Maximum transformer power 1

[Ironhand]

I'm looking for a way to determine the maximum amount of power a transformer can deliver without getting damaged or becoming very inefficient. I know it can be determined by measuring the core dimensions, but this transformer has been submerged into liquid which has subsequently hardened (I don't know the English word for the procedure, but I suppose it's quite common). In other words, I can't reach the transformer itself, only the electrical terminals.

 

[230842]

It is not easy to answer your question. Any additional data are needed even to give any advice. The most important concern the rated voltages and the iron and copper weights.
Any other procedure needs tests and meassurements procedures and equipments.
To determine rated voltages, no load characteristics should be obtained: no load current versus voltage, from which voltage values on both sides can be selected considering iron saturation conditions. And no load losses versus voltage will draw additinal information.
To determine rated currents a short circuit test should be performed to determine short circuit impedance and short circuit losses versus SC current.
With that result values it is posible to get the internal equivalent parameters from which a first aproach to rated values can be made.
To complete study, a full load heating test including efectiveness determination will be imperative. Julian

 

[electricpete]

I don't think that you can reliabily determine the rating of this transformer by test.

#1 - One important limit will be temperature of the windings as a function of load, particularly at the hottest spot. From your description I believe you have know way of measuring that temperature.

#2 - How to determine what voltage the insulation is designed for through testing? No reliable way that I know of without testing it to failure.

 

I am not sure how big your transformer is. It sounds like a small electronic transformer encapsulated in resin.
The power dissapation of electronic transformers is well documented and reported in popular text books specialising in electronic transformer design. Consequently you should be able to get a hot spot temp for the size you have.
Unfortunately you do not know what the insulation of the windings is and consequently you will have to assume the lowest temperature rating of about 90C.
Alternatively you can measure the temperature of the windings at ambient and then apply an assumed full load current. If the winding resistance is checked every few minutes to see it has reached its maximum then the average winding temperature can be calculated and a hot spot temp estimated. You may need several tries at different currents to get the rating

 

[electricpete]

old guy. Good comments, now I am curious.

I agree resistance measurements give average winding temp not hottest as you point out (that's why I highlighted hottest spot in my reply). How can one estimate the hot spot rise above ambient when nothing is known about the construction of the device?

Also I'm curious if you have any proposal for determining what voltage to apply?

 

[peterb]

electricpete, I think that Julian was on the right track in suggesting a mag curve to determine saturation of the windings. The transformer will probably have a kneepoint that is not very much higher than the rated voltage, so the closest standard voltage below the knee is a clue as to a reasonable starting point.
It would definitely be instructive to know the general size of the unit - also what type/size terminals.

 

[electricpete]

you're right peterb, I didn't read Julian's post carefully. Based on his method for determining voltage and old guy's proposal, it looks like you have a reasonable basis for a rough estimate of the ratings.

 

The temperature of the hot spot above average is a bit of a guess but don't let that put you off.
After all the whole transformer is made up out of copper and iron all of which are good thermal conductors so the hot spot is not going to be much above average.
You are also assuming the lowest insulation temperature so if you get it a bit wrong your insulation is likely to be better than you assumed.
Say for instance you get it wrong by a few degrees. The transformer will not fail with a bang. An interturn short may develop which will eventually blow a fuse. The results of this are unlikely (depending on your application)to be catastrauphic. It will reduce life.
As to the amount to allow you must use your judgment. I would expect 10C to be plenty.
With encapsulated transformers the main problem is cracking of the encapsulation if they get too hot. Perhaps this may not cause a failure but it looks unsightly and is not the best way to treat power transformers.

 

I missed you question about voltage. However others have given the correct approach.
You should be able to decide which winding you are going to use as a primary by measuring its resistance.
Step down use the high resistance as primary and vice versa.
Then put a variac in the primary with an ammeter and plot the current as you increase voltage. You will note that as the voltage rises the current starts to increase rapidly.
You need to choose a voltage which does not have too much current.
Again do not get too worried about having the value exactly right. You do not know if the transformer steel is grain oriented or other. However ordinary laminations normally used at 13 kilogauss can be driven to 18 killogauss so there is a plenty of room to choose
When the transformer is on full load you will not even notice the mag current because it is all reactive. All that happens is a bit more heating from iron loss

 

[electricuwe]

If you are sure that you are dealing with a line frequency transformer determine the weight of the unit and compare to catalogue data.

 

[Ironhand]

Thanks for all your suggestions so far. I understand it's not easy to say anything about it based on so little data so I'll give some more info on the transformer block:
The block's dimensions are 123 * 134 * 105 (height) mm and weighs approx. 5.3 kg. The box itself is 0.75 mm metal, except for the top which is plastic (numbered 4022 369 37562, no further markings). The brown resin inside leaves 15 mm of free space under the plastic plate. On top there are 7 terminals: 1 and 2 on one side and 3-7 on the other.

From the resin 12 isolated wires emerge, which I de-soldered from the top plate (which I removed) and from each other. After some simple measurements I found out that what's inside is most likely a transformer with two secondary coils, with two wires on the primary coil (connected to terminals 1 and 2) and five on each of the secondary coils (the two coils connected in parallel to terminals 3-7).
The wires on terminals 1 and 2, as well as both the wires on each of the terminals 3 and 7, are solid copper wires, diameter 1 mm. The other six wires (terminals 4,5,6) have a 0.75 mm diameter and appear to be more iron-like. I suppose the 1 mm wires carried the larger currents while the other terminals were used only for special purposes. When I apply 230 VAC to the primary coil, each secondary coil's total voltage difference becomes 62 V, divided into 4-27-27-4 by the smaller wires.

I don't currently have any accurate measuring equipment, as my (cheap) digital multimeter broke some days ago. I made the voltage measurements using an (even cheaper) analog multimeter, but I plan to make more accurate measurements on the electrical/thermal characteristics as soon as I've bought a new (and hopefully better) multimeter.

Although I'm not sure, the transformer block has most likely been part of a power supply unit for an old electron microscope. The PSU input was probably 230 VAC/50 Hz. I have some other components which are very likely from the same PSU:
- Two 15 mF capacitors, rated 63 VDC, I eff (100 Hz)= 28 A. (height: 110 mm, diameter 65 mm).
- Four diodes: two BYX42-300R and two BYX42-300 . (I couldn't find exact ratings, but I believe they are 300 V / 12 A).
I suppose these things have been used as a rectifier+filter, but I have no idea how far beyond the maximum ratings the actual currents were.

As to the suggestion to compare the weight to other transformers there are two problems (although I like the suggestion):
- I don't know the density of the resin, which makes it hard to know the weight of the transformer itself. (the density of the metal can be estimated accurately enough I think)
- The transformer is probably quite old, and I can imagine the performance/weight ratio of available transformers has improved over the years (but maybe not, transformers are still low-tech after all).

 

[jbartos]

Suggestion: Transformers can be loaded more than their nameplate rated paremeters permit. However, the thermal effects are limitations.

 

[jbartos]

Suggestion: Check
1. ANSI/IEEE C57.91-1981
2. ANSI/IEEE C57.92-1981
1 and 2 provide guidance for loading at other than rated conditions including:
a. Ambient temperatures higher or lower than the basis of rating
b. Short-time loading in excess of nameplate kVA with normal life expectancy
c. Loading that results in reduced life expectancy
Please, note that 1. and 2. are guides not standards.

 

[peterb]

From the size described, the transformer can't be rated at more than a couple of hundred VA. I suggest that Ironhand visits Radio Shack (or equivalent) for a new unit, unless the interest is strictly academic.