Fig. 1.Initial direct connection between VNA and detector coil.

Transcription

1 Date: Tests were conducted to evaluate the potential for use of some types of solid-state semiconductor switches in switching of biosensor detector coil signals. Four (4) different solid-state RF switch types were purchased representing 3 different semiconductor types/materials/structures and we needed to test/demonstrate their RF bidirectional performance, crucial for our S11/Reflectance detector/sensor scanning scheme. The detector coil used in this first series of tests was a cylindrical solenoid detector coil. First, a direct connection was made between a N2PK-style VNA and the detector coil with a 1mm magnetoelastic (ME) sensor. A magnet field bias was adjusted to maximize the coupled ME signal amplitude at resonance (see Fig. 1).

2

3 Fig. 1.Initial direct connection between VNA and detector coil. On this date, two of the four RF solid-state switches recently purchased were tested. One of the two switches exhibited strong, clear consistent and repeatable signals acquired as the signal source was switched from channel-to-channel. Tested successfully was a SP6T Honeywell HRF-SW1030, SP6T non-reflective (internally terminated when off) switch seen in Figure 2. The insertion loss at our 2.24 MHz range is only ~ -0.5dB. It uses a SOI (Silicon-On-Insulator) CMOS architecture which allows them to be DC coupled thus allowing a bidirectional RF path. The evaluation board shown in Fig.2 with SMA connectors is very large in comparison to the small IC chip but makes evaluation convenient. The chip uses 5V Vdd and 0V Vss power pins, and 5V TTL (CMOS) control pins. After building a small interface board, the RF switch was inserted between the VNA and the ME detector coil assembly as shown in Fig. 3. Six (6) comparative traces through all 6 RF channels were acquired as the RF channels were switched.

4 Fig. 2. RF solid-state SP6T switch, Honeywell HRF-SW1030 (left) shown with small interface board (right).

5 Fig. 3. Test setup with first switch (Honeywell HRF-SW1030) was inserted.

6

7 Figure 4 is a composite of all 6 VSWR (proportional to S11) signals through the 6 RF switch channels. The vertical scale/div is constant though the scale reference was shifted slightly for some graphs. Figure 4 is a larger view of the signal through the RF1 channel. Fig. 4. Composite of the 6 VSWR traces obtained from biosensor coil/sensor pair via each of the RF switch s 6 channels shown with each the same vertical scale/div. Figure 5. Typical VSWR trace, this one switched through RF channel #1. In addition to the Honeywell HRF-SW1030 'On' state signal throughput, traces of the 'Unpowered' and Off states were acquired for Isolation verification with resultant VSWR s of >1600 and >730 respectively (see Fig. 6).

8 Figure 6. Traces of 'Unpowered' and Off states were acquired for Isolation verification with resultant VSWR s of >1600 and >730 respectivel. The second switch type was a Hittite HMC253LC4, a SP8T GaAs MMIC non-reflective switch (see Fig. 7). In the initial tests, no ME resonant signal could be detected. Though it frequency range specification is listed as DC-3.5 GHz, its specification sheet calls for "DC blocking capacitors" on all RF ports. The evaluation board includes these and are 100 pf each and at 2.24 MHz, the capacitive reactance is 710 Ohms...this may possibly be why. There are no guidelines or recommendations in their specs regarding 'DC blocking capacitor' sizes. If the capacitor could be changed to ~ 5-10 nf, then it may work. Fig. 7. Hittite HMC253LC4 (SP8T) on evaluation board with 100 pf blocking capacitor. Since only 1 Honeywell was purchased, and the Hittite did not work, a complete X-Y detector coil matrix was not demonstrated on this date. Also purchased, two Analog Devices DPDT switches which should work fine but used a different control voltage scheme than the 5V Vdd used today, have not been tested. We may use one of those to demo a 2x2 (AD x AD) or 2x6 (AD x Honeywell) matrix later, time permitting.