Stratix II: Coil circuit

[Altera_Forum]

I am trying to create a very close range wireless communication using stratix II fpga. 

I want to achieve that using coil - coil communication.  So I connected coil1 to fpga1 with a series resistor of low value around 100ohm. At the secondary coil (with has around 5x turns and bigger radius) I can get as high as 5 volts (AC - open circuit) when fpga1 outputs a high voltage. Perfectly as I need.  Just when I try to connect coil2 as input to another fpga2, the voltage suddenly drops to 0.5volts!  I have a feeling this has to do with impedance mismatching, eventhough I do not understand why since fpga2 is not outputting anything.   Need to mention here that current arrives coil2 is very small (micros). But fpga2 needs only voltage to operate right? No need to care about current I guess!  So what should I do to make sure coil2 sends at least 2volts to fpga2? I tried connecting a series resistor to coil2, did not work.

[Altera_Forum]

A coil with a capacitive load (oscilloscope probe or a FPGA input) is a resonant circuit, connecting a different load changes the resonance frequency. It's basically possible to design a wideband inductive coupling, but most likely an amplifier with suitable frequency characteristic is required. 

 You didn't tell about the waveform that's intended with your application.

[Altera_Forum]

--- Quote Start ---  A coil with a capacitive load (oscilloscope probe or a FPGA input) is a resonant circuit, connecting a different load changes the resonance frequency. It's basically possible to design a wideband inductive coupling, but most likely an amplifier with suitable frequency characteristic is required.  You didn't tell about the waveform that's intended with your application.  --- Quote End ---    A great answer, and a confirmed one. I checked in the lab and what you said is absolutely true.   As a start I want to send a digital clock signal.  Now I have a problem. I am not sure how much capacitance the FPGA has, but if I only connect a coil of 100microH, I have Wo (the resonant frequency) at around 1.6MHz (which gives the maximum voltage to the FPGA input). I want to increase Wo alot, to a magnitude of 100s of MHz.   I know I have two options to play with on the secondary circuit side: Either to make my L (coil inductance) lower, or make my C (capacitance) lower. At this point I have these questions:  1. What is the capacitance at the input of the FPGA (Stratix II GX)? I can connect a capacitor in series to make the capacitance much lower in this case.  2. If I made my L lower, the communication distance between primary and secondary circuits should be reduced as a result. And since I already have the coils almost touching each other, I do not think this is an option. What do you recommend?  3. Is there another better way to operate the circuit on a much higher frequency?

[Altera_Forum]

Pin capacitances are listed in the device handbook, simply assume 5-6 pF for most pins. 

 You can easily see, that it's impossible to achieve a several 100 MHz resonance above a certain coil inductance. But there's no need to have more than one turn with your coil, thus you may be still able to design a coupler with sufficient output level.  You should consider however, that you're effectively operating a RF transmitter with considerable output power (mW range) and that your instrument won't pass EMC tests unless the coupler is completely shielded by a metallic enclosure. Even if you would use a free ISM (industry, science and medical) frequency, the clock signal has to be thoroughly filtered from harmonics.

[Altera_Forum]

--- Quote Start ---    You can easily see, that it's impossible to achieve a several 100 MHz resonance above a certain coil inductance. But there's no need to have more than one turn with your coil, thus you may be still able to design a coupler with sufficient output level.   --- Quote End ---    Could you please elaborate more on this?  I understood what you said, still I do not understand the reason or the explanation of it.     --- Quote Start ---  You should consider however, that you're effectively operating a RF transmitter with considerable output power (mW range)  --- Quote End ---    What is the best way to measure transmitted power in this case? Isnt the average power consumed by the coil is zero? I know if I want to transmit more voltage, I have to transmit less current. But since I am dealing with FPGA, I am more interested to transmit higher voltage. How would I increase the transmitted power?    --- Quote Start ---  and that your instrument won't pass EMC tests unless the coupler is completely shielded by a metallic enclosure. Even if you would use a free ISM (industry, science and medical) frequency, the clock signal has to be thoroughly filtered from harmonics.  --- Quote End ---    okay give me a moment here. 1. I have to make an EMC test?  2. How should I filter my transmitted signal? Just use a bandpass digital filter at the output of FPGA1 (the transmitter)? or programming this FPGA to do that job and output a filtered signal is a better option? 3. Will it really cause me, as a person, a problem to send a wireless signal with some harmonics outside the ISM band?  Thanx! Interesting discussion we have here!

[Altera_Forum]

I simply suggested to use less turns to achieve resonance with the present capacitive load. 

Also the transmitter coil can be operated in a resonant circuit to achieve a higher output level.  You don't need EMC tests for a hobby circuit, but you are required to keep the respective regulations. If your device is disturbing e.g. administrative radio-telephony or public broadcast, you can get in trouble.

[Altera_Forum]

How about the power measurements? 

 What is the best way to measure transmitted power in this case? Isnt the average power consumed by the coil is zero? I know if I want to transmit more voltage, I have to transmit less current. But since I am dealing with FPGA, I am more interested to transmit higher voltage. How would I increase the transmitted power?  And,  How should I filter my transmitted signal? Just use a bandpass digital filter at the output of FPGA1 (the transmitter)? or programming this FPGA to do that job and output a filtered signal is a better option?

[Altera_Forum]

It's not clear to me, in which regard transmitted power is relevant for your application. Although the coil is a reactive load, part of the power feed is consumed by losses, a part is transmitted to the secondary and a small part is radiated. The general limits for radiated emissions correspond to a few µW RF, to give you an idea of the order of magnitude.  

 As you mentioned above, you connected the coild to a digital pin through a series resistor. I don't see, how the FPGA should filter the signal, it's a squarewave anyway. So only an external filter seems to be possible.   I previously mentioned, that a filter would be necessary if you utilize an ISM frequency. Otherwise, your circuit won't comply to EMC regulations anyway, at least if the coil has some extension related to the wavelength and the transmission path isn't shielded. There are additional rules, also some national regulations, that define, which frequencies and how much transmitted power are allowed for particular RF applications.

[Altera_Forum]

--- Quote Start ---  It's not clear to me, in which regard transmitted power is relevant for your application. Although the coil is a reactive load, part of the power feed is consumed by losses, a part is transmitted to the secondary and a small part is radiated. The general limits for radiated emissions correspond to a few µW RF, to give you an idea of the order of magnitude.     --- Quote End ---    Thanx for the clarification.  I need to measure how much power I am transmitting to try and improve my communication distance. In addition, I will be asked how much power I am transmitting and I need to know how to answer! lol!  Could you tell me how to know the power I am transmitting? Its not simply multiply the current going into the coil by the voltage around it, on the resonance frequency, Is it?

[Altera_Forum]

It's not easy to measure because of the complex impedance. RF instruments, e.g. a network analyser would be required for an exact measurement. Supplementing the primary coil with a capacitor to a resonant circuit should ease a rough estimation from voltages at the primary.

[Altera_Forum]

Hi, 

Please dont use 1.6MHz it is the high end of the Medium Wave broadcast band and will cause local interferance, you will be generateing a lot of harmonoics some of these will be in the shortwave region and will go world wide. If you do want to do a home radio link then I would suggest using some radio modules for the unlicenced bands such as 433MHz. A google search for ism 433 mhz gave lots of interesting links.

[Altera_Forum]

--- Quote Start ---  It's not easy to measure because of the complex impedance. RF instruments, e.g. a network analyser would be required for an exact measurement. Supplementing the primary coil with a capacitor to a resonant circuit should ease a rough estimation from voltages at the primary.  --- Quote End ---    Ok I got what you said.  One more question. Since the rate of change in current in the primary coil will produce voltage at the second coil, I think its clear that we need to use a resistance with low value in series with the coil. Isnt that true?  At the secondary coil, parallel resistor with high value should be used to maximize the voltage at the fpga input. Right?

[Altera_Forum]

--- Quote Start ---  Hi, Please dont use 1.6MHz it is the high end of the Medium Wave broadcast band and will cause local interferance, you will be generateing a lot of harmonoics some of these will be in the shortwave region and will go world wide. If you do want to do a home radio link then I would suggest using some radio modules for the unlicenced bands such as 433MHz. A google search for ism 433 mhz gave lots of interesting links.  --- Quote End ---    I hope that I wont have to use that band. I will try to move it to 315MHz.

[Altera_Forum]

--- Quote Start ---  Since the rate of change in current in the primary coil will produce voltage at the second coil, I think its clear that we need to use a resistance with low value in series with the coil. Isnt that true?  --- Quote End ---   The maximum I/O current and voltage of the FPGA have to be kept, this sets a lower limit to the series resistor. External transistors or drivers/power amplifiers should be used, if higher currents are needed.  As said, the highest output level will be achieved with a resonant circuits at both sides and no load at the secondary.

[Altera_Forum]

--- Quote Start ---  I hope that I wont have to use that band. I will try to move it to 315MHz.  --- Quote End ---   Please do not use 315MHz you will cause major interference, it is in the Satalite uplink band. What data do you want to move over this link? What system design work have you done? What modulation scheme do intend to use? Is this a university project, or who is the customer?   James