TRF7960 Power amplifier

Elia,

The amplifier class is a class E amplifier.  As far as literature, I don't now off hand of any single source of information for this amp, but your usual collection of general circuit design, resonant circuit theory, and electromagnetics text books should cover everything.  I just looked at the Wikipedia entry, and it seems to do a good job of describing class E operation, but doesn't give any information on how to design one.  I once read a chapter out of the book "Omnidirectional Inductive Powering for Biomedical Implants" where the author does a good job of describing the class E design process, but I can't remember the author's name or where I found the pdf file.

I too have heard that the IRF510 is not the best for this purpose.  While I am sure that this is true because there is always something that could be better, I personally use this transistor in my design and I have not had any problems with it.  I operate my design, which was copied almost entirely from the TI design, at levels between 100 mW and 20 Watts, and when the amp is driving a properly impedance matched load, the system has better than 90% efficiency.  I use no extra heat sinking or cooling on the transistor, and I regularly have it running for hours at 10 Watts output without the thermal protection ever being activated.

My advice would be to draw the reference design power amp in a circuit design simulation software suite (Pspice, LTspice, etc.) exactly as the reference design circuit, and play around with the circuit until you get a real feel for how it works.  From there you can decide to use the reference design, or you can choose to design your own amplifier.

Sparkchaser

I want to post a little update on the class E amplifier.  I recently came across a piece of freeware software which does an excellent job of simplifying the class E design process, and the website also provides very good information regarding the design methodology.  For those who are interested, the software can be found by googling  "tonne software," and then following the ClassE link.

Sparkchaser

Hi,
Thanks for the replies. I remain however some doubts about topics not covered in the documentation:

1) The circuit diagram of the amplifier contains certain components which must not be for normal operation. I read that they are needed in test phase, but this topic is not treated anywhere. What do you recommend?

2) I am not entirely clear where the information contained in the square signal to be provided at the entrance. For the case treated amplifier also acts as a modulator?

3) Finally, I do not understand what it means "The Reader's Tx output (U3-5) is 4 ohms with a square wave output". As far as I know each harmonic square wave sees a different impedance.

Regards.

1)  During development, it is common to have extra pads on the PCB available to do fine tuning with the matching components.  You place one component which you think will get you in the correct range based on your calculations, and then you add another component to get you to the exact value needed for proper impedance matching according to the measured results.

2)  I don't quite understand the question.

3) To perform impedance matching between the IC and the external network, the IC's output is modeled as a square wave voltage source with a 4 Ohm output impedance.

Hi,

Thanks again for the answers. I need to test this circuit without connecting the TRF7960 but with a signal source external into secondary entrance.However the shematic have not specified the values of the components on this input(C88,C89,R90), you have any suggestions about it?

I also tried to simulate the circuit and is given to me matched ad 12.56Mhz instead 13.56Mhz, very strange.

Regarding question 2 of the previous point: I was wondering where the information was contained to send the antenna; in amplitude of the input signal? The pulse or something?

If you are using a signal generator with a 4 Ohm output impedance, which I would consider not too likely as most are normally 50 Ohm, then you either short or leave open the necesary components of C88, C89, and R90 so that the input signal is applied directly to the circuit input.  If you are using a voltage source with a low impedance output which is capable of driving a low impedance load, then you populate C88, C89, and R90 with elements which will give you a 4 Ohm impedance between the signal source and the circuit input.

12.56 instead of 13.56.  Simulations are usually not perfect representations of what you can expect in real life.  It really depends on how much time and effort was spent designing the Spice models of the components used in the simulation.  The more accurate the models used in the simulation, the more accurate the simulation.

Modulating the input signal will modulate the output of the amplifier.

A little note on impedances.  An output impedance is not necessarily what the source sees as it is looking towards the rest of the circuit, it is what the rest of the circuit sees as it looks back into the source.  A voltage source, such as an opamp, is a low impadance source because when the rest of the circuit tries to pull or push current back into the opamp, the opamp will sink or source as much current as it is rated to to maintain its output voltage.  If a source's change in voltage in response to a change in current is 0, then it is considered a perfect voltage source with 0 output impedance.

Hello,
I made my project and now it seems to work fine. I have a question though: I know that the class-e amplifiers are driven by a square signal on the gate, here the signal is almost sinusoidal at the gate. How do I know this is a class-e?

Hi 

I am a student, and as a degree project I am working on a RFID reader based on the TRF7960 chip , I need the reader range to be about 50 cm, so I have to use a power amplifier

I have a question for sparkchaser 

You say you built an TRF7960 amplifier based on the TI design, and you can obtain power levels between 100mW and 20W, how do you modify this level? by changing the bias point level?

I also want to know if you used a DC source higher than the 12V specified 

and finally, did you tested the amplifier with an antenna? where did you get the information for the antenna?

If anyone else can provide me more information about the power amplifier I'll be very grateful.

Thanks in advance

Hello Ricardo,

I adjust the output power level by adjusting the supply voltage.  The bias point, when it is properly tuned, should be left alone.  Also, while I said that I ran the system at 20 Watts, I found it pretty much impossible to extract the data signal from the carrier at this power lever.  In general, I try to maximize my link quality by employing good antenna design rather than jacking up the transmit power.

The output power level of a class E design is determined by its supply voltage and by its output load impedance.  By output load impedance, I mean the network of inductors and capacitors at the MOSFET drain which causes it to resonate at 13.56 MHz.  There are many different combinations of L and C which will resonate at 13.56 MHz, and each one will have a different impedance.  When this impedance is real (by this i mean there is no imaginary part), then the output power roughly follows V^2/R where V is the source voltage - MOSFET saturation voltage (for the IRF510 Vdsat is between 1.5 and 2 Volts), and R is the output load impedance.

Lets say that the saturation voltage of the IRF510 is Vdsat = 1.6V, then we get an output impedance for TI's amplifier of about (Vdd - Vdsat)^2 / 4Watts = 27 Ohms.   So for me to get 20 Watts, I needed (27 Ohms * 20 Watts)^(0.5) + Vdsat = 25 Volts (Vdsat is not constant, but I have chosen to ignore that fact for this demonstration).  I actually had quite a bit more difficulty with getting the system to work at 100 mW because that is what the gate drive circuitry would inject into the system even though I had Vdd turned almost all the way down.

In TI's output network, they include impedance matching to match the 27 Ohm Amplifier output impedance to the 50 Ohm antenna impedance, along with some bandpass filtering.  The system is design to work with a 50 Ohm load installed, but it will not destroy itself if you forget to connect the antenna, or if you use a horribly tuned antenna.  This is not the case when you are running it at higher supply voltages.  I you have a large supply voltage, and you run the amplifier with no load, it is very easy to very quickly destroy the MOSFET.

I designed my own antennas.  In our lab we have a vector network analyzer, and it makes building antennas very easy.  I simply make the loop, and then using the smith chart function of the analyzer, I tune it to 50 Ohm at 13.56 MHz.

Here is some literature:

"Class-E RF Power Amplifiers" by Nathan O. Sokal    - this is the paper from the inventor of the class E design

Design Equations for Class-E Power Amplifiers by Mustafa Acar, Anne Johan Annema and Bram Nauta  - this paper covers another method of designing the output network without using an RF choke

Analytical Design Equations for Class-E Power Amplifiers with Finite DC-Feed Inductance and Switch On-Resistance  by Mustafa Acar, Anne Johan Annema and Bram Nauta  - this paper covers another method of designing the output network without using an RF choke

good luck,

Sparkchaser

Newbie here - yet oldie in other fields.  Hoping you guys can help a simpleton understand some basic concepts.

I have researched the old  RFID readers TI S4100  is designed around the S6700 Multi-Protocol Transceiver IC, an integrated HF reader system that contains all the high frequency circuitry comprising an Analog Front End (AFE) that decodes the ISO standards protocols.

I am now looking at buying the TRF7960EVM evaluation board as the center core of this project.

When a guy like me looks at putting together these components via schematics he smartly hesitates and looks for alternatives or shortcuts to solutions that already have components within them.  The advantages of using standard design include the variety of ready-to-use designs, applications notes, and test equipment. The resulting system is scalable, versatile, and modular. However, the need for accurate design, dealing with accurate  filters and semiconductor's min-max parameters and ratings, stretches the design and implementation time, and may cause long and tedious system testing and tuning.

So enough of that, to the point - The system I am looking to replicate has 5 components and I am looking for shortcuts and saving both time and money in coming to the same conclusion the study did.  

1) An RFID/NFC reader, (based on multi-functional protocols)  2)  an RF power amplier,  3) load modulation receive buffer, 4) high power antenna and 5) a 12Vdc power supply. The RFID reader generates all the necessary RF signals according to the ISO protocols.  These signals are amplied by the power amplier to generate the RF power which is radiated through the loop antenna. The loop antenna performs the interaction with the RFID
tag, and senses the load modulation signals. These signals are buffered by the Load Modulation Receive
Buffer and fed back to the reader detection input. The Reader communicates with a host system via an
RS232 serial interface.

Now how much has changed since this research?  I am struggling to find a system or RFID reader/writer equipment which has some of or all of  this capability.   Do I have to keep with these 5 separate components to have as much testing or sampling as possible or can I take a short cut with - I see these so called High power HF RF power 4watt amplifier with multiple antenna sma inputs and reader/writer all in one?  But does this remove other functions from my original 5 component design?  Maybe there is a load modulation buffer already embedded in a reader?  I see a new system called PJM from Australia One of my main objectives is to test my system by increasing  the output power and antenna size as these have the most direct influence on read/write range capabilities.  

I research 13.56Mhz tuned power antennas and get these expensive commercial solutions or inadequate PVC (non-conductive) pieces of plastic.   I keep coming back to a simple and cheap solution and that is to buy some 1/2" copper tubing with resonance parallel capacitors C33 and C34 that were merged into one capacitor of 82pF, since the calculated antenna's inductance was around 1 H. and keep coming back to building it myself using the TI antenna cook book.

I am hoping you guys with all this experience will simply help with providing some equipment or component options for me in setting up my testing system.  Are there power amps I should consider? should I have both gain and power control?  Is there one better than the other for RF amplification?  My power supply  will be a non-switchable 12-24Vdc.   When I talk to sales people or distributors they say WHAT?  we only sell the stuff......

I promise not to bug you guys once | get it all set up for testing...  I haven't even started with the software issues yet... Oh by the by,,,,,  I am disabled in a wheelchair and this is what I do for fun in my vast spare time... I hope you can and are willing to help.

Cheers