The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> can bond wire be used as on-chip antenna?
https://designers-guide.org/forum/YaBB.pl?num=1188952464

Message started by sapphire on Sep 4^th, 2007, 5:34pm

Title: can bond wire be used as on-chip antenna?
Post by sapphire on Sep 4^th, 2007, 5:34pm

I am planning to use bond wire to serve as on-chip antenna? One end of bond wire will be attached to output pad, while the other end will point to the air. The measured output power level with Spectrum Analyzer is about -10dBm. The receiver would be placed in several meters range.

Can the bond wire be good monopole or dipole antenna, and works well?

Thanks

Sapphire

Title: Re: can bond wire be used as on-chip antenna?
Post by didac on Sep 5^th, 2007, 7:05am

Hi sapphire,
Assuming that you are at a high frequency(λ/4 at 15GHz are 5 mm which is a long bondwire) and you are using your IC as chip on-board without any epoxy above, I think that several aspects will limit the performance of this kind of antenna:
1)This antenna is by definition a monopole,not a dipole, you are taking the feeding point at the contact point not in a balanced manner in the middle of the antenna. The monopole need a ground plane under it and what you will have is the IC and probably the first layer of dielectric material of the PCB(unless you put a ground plane just under the chip that can complicate the PCB design).
2)Adjacent bondwires can modify the impedance of the antenna due to the proximity, this is important to take into account.
3)From a electromagnetic point of view I don't know in which angle the bondwire will be floating, a monopole has a radition null along it's axis, also I don't know if the bonding pad will suffer from mechanical stress. Another possibility is that the floating bondwire could move and short-circuit to adjacent bondwires.

The most common integrated antennas I've seen are from the printed family(using microstrip), try a fetch at IEEE and take a look

Hope it helps,

Title: Re: can bond wire be used as on-chip antenna?
Post by sapphire on Sep 7^th, 2007, 5:16pm

Hi Didac,

Appreciate for your insightful reply!! I will try to use bondwire for testing first. Actually my design is only 2.4GHz VCO with some modulation. But the receiver can be very close to the antenna. I saw a paper for similar application just use the big on-chip inductor for transmitting RF signal. Their carrier frequency is even lower, around 500MHz.

By the way, what kind of bondwire material you think it's better? Gold or Aluminum? How about thickness? how to control the antenna impedance roughly? Thank you very much! My chip will be embedded into rodent's head, so I want to make it as small and light as possible.

Le

didac wrote on Sep 5^th, 2007, 7:05am:

Title: Re: can bond wire be used as on-chip antenna?
Post by didac on Sep 9^th, 2007, 12:11pm

Hi sapphire,
Could you please post the reference to this paper you say?-it's always good to learn something new. Without more info I will make a wild guess that is an RFID tag or a biomedical device but I'm curious anyway about this paper.
Now let me make some comments:

1)Coils as antennas are used usually at "low frequencies" due to the problems of making a physical length equivalent to a fraction of a wavelength, but their efficiency tends to diminish with increasing frequency, so what it works at 500MHz not necessarily needs to work at 2.4GHz,are they using the chip in the near field region?-distance<λ/2π?.
2)If you could go for Gold I think it's better over aluminum,less resistivity translates into a lower loss resistance that dissipates power in the antenna instead of radiating it.
3)The rule of thumb that I think it's valid is five skin depths(If i did correctly the numbers the skin depth for Gold at 2.4GHz is more or less 1.6um so 8um thick will be ok),making more thick the bondwire won't help to reduce much more the resistance. I suppose the more limiting factor is the availability of a bonding machine that cab handle different thicks.
4)To control roughly the impedance you should take into account what surrounds the antenna, most important in this case adjacent bondwires. Also packaging and the environment could modify the behaviour(radiation pattern) of the antenna, so I think you should make use of a 3D-EM simulator and model as closely to reality as you can.
5) in your previous post you say that the receiver would be at several meters range, and now you say that it can be very close, take this into account when modeling the antenna(near field/far field behaviour).

As always any comments and further discussion will be appreciated.

Hope it helps,

Title: Re: can bond wire be used as on-chip antenna?
Post by sapphire on Sep 11^th, 2007, 11:49am

Hi Didac,

Just saw your reply. Very insightful and helpful. I don't have much experience in antenna-related stuff, but I will try several ways to see the effect. The paper is "A Low-Power Integrated Circuit for a Wireless 100-Electrode Neural Recording System", JSSC, vol. 42, no. 1, January 2007.

Thanks a lot again

Sappire

Title: Re: can bond wire be used as on-chip antenna?
Post by didac on Sep 14^th, 2007, 2:02am

You're welcome sapphire, was a pleasure. Thanks for the reference, pretty interesting to read, thank you very much.

Title: Re: can bond wire be used as on-chip antenna?
Post by carlgrace on Sep 19^th, 2007, 1:31pm

It amazes me there is so much research in high-tech ways to torture small animals.