Early Unusual Transformer Design

[ChristianFletcher]

I guess I can also measure the transformer regulation

[kalee20]

Originally Posted by ChristianFletcher

I assume I need to be measuring true power rather than VA

Oh yes, definitely - an electronic wattmeter or a twin-coil dynamometer instrument (seeing this is a vintage forum, preferably the latter!)

[kalee20]

Originally Posted by turretslug

Something rings a bell about Mr. Faraday's first transformer (or at least a demonstration item) having a bundle of iron wires as its core. "Sorry Mike, no lams here, no-one's invented a use for them yet...."

And when he'd got his core, and came to wind on it, "Sorry again Mike, no enamelled wire here, no-one's invented a use for it yet..."

He insulated his bare copper wire by lapping calico around it as he wound it on.

Michael Faraday - some day I will get to kneel beside his grave. I owe so much to him...

[broadgage]

ISTR reading the instructions for home building a small transformer of the type illustrated. In a 1930s issue of the "model engineer".

"Florists wire" was recommended as the core material.

The main drawback of the design is that heat from resistance losses in the windings can not be readily dissipated as the windings are covered by the iron wires that form the core.

[ChristianFletcher]

Florist wire very well describes what i here

[ChristianFletcher]

Originally Posted by kalee20

Oh yes, definitely - an electronic wattmeter or a twin-coil dynamometer instrument (seeing this is a vintage forum, preferably the latter!)

Quite right! I did actually find my twin coil dynamometer but it’s 500 watt reading so not really suitable. But did find this blast from the past and did remind my off the power theory classes at Huddersfield university. Hard to imagine sometimes being young

Also known as Stubs wire, very soft iron, great for all sorts of magnetic things.

[Ed_Dinning]

Often used as the core on induction coils, packed into a plastic drain tube

Ed

[ChristianFletcher]

I’m just in the process of measuring my open circuit iron losses and my short circuit copper losses. But the formula for efficiency calls up the transformers rated power in VA. Given that this is an unknown transformer I don’t have the VA rating.

From a specification sheet for the WIB it does say the output current is 20mA and states output voltage of 130 but the output voltage would be considerably lower at the output terminal give the drop across the smooth choke and the rectifier So the secondary voltage would have been higher.

The only thing I can think of is to say the VA rating is equal to the secondary current multiplied by the secondary voltage. So this would be 20mA x 160V = 3.2 W with the transformer loaded at 20mA

[ChristianFletcher]

I’m still confused about the VA rating above but wanted to do a sense check. So I have the transformer connected via a variac. I have transformer on the 200-220 Volt tapping so splitting the difference I have 210 on the primary.

I have arranged a bank of resistors on the secondary and have a secondary voltage of 160V at 20mA secondary current. So I have 160 x 20mA developed across the load I.e 3.2 Watt.

I have a watt meter in the primary circuit and the is indicating 7.4 watts. So I have 7.4W - 3.2W = 4.2 watt off losses in the transformer.

Efficiency = Pout / Pin = 3.2 / 7.4 = 0.43 or 43% efficiency

[kalee20]

That's pretty lousy for efficiency, but the methods and calculation is spot on.

The secondary load being resistive, a voltmeter and milliammeter is OK for the load power (else you'd have needed to use the wattmeter there also).

As a sanity check, you could of course remove the transformer, reduce the Variac to 160V, and connect your load (with voltmeter and milliammeter) directly to the wattmeter load terminals, and check it reads 3.2W.

Small transformers aren't renowned for their efficiency, though 60-70% is generally achieved by the 1 - 5W items.

[ChristianFletcher]

Thanks yes I did check the watt meter against just the resistive load and it measured closely. I found my plug in wall wart energy monitor was inaccurate but the feedback meter works much better. The above method works but I really wanted to use the standard formula and separate my losses into core and copper losses. I’m guessing I can still do that using the open and short circuit method and the see if the sum to 4.2watts. I’m guessing copper losses will be relatively low compared to the iron losses for such a lightly loaded device.

Just to bump up the question. Is there a way to determine the VA rating of a unknown transformer

I think I was over thinking the problems as the formula I was looking for calculating the efficiency said the output power is equal to the rating of the transformer in VA. So that’s really just saying the power being delivered to the load. Which in my case is 3.2w

There is the limiting factor, an unloaded or output shorted transformer is 0% efficient.

[Refugee]

I have done multi output transformers in the past.
The first thing I did was to guess the regulation figure and use that as a bench mark.
I loaded it up with resistors until the drop in output was equal the regulation estimate.
Using the DC resistance of the winding helps with load resistor selection as it is pretty representative of copper losses. Any more voltage drop would account for any on load iron losses that may occur.
With a modern transformer the weight gives a good indication of overall VA rating.

[ChristianFletcher]

I am planning on measuring the regulation and will assume the 20mA specified for the W1B is the full load current. I also noted the W1B states operation between 50 to 100hz so I was going to sweep the transformer and see how if it effects the regulation. Are there any other useful or interest test I could do.

[Radio Wrangler]

Originally Posted by ChristianFletcher

Just to bump up the question. Is there a way to determine the VA rating of a unknown transformer

Load it up and go by temperature rise. Might be best to try to sneak a thermocouple in.

David

[Ed_Dinning]

Industrial method of checking temp rise accurately is the measure a winding resistance at ambient using a bridge circuit and accurately. Run transformer on load for a good time until stability is reached then re measure winding resistance.
Check for accurate figures but copper has a temp coeft of resistance of 0.3% per deg C
Increased resistance now gives transformer copper temp rise

Ed

[kalee20]

Originally Posted by ChristianFletcher

Just to bump up the question. Is there a way to determine the VA rating of a unknown transformer?

Not really! It depends on the application, and why a rating exists at all...

If you buy a transformer rated at 100W and load it to 101W, it's not going to expire like a nuclear holocaust.

There is of course a maximum power that can be drawn from a transformer. As you draw more current, the output voltage falls due to winding resistance. Eventually, you reach a point where the extra load power obtained by increasing load current is more than offset by the drop in voltage... it's all downhill from there. It happens when the output voltage has fallen to about half the off-load voltage. It's a manifestation of the Maximum Power theorem. But, the figures are such that most transformers would severely overheat if you tried to do this.

The exception might be transformers which are loaded only briefly, like a doorbell transformer, or a welding transformer.

As for continuous ratings, it depends what the limit is. It might be tolerable output droop, important if the load is variable (like a 12V lighting bus for an array of halogen lamps). Too many lamps might pull the voltage down so that even though the transformer is at a comfortable temperature, it's still no use because the lamps are dim... and even if the transformer is designed with a bit of margin, you can't have the voltage rising at lighter loads else the lamps still connected will go pop.

There again, it might be temperature rise. So if your transformer is running rather hot (approaching the temperature where the insulation materials used within it start to degrade), you could get around that by operating in a cooler environment. Or fan-cooling it. So, the rating would be highly dependent on the ambient conditions.

Ed Dinning gives a very good method for determining temperature rise. There are rules of thumb for estimating temperature rise knowing the winding resistances, the currents, and the surface area if the transformer, which can be worked backwards so that given a transformer size, setting a maximum temperature and measuring its winding resistances you can work out the maximum currents, but reality rarely gives agreement to better tha 10%.

[ChristianFletcher]

Thanks for the comments really useful. Always interesting to explore a subject more deeply than is required for typical repair job.

I did my open circuit test and got an indicated core loss of 4 watt. I then did my short circuit test to measure the copper losses and got 0.3. I previously measured the total lost as 4.2 so I have a slight error but I’m sure I’m in the ball park. But interesting my core losses for the hedgehog are really high but that’s probably to be expected for such a crude manufacture of the transformer.

I have run the transformer for about an hour and it just gets warm to the touch. Just by feel I don’t think the temperature change is very much different for on or of load. Again I guess this indicates that for this transformer it the core losses that dominate.

[kalee20]

Originally Posted by ChristianFletcher

I have run the transformer for about an hour and it just gets warm to the touch. Just by feel I don’t think the temperature change is very much different for on or of load. Again I guess this indicates that for this transformer it the core losses that dominate.

It sounds very much like it!

While you have the measuring equipment out, can you share with us the winding resistances; the open-circuit voltage (tell us the primary voltage!); and the full-load voltage (at the same primary voltage).

The no-load and full-load primary currents would be good to know, too.

I can do some number-crunching then!