This page contains information about building your own WISPs.
S1000 - Direct from Seiko
HSMS-285C - Direct from Avago *
NLSV1T244 - Direct from On Semi
NTK3134N - Mouser
Inductor - Direct from Coil Craft
ESD5Z - Direct from On Semi
Maxcap - Direct from manufacturer
* Sometimes available from distributors. Try searching www.octopart.com
Note: many parts can be obtained as samples in quantities up to 100, which is fine for WISP prototyping.
1×50 Male:
S9014E-50-ND (cheaper & scored)
ED8250-ND (not scored)
1×50 Male Rt. Angle:
ED8650-ND
1×50 Female:
S9008E-50-ND, ED90265-ND (not scored)
1×50 Female Rt. Angle:
ED8850-ND
More connectors here: Fine Pitch Connectors
Material Type: FR4
Copper Weight (inner): 1oz
Copper Weight (outer): 2oz
Finish Thickness: 31mil (0.031“)
Generally each batch needs to be matched using a VNA due to changes in the design and/or process. An inductor kit from CoilCraft is handy for this task. High-Q inductors are critical for achieving high RF sensitivity.
For tuning the front end, mechanical trim caps can be really handy. You pick an inductor, and them tune the cap to resonate at your desired frequency. For example, digikey part numbers: GKRP10066 GKRP20066
More Details and Tutorial on how to tune the RF Front End here.
It is 14.6cm in total length. The pcb extends about 1mm beyond the antenna. The copper is only on the top side of the pcb.
Question:
Hi! I'm a reasearcher of T3LAB of Bologna (Italy). I'm interested in creating a UHF battery-assisted tag for RF-ID. I saw WISP project and I trying to understand how the RF circuit works. I use OrCAD Capture (PSpice) for building a similar circuit, but I can't find the BF1212Wr component. On NXP site this componenst has 2 MOSFET + 2 diodes, but in WISP schematics there's only a N-channel MOSFET. Why?
Could you help me in re-building the RF circuit? I will develop a project on 868 MHz (EU UHF RFID).
Answer:
You are correct the BF1212WR is a dual gate NMOS with 2 diodes.
In the schematic we did not draw the second diode or redundant gate (instead we defined both pads in the footprint to be connected to pin 3). In the layout images you can see that both gates are connected to the ‘Transmit’ net.
Also the BJT Q1 (BFT25A) is not necessary. We wanted the option to using either part on the WISP production runs. Thus, we solder only one of the two parts on the boards, but never both.
There is an impedance matching network on the WISP that matches the antenna to the analog front end. See that schematic of the AFE (page 2)
http://wisp.wikispaces.com/file/view/WISP4.1DL_Schematic.pdf
The antenna terminals are contacted to ‘gnd’ and to the signal labeled ‘Ant’. (Side note: RFID tags do not need to use a baluns since they are floating with respect to earth ground, and capacitive coupling to earth ground is assumed to be very small.)
The inductor ‘L1’ and capacitor ‘CV1’ create a L-match network that transform the impedance of the antenna (~70 ohms) to the input impedance of the rectifier, denoted as signal ‘Feeder’ referenced to gnd.
The values used for the impedance matching network depend on heavy on the dielectric constant of the FR4 used, the layout, and parasitic capacitance. We typically need to recheck the impendence for each run of PCBs, even when we’ve ask for the same board build up. For the last years batch of WISP4.1 boards the L-match values are (7.5nH and 8.0pF). Since you fabbed your own WISPs the you will have to measure your impendence.
Additionally you should also measure the impedance of your antenna. Since the dielectric constant of the FR4 will affect the electrical length of the dipole you may find the resonator frequency is a little off. Measure the antenna’s impedance at the center frequency (916MHz in the US), and match the WISP’s AFE to the complex conjugate of the antenna.
We should be getting the next batch of WISP back for assemble in the next week. I will need to verify the quality of the matching network and will write up a tutorial on impedance matching.