5:1 AC Step Down Transformer

Hi Everyone,

I'm looking for a 5:1 step down transformer for an AC signal. More specifically, I'm looking to step down 5v to 1v. Also, the frequency of the AC signal is 134.2 kHz.

I would have thought that to find this would be fairly simply, but having spent several hours search the internet, I've found nothing. What am I missing here?

Thanks for the help!


Do you want to buy one, or make one ?

Why would you want to do this ? It seems rather pointless. Nobody else wants to do it, which is why such a product probably doesn't exist.

If you want to try making one, there are plenty of online tutorials and youtube videos about how to make your own transformers.

If you simply want to measure or detect this signal, and you want to reduce the voltage for compatibility with an A/D converter, or something, then the simple answer is a voltage divider using two resistors.

Thanks for the responses.

The reason I would like to use a transformer is because I'm trying to isolate two circuits. In one, the voltage is higher, but there is a current limitation due to an IC. On the other, I would like to have an increased current (by using a step down transformer).

I found several transformers on digikey. I have one quick question: I understand the significance of the turns ratio, but when a manufacturer publishes a voltage for the primary and secondary winding, is this a maximum voltage or the exact voltage that must exist across the winding.

In other words, if the primary winding has a voltage rating of 5 v and the secondary winding has a voltage rating of 1 v, does this mean the voltage step down is 5:1 or that 5 v must exist across the primary winding?

You can use a transformer at any voltage until the insulation breaks down. You are not going to do this at 5V so don't worry about the voltage. Sometimes the turns ratio can be expressed as voltage in and out. However you have to be cairfully of the input impedance of the transformer and make sure it is not too low for your driving circuit.

You specified a frequency of 134.5kHz. If that is correct then it is unlikely that an iron cored transformer will meet your needs. You will probably need a ferrite core unit and at those sort of frequencies you should be able to wind your own using shellac coated (or equivalent) copper wire. Or alternatively buy something like one of these http://uk.rs-online.com/web/p/telecom-transformers/0196369/. From its data sheet you'll see it offers high frequency bandwidth and high voltage isolation. Whilst its ratio is 1:1 rather than the 5:1 you desire, you can use a simple resistor divider on the output stage to get your desired ratio.

What kind of isolation voltage are we talking about? If its really low a transformer might not be the best approach. If its really high you'll need a transformer rated for that isolation voltage.

Sounds like an application for a pulse transformer perhaps?

There are kits for winding one's own transformer/inductors, BTW.

More info on the driving and receiving circuitry is useful.

Perhaps you should explain what you want to do, rather than how you think you want to do it.

Sorry I didn't originally post the entire problem. I agree that this approach would be most helpful.

The overall application is an RFID reader. As of now, I am using an EM4095 IC, which drives an inductor and cap in an RLC circuit. My issue is that I need to increase the RFID read range, which I plan to due by increasing radiated power. Unfortunately, the EM4095 is already outputting its max operational current of .25A. In addition, the EM4095 has two voltage drivers, both operating 0-5v but in opposite directions, providing a p-p ac signal throughout the RLC circuit of 10 v p-p. Per my calculations, amplifying the power to roughly 5 times what it is now should be adequate.

The plan I originally suggested was to connect a transformer across the two voltage drivers, with the RLC components existing in an isolated circuit due to the transformer. I realize this approach brings up several problems:

  • The transformer must be able to operate at 134.2 kHz
  • The inductor's inductance is 2.8 mH, so preferably the transformer doesn't add too much inductance. I realize I can compensate by adjusting the cap, but the RLC circuit already has a high Q, and I don't want to narrow the BW too much.
  • The current output from the EM4095 is dictated by the intrinsic resistance of the inductor. If I were to isolate the RLC circuit and step down the voltage, the resistance of the inductor would also have to be dramatically reduced in order to allow an increase in current in the isolated circuit

I hope I stated this clearly enough to help answer anyone's questions. Please feel free to ask more if anything is still unclear.

Thanks again for the help.

amplifying the power to roughly 5 times what it is now should be adequate.

The snag with that is that a transformer does not amplify power. In fact you will loose some power going through a transformer.

The only way to get greater range with that chip is to design a better coil.

The chip not only generates the AC feed, it senses changes in the load due to the RFID chip modulating its tuned circuit.

The tuned circuit of the reader itself "transforms" the voltage up automatically to the maximum the available power and the Q of the circuit allows - assuming component values have been correctly chosen.

The way to increase range is choose a larger coil (range is proportional to the size of the coil) and increase power to match. Also having the most sensitive receiver helps

With a larger coil you get less signal (proportionally) from the RFID as its cutting a smaller proportion of the flux from the coil, so more sensitive receiver is needed to match, I believe.


Correct me if I'm wrong, but although the dissipated power won't increase, by stepping down the voltage a higher current can exist in the isolated rlc loop. Of course this will only happened if the intrinsic resistance of the coil is decreased.

Also, the radiated power is dependent upon the current through the coil, not the driver voltage.

One thing I forgot to mention is that a sensing lead is connected between the coil and cap which carries that signal to a sensing pin on the em4095.


SkiBum326: Correct me if I'm wrong,

Oh, he will!

In short, you must assume the design is already optimised to obtain the best performance from the parts available and that no "simple" modification will drastically enhance its performance.

A larger coil will increase range, but you then have to work out how to design it so that it tunes to the frequency. Not trivial.

I think we have forgotten about the transformer by now. :D

Would an Op-amp with a transistor on the output fit the bill?

The op-amp has high input impedance and lets you regulate the output with the transistor/FET providing the current required. Choosing the right op-amp and transistor, the frequency would be maintained accurately as well.

Edit: Just read that the coil needs to be 'read' at the same time so the above may be moot.

lemming: Just read that the coil needs to be 'read' at the same time so the above may be moot.

And that indeed, is the "trick".

Using a single-chip implementation of this sort of function prevent you from using "outboard" amplification in the separate transmit and receive paths.

And a "plain" op-amp is not the most suitable for these frequencies.

A mentioned before, you must assume the design is already optimised to obtain the best performance from the parts available and that no "simple" modification - if even practical - will drastically enhance its performance.


Correct me if I’m wrong,

Oh, he will!

Ah you know me so well.
You can not get a power increase from a transformer because that would break the conservation of energy principle and allow things like a perpetual motion machine. The second law of thermodynamics has something to say about it as well.
There are three main causes of losses in a transformer.

  1. I2R losses - Heating of the coil caused by resistance in primary and secondary windings.
  2. Dielectric losses due to the coil’s core material.
  3. Coupling loss due to the fact that the windings will not share all the magnetic field.

That is quite a good chip for a simple RFID reader but it is never going to win any prizes for range.

It has a PLL that is constantly twiddling with the excitation frequency to keep the coil in resonance. This means that while the antenna is in resonance the token is not being driven at its resonant frequency. You can adjust the coil or capacitor so that the twiddled excitation frequency is closer to the excitation frequency of your token. The token excitation is different in different tokens but normally has a certain range, which is less than this system can span. This will improve the range.

However, this chip can also have it’s range improved, it says in the data sheet:-

Read sensitivity (and thus reading range) can be increased by using external envelope detector circuit. Input is taken on antenna high voltage side output is directly fed to CDEC_IN pin.

But for ultimate range throw that chip away and make a reader from it’s discreet bits. The maximum range is not achieved at the resonate frequency but just at one side or other of this.


Thanks for the response. I read up on envelope detector circuits, so I'm somewhat familiar with their operation. Is there any way for you predict how much of an impact such a circuit could have on read range? Or is that something that I can only find out through actual testing?

On a separate note, I originally decided to go down the path of using a transformer because I was under the impression that I was already operating under maximum current draw from the EM4095. This impression was based on my estimation that the intrinsic resistance of the coils I'm using is roughly 20 ohms. However, I just tested the resistance using an LCR meter. One thing I'm confused about is that I can select a test frequency for resistance...I thought resistance was a dc characteristic. If the frequency does indeed effect intrinsic resistance, then the reading I got at 100 kHz of 60 ohms would indicate that I'm currently operating at 1/3 the max operating current of the EM4095..

Sorry for being all over the place; I'm just exploring all options that could lead to a solution.

Thanks again for all of the help.

I think I may have just answered my question. Wouldn't the resistance increase due to skin effect at higher frequencies..the conduction path is narrower because of skin effect.

Wouldn't the resistance increase due to skin effect at higher frequencies..the conduction path is narrower because of skin effect.

No not at the frequencies we are talking about. Unless you are a self deluding Hi-Fi nut.

I just tested the resistance using an LCR meter.

thought resistance was a dc characteristic.

It is.

If the frequency does indeed effect intrinsic resistance, then the reading I got at 100 kHz of 60 ohms

There is the property of inductive reactant, perhaps your meter is measuring that, it is measured in ohms as well.