RF Power Transmission and Tissue

[Stiganator]

This is related to remotely powering a brain implant from a power unit that you could put in your pocket.


I know you can power devices via RF, but I don't really understand how it works.

So, say you transmit at 2.4GHz and 1 watt. Some device can pick it up an an antenna. How does that get turned into say, 5V supply for the circuit components to use?

Does anyone know anything about RF attenuation in tissue? How can I do some ballpark calculation of the attenuation, so I know power much power the signal must have to get through the skull.

i.e. Say that implant needs 250mW, how much power does the external unit need to beam over to get that when you account for tissue attenuation.

Then, how strong of a signal can the implant produce to send data back through the skull to the base unit (presumably it can only use as much power as gets beamed to it, so 250mW).

Would attenuation be a proportional thing (1w--->0.5 watt, 2w--->1w) or is the attenuation have a barrier so, (1w ----->0.5 w, 0.25w---->0w)?

Does anyone know off of hand how power corresponds to distance in air and/or tissue for RF transmission? I would imagine it depends on power and frequency, I'm only concerned with 2.4GHz.

i.e. 1 watt@2.4 GHz will travel 10 ft before it is attenuated -3db or some such.

Could an antenna in the base power unit, be sensitive enough to pick up the attenuated transmission from the implant?

Would it be easier if the tip of the implant antenna were sticking through the skull? You wouldn't have to worry about tissue attenuation then or not as much?

 

[PottedMeat]

I believe that it's simply RF -> Antenna -> Rectification. I don't really think that a pocket power transmitter + skull receiver is practical though.

In RF classes we just used the Friis Transmission Equation: http://en.wikipedia.org/wiki/F..._transmission_equation

You need to know the gain of both antennas pointed at one another ( from their radiation pattern graphs ). I don't know the attenuation of human tissue/bone but you can probably just add a few terms to the transmission equation to adjust for that.

 

[QuixoticOne]

Now there's nothing *special* about 2.4 GHz other than it being a high UHF frequency, so it will act similarly to other similarly high frequencies.

That being said, 2.4GHz is the same frequency used in microwave ovens, and as one might expect it is highly effectively absorbed by wet tissue materials.
It is also the same frequency used by wireless LAN and some newer cordless telephones.

Although it does propagate somewhat through wet tissue, I'd suggest that for power transmission purposes you'd be much better off using a much lower frequency.

Cordless regcharging of things like razors and toothbrushes uses inductive (simple transformer like) coupling between a powerful AC signal in the base unit and a magnetic coil in the remote device. They must be installed in close proximity for best results (usually within a few of centimeters, with a couple of millimeters being ideal). That can be a highly efficient mode of power transfer that is certainly capable of sending milliwatts to watts of power over short distances.

For modeling the tissue absorbtion of the body and its skeleton, consider looking up the "SAR" (specific absorbtion rate) studies used to determine the biological heating and RF dose effects of devices like PCS cell phones on the users' bodies and heads. Such cellular phones typically operate around 1.9GHz with output powers in the 0.25 to 1 Watt range with their antennae often positioned right next to the user's head. The models and actual measurements have been conducted quite precisely and can easily quantifiably answer absorbtion / heating questions. Though they've been deemed safe, clearly some degree of question / controversy exists about that if non-thermal damage is possible.

You might consider looking into power transmission via low frequency AC (100s of Hz up to 10s of kHz) and data transmission over RF.

 

[Stiganator]

I don't want to use inductive coupling, because it does have short range and requires coil alignment. Presumably the power unit will be cellphone sized and be at your waist or in your pocket.

I recall seeing in a paper saying that lower frequencies absorb much more than higher frequencies (I think due to cell capacitances, higher ones short out, lower ones don't?). I think that is one of the advantages of higher frequencies along with the ability to have a shorter antenna since the there is more energy. Maybe I misunderstood what they wrote, I'll check?


Edit: a portion of the ~2.4GHz is classified for medical communications also.

 

[bobsmith1492]

Water resonates at about 900MHz and 2.45GHz. These are also the commercially available frequency bands and are used for just about any consumer product that requires an RF signal.

They were chosen for the very reason that water resonates at these frequencies and thus absorbs the signal very well. The idea was that, with an upcoming proliferation of RF devices, the extra attenuation would help limit interference from distant sources.

It turns out that atmospheric absorption due to H2O is not that big of a deal but the fact remains that water is extra absorbent of RF at those frequencies.

There is another, lower-frequency band available - I think they use it for remote-controlled cars and airplanes. You might check that out, too. Like you say, at lower frequencies you'll need a bigger antenna, though.

 

[CycloWizard]

I actually cited a paper on the dielectric properties of the brain and ocular tissues in one of my recent papers...

Schmid, G., Uberbacher, R., 2001, ?Age dependence of dielectric properties of bovine
brain and ocular tissues in the frequency range of 400 MHz to 18 GHz,? Phys. Med.
Biol., 50(19), pp. 4711-4720.

It's been a while since I read it, but it was a pretty straightforward paper and should at least get you going in the right direction.

 

[QuixoticOne]

There isn't a specific sharp (or broad) quantum resonance in H2O that is used by 'microwave' ovens motivating their tuning to those particular frequencies, it just takes advantage of the generally lossy / absorbent properties of wet items and similarly lossy materials at UHF/SHF frequencies. Those frequencies are just convenient in wavelength / efficiency / regulatory availability and have sufficiently high loss factors to make them practical for such uses.
Certainly the dielectric loss does increase with frequency with most materials, and it does vary with frequency otherwise. So you're right that there is loss due mostly to wet items in that frequency range, but it is a bit of a myth that water "resonates" particularly in that spectrum as compared to most any higher UHF/SHF frequency.

The frequencies you refer to are ISM (Industrial Scientific and Medical) frequencies which are covered by specific parts (e.g. part 18) of the FCC regulations to make them easily available for such uses so that such equipment won't interfere with communications activities on other frequencies.

6.765 - 6.795 MHz
# Industrial, Scientific and Medical [Part 18]

13.553 - 13.567 MHz
# Industrial, Scientific and Medical [Part 18]

# 26.96 - 27.28 MHz
* Industrial, Scientific and Medical (26.957 - 27.283 MHz) [Part 18 - Subpart C]

40.66 - 40.70 MHz
# Industrial, Scientific and Medical [Part 18]

150.0500 - 150.8000 MHz
# 150.7750 and 150.7900 - Medical radiocommunication systems [90.265]

162.0125 - 173.2000 MHz
# Secondary uses:
* 163.2500 MHz - medical radiocommunication systems [90.265]

174 - 180 MHz
# Biomedical telemetry devices, 1500 microvolts/meter at 3 meter max. [15.241]
180 - 186 MHz
# Biomedical telemetry devices, 1500 microvolts/meter at 3 meter max. [15.241]
186 - 192 MHz
# Biomedical telemetry devices, 1500 microvolts/meter at 3 meter max. [15.241]
192 - 198 MHz
# Biomedical telemetry devices, 1500 microvolts/meter at 3 meter max. [15.241]
198 - 204 MHz
# Biomedical telemetry devices, 1500 microvolts/meter at 3 meter max. [15.241]
204 - 210 MHz
# Biomedical telemetry devices, 1500 microvolts/meter at 3 meter max. [15.241]
210 - 216 MHz
# Biomedical telemetry devices, 1500 microvolts/meter at 3 meter max. [15.241]

402 - 403 MHz
# Medical Implant Communications (MICS) [Part 95 Subpart I and 95.628]
403 - 405 MHz
# Medical Implant Communications (MICS) [Part 95 Subpart I and 95.628]

470 - 476 MHz
# Biomedical telemetry devices [15.242(a)]

476 - 482 MHz
# Biomedical telemetry devices [15.242(a)]

482 - 488 MHz
# Biomedical telemetry devices [15.242(a)]

902 - 928 MHz
# Industrial, Scientific and Medical [Part 18]

# 1395 - 1400 MHz - Wireless Medical Telemetry Service [FCC News Release - June 8, 2000]

1429 - 1435 MHz
# 1429 - 1432 MHz - Wireless Medical Telemetry Service [FCC News Release - June 8, 2000]

2400-2402
# Industrial, Scientific and Medical [Part 18] - Microwave ovens

2402-2417
# Industrial, Scientific and Medical [Part 18] - Microwave ovens
2417-2435
# Industrial, Scientific and Medical [Part 18] - Microwave ovens
2435-2450
# Industrial, Scientific and Medical [Part 18] - Microwave ovens
2450-2465
# Industrial, Scientific and Medical [Part 18] - Microwave ovens
2465-2483.5
# Industrial, Scientific and Medical [Part 18] - Microwave ovens
2483.5-2491.5
# Industrial, Scientific and Medical [Part 18]
2491.5-2499.5
# Industrial, Scientific and Medical [Part 18] - Microwave ovens
2499.5-2500
# Industrial, Scientific and Medical [Part 18] - Microwave ovens
5725 - 5785 MHz
* Industrial, Scientific and Medical [Part 18]
5785 - 5815 MHz
* Industrial, Scientific and Medical [Part 18]
5815 - 5825 MHz
* Industrial, Scientific and Medical [Part 18]
5825 - 5850 MHz
* Industrial, Scientific and Medical [Part 18]
5850 - 5875 MHz
* Industrial, Scientific and Medical [Part 18]
24,000 - 24,050 MHz
* Industrial, Scientific and Medical [Part 18]
24,050 - 24,075 MHz
* Industrial, Scientific and Medical [Part 18]
24,075 - 24,175 MHz
* Industrial, Scientific and Medical [Part 18]
24,175 - 24,250 MHz
* Industrial, Scientific and Medical [Part 18]
59 - 64 GHz
# Industrial, Scientific and Medical (61 - 61.5 GHz only) [Part 18]

Originally posted by: bobsmith1492
Water resonates at about 900MHz and 2.45GHz. These are also the commercially available frequency bands and are used for just about any consumer product that requires an RF signal.

They were chosen for the very reason that water resonates at these frequencies and thus absorbs the signal very well. The idea was that, with an upcoming proliferation of RF devices, the extra attenuation would help limit interference from distant sources.

It turns out that atmospheric absorption due to H2O is not that big of a deal but the fact remains that water is extra absorbent of RF at those frequencies.

There is another, lower-frequency band available - I think they use it for remote-controlled cars and airplanes. You might check that out, too. Like you say, at lower frequencies you'll need a bigger antenna, though.

 

[QuixoticOne]

Unless you're looking for milliwatts or microwatts of power density at the location of the cranial implant you can forget a waist located power transmission antenna for power coupling purposes (unless the receiver is also near waist level).

You could do something like integrate an antenna in the fashion of a necklace, collar, hat, pillow, et. al. and have that couple power and telemetry much more efficiently to the site of a cranial implant either through inductive coupling or via RF (near-field at long wavelengths or borderline far-field at short wavelengths).

You'll have to dig into the papers more thoroughly; whole BODY loss is indeed higher at some lower VHF frequencies, if the whole body is equally illuminated.
Also penetration is generally lower at higher frequencies due to skin depth effects, scattering effects, absorbtion loss, et. al.

Though the loss per cubic centimeter depth of moist tissue generally increases with frequency. Near field (within a couple of wavelengths of the transmitter) power absorbtion differs from far field mechanisms. And absorbtion / dose effects for sources that are in close proximity to part of the body will significantly differ from cases where the body is in a more uniform field from a distant source.

Generally the penetration depth (relative to a given amount of loss) into any conductive / wet material decreases monotonically with frequency in VHF/UHF/SHF.

Check the SAR data for particular tissue types and you'll find all the details for many frequencies especially 700MHz - 5.8GHz where telephone / communications related device safety has been extensively researched.

If you want to know about antenna field power density vs. distance just do some basic notepad calculations based on field strengths and antenna patterns in free space and you'll get a rough answer to confirm that the power density falls off rapidly (even in free space) at a one meter distance. Then account for absorbtion and scattering due to tissue and you'll be reducing the power density still more. Use MEEP (free/general) or 4NEC2 (free/general) or specialized SAR related code packages if you want some more quantitative models.

IMHO you might look at just having something like a 'pillow' for night-time telemetry and close proximity charging; that would tend to get within 1" to 4" of the skull for hours a day. A hat/cap antenna would improve that still more.

http://www.icnirp.org/documents/emfgdl.pdf
http://en.wikipedia.org/wiki/Specific_absorption_rate
http://ceta.mit.edu/pier/pier.php?paper=0406251
http://www.ece.ncsu.edu/erl/ht...SU-ERL-LAZZI-00-01.pdf
http://www.lss.supelec.fr/~pub...JDS0VM_AMTA-03-115.pdf
http://irpa11.irpa.net/pdfs/8f4.pdf
http://www.ece.utah.edu/people/profiles/gandhi.html
http://ieeexplore.ieee.org/iel...73.pdf?arnumber=942573
http://www.ursi.org/B/EMTS_2007/O9-51/10-Eldeeb142.pdf
http://journals.tubitak.gov.tr...-3/elk-6-3-5-98022.pdf
http://piers.mit.edu/piersonli...FQYWdlMTA0dG8xMDkucGRm

Originally posted by: Stiganator
I don't want to use inductive coupling, because it does have short range and requires coil alignment. Presumably the power unit will be cellphone sized and be at your waist or in your pocket.

I recall seeing in a paper saying that lower frequencies absorb much more than higher frequencies (I think due to cell capacitances, higher ones short out, lower ones don't?). I think that is one of the advantages of higher frequencies along with the ability to have a shorter antenna since the there is more energy. Maybe I misunderstood what they wrote, I'll check?


Edit: a portion of the ~2.4GHz is classified for medical communications also.

 

[Stiganator]

yeah, I'm guessing fully integrated it would take 25mW to operate the implant. I just guessitmated that there would be 80% loss, figured that would be about right.

 

[Paperdoc]

It's been too long, but I did research in absorption of microwaves, VHF, UHF, etc by organic molecules. Others are right to say that water absorbs stongly in the 0.9 to 2.5 GHz range, but also outside that range, too. It is not a resonance process, such as one sees in isolated molecules in the gaseous state. What's really going on is that molecules that are not completely symetric, like water and MANY organics in tissues, have an electric dipole - that is, they have a more positively-charged end, and a more negatively-charged end. The all are rotating naturally because of their energy, and each has natural rotational speed or frequency. Now, in the liquid state they keep bumping into each other and changing energy slightly, so their rotational speeds keep changing a little; hence there is a range of rotational rates among the billions of billions of molecules in the system. But most of these happen to have a rotational rate around 1 to 5 GHz. If you apply an external alternating electric field to this situation, the electric dipoles will sort of be turned a little harder by the applied field as it swings around. If the frequency of the applied field matches the natural rotational rate of the molecules, the transfer of energy from the applied field to the molecules is more effective, and the molecules absorb energy from the field, thus decreasing the field intensity. From the electrical engineer's perspective, this is measured as attenuation of the field as it passes through the material. From the molecular perspective, the increased rotational energy they gain means more movement, which we people sense as a material with increased temperature. The microwaves heat the tissue!

Because of natural rotational speeds, this absorption is much less at much lower frequencies - say, below 100 MHz where TV, FM radio, etc. operate, and that might be a frequency range you should consider. However, that also is a closely regulated area, so you have to find out the rules. On the other side, at very high signal frequencies there also is less absorption because the molecules don't rotate that fast. However, by "very high" I am thinking of 50 to 100 GHz (maybe up to 200), and communications equipment for that area is harder to get and use.

Safety is surely a significant concern for what you are considering. In the 0.5 to 10 GHz region, as I said, the effect of the radiation field is to create local heating in particular molecues or parts of them. In something as sensitive as living human tissue in the brain, the impact of higher-than-normal temperatures, even for short times, is a great concern. So, for example, cell phones that operate in the 0.7 to 1.9 GHz range near heads are limited to a maximum output power of 500 mW, and usually operate lower. Whether they know it or not, cell phone users are implicitly agreeing to accept the increased possibility of brain damage when they use the devices.

 

[Stiganator]

SAR values have to be met this is true, 1.6 W/kg. 25mW=X*.2 so technically, our device could get by with 125mW, so I think 500mW gives a lot of leeway.

We did some thermal simulations from the chip which was very low, only 1 degree at 1 watt. But I haven't looked at the RF heat generation. I have seen similar studies though that had only marginal increases. Technically, studies have shown that brain tissue isn't damaged until 42 degrees, but that doesn't mean that it is safe.

 

[Modelworks]

If you get some time read up on Tesla.
Some of his patents are very interesting on the subject.
He was very interested in transmitting power via RF.

He actually had several patents that worked quite well.
One was able to light an entire theater with no wiring installed for the lights except a receptacle for his transmitted power for each string of bulbs.

http://www.tfcbooks.com/patents/system.htm

 

[Paperdoc]

Originally posted by: Stiganator
SAR values have to be met this is true, 1.6 W/kg. 25mW=X*.2 so technically, our device could get by with 125mW, so I think 500mW gives a lot of leeway.

We did some thermal simulations from the chip which was very low, only 1 degree at 1 watt. But I haven't looked at the RF heat generation. I have seen similar studies though that had only marginal increases. Technically, studies have shown that brain tissue isn't damaged until 42 degrees, but that doesn't mean that it is safe.

Yeah, it's the tissue heating by RF absorption that may be the most important factor, not the heat dissipated by the device itself. Don't forget that, if you consume 125 mW in the device, you will have to expose the tissue between the device and the outside of the head to much higher power. For example, applying a 500 mW beam from the outside with only 6dB attenuation of the beam as it travels through the tissue will deliver 125 mW at the device. As I understand it, the 500 mW max power level for cell phones is considered an acceptable level for frequent exposure (not quite constant exposure, the way cell phones are used), whereas your application seems to be for very infrequent exposures. The safety rules may be different for that - I don't know.

In any case, if you use a frequency much lower or higher than 2 GHz, thereby getting to a range of less intense absorption of RF power, maybe the exposure limits are different. And of course, the advantage is that the attenuation of the beam is lower, so the initial beam power also is lower to meet the same field intensity at the device.