Why is my RFID range so low?

[Mad ID]

Hello,
I'm developing a 134kHz HDX (frequency modulation) RFID reader which must be able to read spiral tags like the one in the picture.


I'm using EM4095 reader IC. The antenna is 1.5mH and is also of solenoid type because I figure it's the only way to read a tag like the one in the picture which is aligned in parallel with my reader antenna.

The space inside is free air, length 65mm, diameter 15mm.

My reading range is 4cm which is bad ,but good enough for my application. To be honest, I don't have a clue how to get a better reading range with a tag of this size? What's the optimal antenna diameter if tag diamater is cca 3mm? From my measurements, bigger diameter antenna has a better range.

The problem is that I have to enclose the antenna in a thermoplastic potting case (Polyoxymethylene (POM) or polyacetal) and then fill two component polyurethane resin inside. When I do that my reading range drops TWO TIMES

First, the frequency changes from 134 to about 124kHz, but I offset this by tuning the frequency to 144 before potting. The problem is even when the frequency is 134 or close, my reading rage is lower than before potting.

When I enclose the antenna in the potting box, reading range falls from 4cm to 3cm even without potting?

Here are my questions:
1) Why does the frequency change? I know that 10kHz RLC metar does not see a frequency change, but 100kHz meter reads 10kHz drop. It confuses me because both thermoplastic and resin are not magnetic materials, i.e. their permeability is close to 1 like air.

2) Even with the frequency good, the reading range is much lower. Why?!

If you can give me any clue of what to investigate I would be grateful.

Thanks.   

[PA0PBZ]

Your antenna is wrong, you should really use something like this:



Also, have a look at page 7 of the application note to see how to calculate the antenna properties, 1.5 mH seems too much:

http://www.emmicroelectronic.com/sites/default/files/public/products/datasheets/an404.pdf

[Mad ID]

Your antenna is wrong, you should really use something like this:



Also, have a look at page 7 of the application note to see how to calculate the antenna properties:

http://www.emmicroelectronic.com/sites/default/files/public/products/datasheets/an404.pdf

Nop, with the antenna you suggest zero field lines pass through the tag's solenoid. Please take note that the tag's solenoid is mounted parallel in respect to my reader and this cannot be changed.

[PA0PBZ]

Nop, with the antenna you suggest zero field lines pass through the tag's solenoid. Please take note that the tag's solenoid is mounted parallel in respect to my reader and this cannot be changed.

I don't think you are supposed to use them like that, so that could be your problem:



Did you try a ferrite core?

[T3sl4co1l]

You can't put two solenoids beside each other and expect them to couple any more than a few diameters, if that.  So, your range is so low because you made your coil so small.

A large loop coil is the only way to achieve good coupling over any distance.  And yes, the coils need to be on a common axis.

If you can't tilt the TX loop into a position where it couples into the RX device, perhaps you can make it so large that the loop can fit entirely over the object.  Compare: car-detection loops embedded in roadways, or those sensors placed near doorways at stores, libraries, etc.

Tim

[Mad ID]

I don't think you are supposed to use them like that, so that could be your problem:
Did you try a ferrite core?

I understand your point, but unfortunately the tags I have to read are 90 degrees rotated with respect to your picture. There is a reader from one firm (I cannot get it) which can read these tags from 10cm+. I don't have a clue how can they do it ?!

I have not tried ferrites, would it make any difference? I had an idea to shape the field somehow but it all failed because I can direct the filed to one side but the reading range always drops.

You can't put two solenoids beside each other and expect them to couple any more than a few diameters, if that.  So, your range is so low because you made your coil so small.

A large loop coil is the only way to achieve good coupling over any distance.  And yes, the coils need to be on a common axis.

If you can't tilt the TX loop into a position where it couples into the RX device, perhaps you can make it so large that the loop can fit entirely over the object.  Compare: car-detection loops embedded in roadways, or those sensors placed near doorways at stores, libraries, etc.
Tim

Again, it common sense to have a common axis and in this case there's nothing I can do. I suspect that the company made it like this on purpose so it is difficult to change the reader. They read 10cm+.

My antenna size is limited. The thing is that 4cm is enough but the enslosure and potting reduces the range further and that's my main issue at the moment. Do you have any idea why would it happen?

[Fred27]

I'll be very interested in seeing how you get on. I've got a similar 13.56MHz NFC tag implanted in my hand. Most off the shelf readers use a PCB antenna that's intended for cards. To read I have to have my hand right against the trace. I'd like to make an antenna that's tuned better to the small coil, but that's not my area of expertise.

[PA0PBZ]

I'm not sure that ferrite will solve your problem completely, but I feel like it is worth to give it a try.
Potting probably decreases the Q of your circuit, and looking at the way you wound the 'antenna' I think the Q is not very good to begin with.

About the detuning: I guess you have an external C to tune the LC combination to 134 Khz? So all you need is a different C to get the stuff back on frequency.
Or are you just using the coil to get to 134 Khz? 

Get a ferrite rod, wind it nicely and measure the inductance. Try to get something in the ballpark of the application note (300-800 uH) and calculate the C you need.
Then test the range again before potting.

[Mad ID]

I'm not sure that ferrite will solve your problem completely, but I feel like it is worth to give it a try.
Potting probably decreases the Q of your circuit, and looking at the way you wound the 'antenna' I think the Q is not very good to begin with.
About the detuning: I guess you have an external C to tune the LC combination to 134 Khz? So all you need is a different C to get the stuff back on frequency.
Or are you just using the coil to get to 134 Khz? 
Get a ferrite rod, wind it nicely and measure the inductance. Try to get something in the ballpark of the application note (300-800 uH) and calculate the C you need.
Then test the range again before potting.

Hello. You are saying that the solenoid has to be wounded perfectly (how?) to get the best antenna, i.e. none back and forth? I'm doing it by hand on a piece of syringe The turns are not the best.. Can you point me to some literature about how to wound it properly? I'm tuning the circuit with a capacitor bank. The problem is that the electronic module which contains the capacitor bank is potted at the same time (believe me, it's not possible to do the antenna first in my case) and I cannot change it afterward.

Your theory is that the Q droops so much that there isn't a nice frequency peek at 134kHz but it's spread over a large bandwidth? This would explain the droop in the reading range... Is it possible that the resin increases parasitic capacitance somehow between the individual turns (dielectric=4 vs 1 for air) and that this decreased the Q? Can low Q be the reason the 100kHz RLC meter reads a droop in inductance?

What Q should I aim for?

Another question, if I understand correctly the larger the L, the stronger the field for a given drive current. That's the reason my L is 1.3-1.5mH, because I want the strongest field. Downside is that the voltage developed will be large but my caps can handle it.

Thanks. 

[Mad ID]

I'm not sure that ferrite will solve your problem completely, but I feel like it is worth to give it a try.
Potting probably decreases the Q of your circuit, and looking at the way you wound the 'antenna' I think the Q is not very good to begin with.
About the detuning: I guess you have an external C to tune the LC combination to 134 Khz? So all you need is a different C to get the stuff back on frequency.
Or are you just using the coil to get to 134 Khz? 
Get a ferrite rod, wind it nicely and measure the inductance. Try to get something in the ballpark of the application note (300-800 uH) and calculate the C you need.
Then test the range again before potting.

Hello. You are saying that the solenoid has to be wounded perfectly (how?) to get the best antenna, i.e. none back and forth? I'm doing it by hand on a piece of syringe The turns are not the best.. Can you point me to some literature about how to wound it properly? I'm tuning the circuit with a capacitor bank. The problem is that the electronic module which contains the capacitor bank is potted at the same time (believe me, it's not possible to do the antenna first in my case) and I cannot change it afterward.

Your theory is that the Q droops so much that there isn't a nice frequency peek at 134kHz but it's spread over a large bandwidth? This would explain the droop in the reading range... Is it possible that the resin increases parasitic capacitance somehow between the individual turns (dielectric=4 vs 1 for air) and that this decreased the Q? Can low Q be the reason the 100kHz RLC meter reads a droop in inductance?

What Q should I aim for?

Another question, if I understand correctly the larger the L, the stronger the field for a given drive current. That's the reason my L is 1.3-1.5mH, because I want the strongest field. Downside is that the voltage developed will be large but my caps can handle it.

Also, I can use thicker wire to increase the Q, would it make any sense?

Thanks.

[Bud]

... I've got a similar 13.56MHz NFC tag implanted in my hand.

Can we ask why ?

[Fred27]

... I've got a similar 13.56MHz NFC tag implanted in my hand.

Can we ask why ?

Just for fun. I can unlock my garage door and phone. When you scan my hand with an NFC enabled phone it opens this page. (I did have it transferring my contact details but changed it.)
http://0xfred.wordpress.com/2014/05/23/my-nfc-implant/

I'll do more with it when I find the time.

[kt315]

As many fun things on this forum,  RFID implants are known to cause cancer.

[miguelvp]

As many fun things on this forum,  RFID implants are known to cause cancer.

On mice that are scanned often. But now I'm going to keep an eye on both my dogs because they are rescued dogs and we had to chip them, although they only get scanned if we change vets. So the chips have only been active around 3 times over the last 6/7 years.

[Fred27]

As many fun things on this forum,  RFID implants are known to cause cancer.

I quick search shows this was only observed in mice. Apparently mice get cancer far more easily than other animals and the chip is relatively speaking far bigger. Nothing reported in the far larger number of cats and dogs. I'll take the risk, but thanks for the info.

[Neganur]

As many fun things on this forum,  RFID implants are known to cause cancer.

I truly hope you're not referring to the article in The Washington Post from 2007 quoted by Wikipedia. You may as well claim that breathing is known to cause cancer.

Don't get me wrong, I'd gladly support your claims if the evidence was good enough. But as always, words are being bent to claim whatever people want to promote. Even if foreign objects injected into a human body were 100% proven to cause tumors, a tumor is not automatically malignant. (I'm pointing out that one research claims that foreign objects cause tumors but it does not say if benign or cancerous. So is the glass half full or half empty?) It also seems to me that it is being emphasized that it is the microchip that causes the cancer. Why not the RF-transmission or something else entirely?

I don't like how some of those researchers claim that "cancer is a really bad disease. Life-threatening" and then try to link it to "microchips cause cancer" while no real controlled tests were conducted to prove the link.

In the European Nordic countries, a lot of animals and pets had those glass encapsulated tags injected in the past. I find it a bit far fetched to claim that all of them will develop cancer due to said implant.

[Neganur]

Regarding your antenna. If memory serves, you should be able to get a reading distance of up to some 30 cm- more if your antenna is very large.
You definitely want a ferrite core for your reader antenna and the correct gauge & number of turns are critical for optimal results.
Make sure you only have _one_ very neat layer of wire wound. You might even want to tune the antenna exactly for the frequency you need with a voltmeter and remove/add turns until you get the best result at your resonance frequency.

EDIT: not the link that I had in memory but it appears useful anyway (verichip reader diy attempt): http://cq.cx/vchdiy.pl