Measuring Radiation from Mobile

[Olihou]

Hi Happy New Year Every Body !

I am asked to compare the radiation levels from mobile phones. Without the accession to professional radiation meters, I am thinking building up a circuit for this purpose.

Unfortunately I have not been able to find a circuitry or anything that would guide me to a start.

What I have in mind is a circuit with broadband (say 800 MHz to 2 GHz) antenna and amplifier, with a DC output of a fraction of a volt will do.

Any advice and help please !

 

[justDIY]

pretty sure all you need is a suitable dipole, a microwave rated diode and a millivolt meter

similar to those replacement antenna they sell that flash and do all sorts of tricks when exposed to rf energy

 

[Olihou]

Thanks justDIY.

I have disassembled one of the "handset flashers" you just mentioned. Components are very small and packaged that I am not able to trace out the circuit, and not even identify if there is a microwave diode, not to mention if it is a good dipole design.

Besides circuits, I am looking for some theoretical discussion so that I will have confidence in what I have built, say the sensitivity of the meter will remain more or less the same through 800 MHz to 2.4 Ghz (to take care of microwave oven leakage as well).

Any website I can visit ?

 

[Papabravo]

Hi Happy New Year Every Body !

I am asked to compare the radiation levels from mobile phones. Without the accession to professional radiation meters, I am thinking building up a circuit for this purpose.

Unfortunately I have not been able to find a circuitry or anything that would guide me to a start.

What I have in mind is a circuit with broadband (say 800 MHz to 2 GHz) antenna and amplifier, with a DC output of a fraction of a volt will do.

Any advice and help please !

Whoever asked you to do this without the proper equipment is asking a great deal. I think your chances of building a simple circuit and making reliable measurements without knowing what you are doing is exceedingly remote. Even if you had a circuit how would you develop any confidence in your measurements unless you had some way of calibrating your device.

Tell the person who gave you this assignment that you'll make those measurements if he'll pee up a rope.

 

[Olihou]

I believe I will be able to get the DIY meter calibrated but this is meaningful only if I have some confidence in its essential sensitivity and spectral linearity.

It is by no way an easy job, but probably not as much as "peeing up a rope".

There must be some one experienced who could advise, please.

 

[Analog]

I have experience in compliance engineering for EM emissions from electronic equipment, and it takes quite a bit to get a decent measurement. Everything from the environment taken (faraday cage if required), to absolute directionality, distance, orientation etc. This is not going to be a trivial task.

 

[mramos1]

Would the antennas need adjusting for the different frequencies as well? 800Mhz vs 2Ghz?

 

[Nigel Goodwin]

Would the antennas need adjusting for the different frequencies as well? 800Mhz vs 2Ghz?

I would have thought so, microwave oven detectors are designed for a specific single frequency - and require sending back yearly for calibration checking and alignment (apart for the toy ones, which aren't even calibrated in the first place).

 

[Papabravo]

Would the antennas need adjusting for the different frequencies as well? 800Mhz vs 2Ghz?

If you are using a 1/4 wave vertical you can compute the approximate length from:

0.25 * (3e8 / 800e6) = 9.375 cm
&
0.25 * (3e8 / 2e9) = 3.75 cm

This is well within the capability of a telescoping whip. At these frequencies precision is everything. An error of one tenth of a centimeter in distance is quite a difference in frequency.

If you need a directional antenna your mechanical adjustment choices become more limited. I don't know if such a directional antenna with a broadband response is even possible. I'm no microwave antenna expert, but I kinda doubt it.

I do know that on the 23 cm band (1296 MHz.) we loose half our power in the feedline to a 35 element yagi and the beam is quite narrow.

 

[mramos1]

So sounds like he can make a detector, but it will not tell him more that radiation is present. Sounds like a $19.95 sniffer from the internet will do.

 

[Olihou]

So sounds like he can make a detector, but it will not tell him more that radiation is present. Sounds like a $19.95 sniffer from the internet will do.

Sniffer ? What is this ?

 

[Olihou]

Hello friends

I have been away for a while. Nice to be back to the forum and see all your comments.

First I need to tell every body that I am not trying to make any thing complying to any industry standard, such as one for EMC measurements. I have simply been asked by buddies about the level of radiation from mobile phones, and the potential hazard to health. So as a hobbyist I assigned my self this task for building a detector that could compare the different levels of radiation from different handsets, and at different receiving conditions, etc.....

As already pointed out the most difficult part is probably to build an antenna and RF front end that responds equally (or just roughly the same) to radiations at a wide band of frequencies, say from 800 MHz to 2.4 GHz.

I have seen a design of self-powered microwave oven leakage detector: simply an antenna (quarter wave ?) plus a microwave diode plus a microammeter. However this is only for a single frequency (2.4 GHz).

I just wonder if my proposed wide band of 800 MHz to 2.4 GHz is technically feasible ? If not, what if there is any feasible design available ?

I understand that a serious theoretical treatment of this subject is beyond my capability, but I would be very much grateful if some one could give me a hand please !

 

[Olihou]

After much hard work of searching, I have eventually found a website relevant, providing a clever method of "electric field and leakage detector" at

https://www.4qdtec.com/putpr.html#efld

There is no mention as up to what frequency it can work. Apart from transistors, nothing seems to be there to limit the frequency of response. However the circuit works like a Geiger Counter, and the output (either in frequency/amplitude of the output pulse) may not be proportional to the amplitude of input radiation.

So this is a wide band RF detector that may be modified to work as a wide band radiation meter?

Any comments on this circuit please?

 

[Nigel Goodwin]

I don't think it's any use whatsoever to you!.

 

[RadioRon]

Oh my gosh, there are so many variables in this type of task that I don't know where to start in giving advice. First off, the detector circuit in that example will not be useful at all since I believe it operates best at fairly low frequencies and so will trigger from any charge pickup due to mains electrics, MW broadcast, static electricity, and hundreds of other low frequency sources. I suspect it would not operate well above 10 MHz in any case.

Detecting electric field, magnetic field, or electromagnetic radiation at specific frequencies, as the others have pointed it, is relatively easy. Measuring, as opposed to detecting, is very difficult. By measuring, we mean determining the amplitude of the field against some accepted scale or relative to a reference. This implies a certain amount of accuracy, and this is where a hobbyist's attempts will not succeed, as accuracy in measuring fields is a science and an art that takes a lot of effort and experience to get to useful levels. But, if you want to try measurement and aren't too worried about accuracy, it might make a useful and fun project anyway.

The best way to measure RF levels simply and quickly for mobile phones is to get your hands on an evaluation board for the Analog Devices AD8362. You may be able to purchase this directly from Analog Devices using part number EVAL-AD8362EB as noted in this link:

https://www.analog.com/UploadedFiles/Evaluation_Boards/Tools/609195956AD8362EB_0.pdf
(heres the general info on this chip **broken link removed**)

This board is valuable as it is unlikely that you can make the chip work to specification without some RF design tools and experience. If you were to copy the eval board layout exactly and use exactly the same parts they do, then you could build your own. Better off to get one of theirs, then you know it will work.

You would feed this board using specific antennas for each frequency band. Mobile phones will be transmitting somewhere in the range of 824 to 915 MHz if low band, or 1710 to 1980 Mhz if high band. A simple half wave dipole made out of some wire mounted to an SMA coax connector will work but you have to make one for low band and one for high band. In operation, you have no choice but to connect one or the other and measure for that band only as automatic switching between the two is beyond the reach of the beginner hobbyist. Determine the length of the wire (use #18awg solid copper) by calculating the half wavelength in free space and then multiplying by 0.4.

The antenna will be built around an SMA connector intended for pcb mounting. The input to the detector board will be an SMA coax connector. So you will need an SMA-SMA coax cable to connect the two. Keep it as short as practical, say 3 feet or less is fine.

Half wave dipole antennas have some useful selectivity, so such a setup may be more than enough for measuring fields fairly close to the phone. They will be useless when you get too close though. You need to keep the antenna further than about 3 wavelengths from the phone or your measurement antenna will actually disturb the phone's emitted field enough to mess up your measurement.

Such a setup also is completely useless for measuring how much radiation is going into the user's head. In other words, if you put your dipole a few centimeters from the phone, the levels you measure will simply be wrong. This is because you are connecting two radiators, the phone and your dipole, using near field coupling, and creating a sort of transformer that encourages energy to move across the transformer rather than radiate as it normally would. This coupling will be different than the coupling that normally exists between phone and head, so your result will be quite wrong.

Professionals who measure mobile phone levels at or into the human head use a fake head made of a hollow fibreglass mould filled with electrolytic fluid. They put an antenna inside that and move it around to measure field strength. This antenna is an electrically small antenna (not a half wave dipole) that has little effect on the measured field. Its quite complicated really.

 

[Olihou]

Thanks a lot RadioRon, for the very detailed and enlightening explanation.

I did not realize that there are sophisticated ICs such as AD8362 for RF power measurement. However this is well beyond what I need indeed -- I think a measurement range of 10 dB is good enough for me, and not the 60 dB as the AD8362 is capable of.

While I still have no idea why a wide band antenna could not be designed to cover both high and low band (even sacrificing the Q value), I shall take your advice in my attempt.

Having said that I believe that I have seen commercial field strength meters claiming wide frequency range from MHz to GHz, seemingly with one single probe only !? .......

 

[RadioRon]

A wide band antenna can be constructed to cover both bands. Indeed it was not my intention to imply otherwise, but such an antenna is more difficult for the beginner to build. Some common wideband configurations include the log-periodic antenna, the spiral antenna, the sinuous antenna, some horn antennas and others. The more common approach in the cellphone business is to make a "dual band" antenna. That is, one that works well from 824 to 960 MHz, and from 1710 to 1980 MHz but poorly otherwise. So these are not "broadband" as you and I would understand, but they get the job done in the bands that the cell business cares about. You could consider making a dual band antenna, but again such a thing is difficult without the aid of the right test equipment.

I think that the most likely broadband antenna which you can make using a textbook recipe and no test equipment at all would be the planar spiral antenna. Perhaps you should look for a web or magazine article on how to build one of this type and give it a try. It mainly involves etching a pcb.

To get back to basics for a minute, another approach, as you alluded to, was to simply de-Q a simple wire antenna. This can be done easily with a resistor across the antenna terminals, but we antenna designers refer to such techniques as "going over to the dark side" (per Star Wars philosophy) in that this is an inefficient method that burns off valuable rf energy. In an industry where noise figure is hard fought for, doing this sort of thing is the height of blasphemy. But in your case maybe not. Build a dipole designed for 1450 MHz and then put a 50 ohm resistor across its terminals, and you will have a broad band antenna. It will be somewhat mismatched at your load, but I suspect that this won't matter too much.

Another approach is to ignore trying to make the wire antenna any resonant length, and just make it relatively short, like about 40 mm or so and feed it in the center as you would a half wave dipole. Such a short dipole will not be a good impedance match to your transmission line and receiver, but it also won't be resonant below about 2800MHz and so will have a nice flat frequency response. Same can be said for a small circle of wire looped around from the center conductor of the coax to the ground braid, with a circumferance of about 40 mm. Its gain will be lousy, but maybe that's ok in your case. By lousy, I mean instead of a gain of about 0dBi, you might get a gain of -10 dBi, which might still be OK for detecting fields relatively close to the phone. This approach may be just as effective as the halfwave dipole with the resistor across it.

 

[RadioRon]

Your receiver may be as simple as an RF diode detector. By this I mean a simple AM envelope detector using a diode, followed by a high impedance DC amplifier. The key here is to use a good RF Schottky diode, such as the SMS7621, for example, and to use good RF construction practices. This means to use surface mount parts and mount them as close together as possible (like within 5 mm of each other).

You may be disappointed with the sensitivity of such a detector. It can be improved by putting a little bit of DC bias current through the diode to shift the operating point closer to the inflection point on its V-I curve. You may also need to add an RF amplifier before the detector to give yourself another 10 to 15 dB better sensitivity.

If you can't find a good detector circuit, then PM me and I can suggest something.

 

[audioguru]

Don't cell phones transmit with different power levels depending on the distance and traffic of the nearest base office? Even during each call?

 

[Nigel Goodwin]

Don't cell phones transmit with different power levels depending on the distance and traffic of the nearest base office? Even during each call?

Yes, but he mentioned that the idea was to test under different signal conditions.