the test
resistance by the voltage drop ratio. A DVM that can measure 1 millivolt
will be able to measure ~0.01 ohms with 1V applied to the circuit. You'll
have about 100 ma through the circuit, but virtually NONE through the meter.

To measure smaller resistance values you could even build a simple DC
amplifier with an op amp, and after zeroing and calibrating it, you read the
output of the op amp on a DVM. You could build it on a breadboard from
spare parts, even.

[Patrick Turner]

On Mar 6, 8:17*am, Big Bad Bob BigBadBob-at-mrp3-
wrote:
On 03/03/11 19:29, John L Stewart wrote:

Ever wonder how you could make accurate measurements of small
resistance s common in audio or power transformer windings?

I would build a "half bridge" of sorts, maybe a 10 ohm resistor in series
with the winding, with a very low voltage DC power supply (1v) going through
the two resistors. *Then you'd measure the voltage across the 10 ohm
resistor directly on the component, and the voltage across the test
resistance directly on the component, using a decent digital multimeter that
can measure very low voltages (down to 1 millivolt, for example). *Then you
make sure your 10 ohm resistor is accurate (maybe purchase a 1% or 0.5% or
measure it with a multimeter first). *Then you can calculate the test
resistance by the voltage drop ratio. *A DVM that can measure 1 millivolt
will be able to measure ~0.01 ohms with 1V applied to the circuit. *You'll
have about 100 ma through the circuit, but virtually NONE through the meter.

To measure smaller resistance values you could even build a simple DC
amplifier with an op amp, and after zeroing and calibrating it, you read the
output of the op amp on a DVM. *You could build it on a breadboard from
spare parts, even.

Your idea sounds like my approach using 10Vdc and a 100 ohm series R
to generate a virtually constant current of 100mAdc

But a 13V winding from some old transformer might be more likely to be
in your junk box and if used in a simple doubler rectifier you get
aboout +33Vdc, so R would be 330 ohms.

Measurement of 0.01 ohm gives only +1mVdc, but an opamp can be made to
give accurate gain = 100 at dc so its easy to measure without a
bridge, although a bridge was the traditional way. I suggest an OPA134
or its dual version, OPA234 is an excellent opamp with high input
impedance for such a measurement. A meter can be built into the NFB
loop. Then one may as well build your own multimeter. Oops, another
couple of sundays are doomed!

Patrick Turner.

[Big Bad Bob]

On 03/05/11 14:16, Patrick Turner wrote:
Measurement of 0.01 ohm gives only +1mVdc, but an opamp can be made to
give accurate gain = 100 at dc so its easy to measure without a
bridge, although a bridge was the traditional way. I suggest an OPA134
or its dual version, OPA234 is an excellent opamp with high input
impedance for such a measurement. A meter can be built into the NFB
loop. Then one may as well build your own multimeter. Oops, another
couple of sundays are doomed!

microcontroller?

one thing worthy of mention - no regulation is needed as long as one of the
two resistors is known. You then calculate the unknown resistor using the
voltage drop ratio rather than trying to ONLY measure the voltage on the
unknown resistor.

[Phil Allison[_3_]]

"Phil Allison"

**Correction:

The fractional increase in resistance is simply multiplied by 245 to get
the actual temp rise of the copper - which is the thing that matters most.


** The number should be 234.5 for Copper and 228.1 for Aluminium.

From the Australian Standard on transformers for measuring temp rises of
windings.




...... Phil

[Maxnumar]

I almost for got .
You might look into using FRED Diodes Fast recovery epitaxial diodes for the bridge rectifier. they have alot less sag then the regular ones.
wlrs

[Phil Allison[_3_]]

"Maxnumar"

I almost for got .
You might look into using FRED Diodes Fast recovery epitaxial diodes
for the bridge rectifier. they have alot less sag then the regular
ones.

** Huh ??????????????




...... Phil