14
\$\begingroup\$

Suppose we have two transformers with same turns ratio, \$N_2/N_1\$. First transformer \$A\$ is of \$100:10\$ and the second transformer \$B\$ is \$1000:100\$. Both have same turns ratio so both can step down the voltage by 10 times theoretically. If we assume size of wires are identical at primary and secondary, current and power ratings are same (I am not sure). But eventhough turns ratio is the same for both transformers, the number of turns is different in primary and secondary.

So my question is does the number of turns in turns ratio matter? I have a feeling that \$ B \$ has some advantages. But I am unable to figure out.

\$\endgroup\$
5
  • 7
    \$\begingroup\$ I wish it didn't! It'd be a lot faster to turn a 1:10 transformer instead of a 100:1000 transformer. \$\endgroup\$
    – DKNguyen
    Jun 17 at 14:47
  • \$\begingroup\$ Here is a good tool ??? play.google.com/store/apps/… \$\endgroup\$
    – user288518
    Jun 18 at 15:31
  • \$\begingroup\$ Or this one, for "sizing" ??? changpuak.ch/electronics/Power_Transformer_Design.php \$\endgroup\$
    – user288518
    Jun 18 at 15:53
  • 1
    \$\begingroup\$ @Antonio51 Are either of those sites / apps yours? You are allowed to post links if they are but if so please note the ssociation. They look useful. \$\endgroup\$
    – Russell McMahon
    Jun 19 at 9:21
  • \$\begingroup\$ Ok. It is somewhat difficult to note the association or private person. Thanks. \$\endgroup\$
    – user288518
    Jun 19 at 9:26
26
\$\begingroup\$

Number of turns on a transformer matters greatly. The turns ratio is one of many considerations in designing a transformer. The following are high-level considerations when designing a transformer.

  • As Tobalt stated, magnetizing inductance is important. Too few turns will consume excess current, even without a load.
  • You need a minimum of turns to prevent saturation of the core. Number of turns controls the flux density. If the core saturates, your primary starts to look like a short circuit to the driver. You can also control when a transformer saturates by changing the effective core area.
  • Increasing number of turns will increase leakage inductance (can be modeled as a series inductance either in the primary or secondary leads) which may be undesirable. There are cases where leakage inductance is desirable. Leakage inductance can also be controlled by controlling the physical positioning of the windings and can increase the cost of increased build complexity.
  • Increasing the number of turns will lower the resonant frequency of the transformer (larger inductance, larger self-capacitance). You want to keep the resonant frequency above the highest frequency of operation by at least 5x (my rule of thumb).
  • Wire diameter and type (single, bunched, Litz) is important and affects the winding (copper) loss. For a transformer operated at a single frequency, there is an optimum wire diameter which balances AC copper losses (eddy losses: proximity effect which is prominent at frequencies below apx 1 MHz, and skin effect) and DC copper losses. Proximity effect can be reduced by using bunched or Litz wire, skin effect can be reduce by using Litz wire - both are expensive options. Copper loss is an important consideration in power transformers.
  • Core losses are affected by flux density. Less turns have a higher flux density which mean higher core loss. Core loss is an important consideration in power transformers
\$\endgroup\$
1
8
\$\begingroup\$

Any given core material, cross section and frequency has a maximum volts per turn. If you want to use the winding at a certain voltage, then you need enough turns to support that voltage.

For instance, low frequency mains transformer steel will only run up to a peak field of 1.7 T or so before it saturates. If you had a core that was 10 mm x 20 mm, and wanted to run it at 50 Hz, then the fastest the field swings, for a sine wave voltage, is 2pi.B.f = 6.28x1.7x50 = 534 T/s. In the 200 mm2 of the core, that's a peak flux change rate of 0.1 Weber/s, meaning that core at that frequency will only support 0.1 V/turn.

A 240 Vrms mains winding on that core would need to support 340 Vpeak, so would need 3400 turns minimum.

\$\endgroup\$
1
  • \$\begingroup\$ I think you should explain tte formulas being used, or cite from where they are. The question ws somewhat generic, and so should the answer be IMHO. \$\endgroup\$
    – U. Windl
    Jun 19 at 9:04
5
\$\begingroup\$

Obviously, B has some advantages indeed. Otherwise all transformers would use a tiny number of turns. More turns generate more magnetizing inductance \$L_M \propto N^2\$.

Imagine a transformer with the secondary open, which is essentially a large inductor. If you connect an AC voltage with frequency \$f\$ to the primary, it will see an impedance of \$2 \pi f L_M\$. If this impedance is too low, a lot of current will flow through the primary and be wasted with no effect. This excess current will be also present in addition to the regular load current when the secondary is connected. Therefore the excess current should be minimized and \$L_M\$ should be maximized. In practise, there is a compromise between a sufficiently large \$L_M\$ and a small/cheap transformer, so \$N\$ will be neither <10 nor extremely high.

Examples:

  • For an SMPS transformer, where the lowest frequency is perhaps 100 kHz you don't need a lot of \$L_M\$ to prevent excessive primary current, so a rather low number of turns will be fine.
  • For a microphone transformer, you have a low end frequency of perhaps 20 Hz and in addition you don't want to load the signal source too much, so you need a lot of \$L_M\$, which can mean 1000s of turns on primary.
  • 50 Hz mains transformers also need large \$L_M\$, because the primary voltage is high, frequency is low and you want the idle current to be low.
\$\endgroup\$
2
\$\begingroup\$

Transformer design has to optimize a lot of different variables. And I am not an expert. But the transformer primary needs to have enough turns so that it acts like a big inductor and prevents excessive current from flowing. If you take a particular transformer core, you can either wind the primary with many turns of fine wire (high voltage primary) or wind it with fewer turns of thicker wire (lower voltage primary).

But in both cases you MUST fill the whole winding area with copper to make sure the transformer will work well and have low resistance and handle its rated power.

You cannot just put one turn of thin wire. That is basically a short circuit, not a transformer. So, if you choose a transformer core based on power requirements, then choose your operating voltage, then you can calculate how many turns are needed to prevent oversaturation of the core. THEN you choose a wire diameter which will fill the available space efficiently.

That is more or less how the design process works. Once the primary is settled, you choose the number of secondary turns to get the desired secondary voltage. Once you know the number of secondary turns, you choose a wire diameter that fills the space effectively.

Generally, you end up with a lot of turns on your primary for 50 or 60 Hz power transformers.

\$\endgroup\$
4
  • \$\begingroup\$ Does the number of turns at primary n secondary matter for flux linkage as well?? \$\endgroup\$ Jun 17 at 6:41
  • 1
    \$\begingroup\$ Yes, it is N x Φ where N is the number of turns and Φ is the flux. \$\endgroup\$
    – mkeith
    Jun 17 at 6:48
  • \$\begingroup\$ Other question would be about heat and voltage: Will a "thinker winding" cause accumulation of heat, and (assuming more turns lead to thinner wires) will more tuns make it more likely for the voltage to "jump over", damaging the wires. \$\endgroup\$
    – U. Windl
    Jun 19 at 9:09
  • \$\begingroup\$ @U.Windl If the transformer core is designed correctly, and you wind it efficiently, filling all the space, and don't over-saturate the core by using too high a voltage, it should not overheat regardless of the wire diameter. \$\endgroup\$
    – mkeith
    Jun 19 at 9:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.