8

To evaluate the HMPP-389T as shunt switch, it was

mounted on the test line and then the appropriate bias-

ing voltage was applied. In our prototype, the worst case

return loss was 10 dB at 5 GHz (Figure 22). The return loss

varied very little when the bias was changed from zero

to -20V.

Figure 19. Insertion loss of HMPP-389T at -20V.

The PIN diode’s resistance is a function of the bias current.

So, at higher forward current, the isolation improved.

The combination of the HMPP-389T and the SK063A

demoboard exhibited more than 17 dB of isolation from

1 to 6 GHz at If ≥ 1mA (Figure 25).

Figure 17. Return Loss of HMPP-389T

Mounted on Test Line at 0V and -20V Bias.

FREQUENCY (GHz)

-5

-15

-25

-35

-45

-55

1

6

2

4

5

3

Figure 18. Insertion Loss of HMPP-389T at

0V.

FREQUENCY (GHz)

0

-0.2

-0.4

-0.6

-0.8

-1.0

1

6

2

4

5

3

Figure 19. Insertion Loss of HMPP-389T at

-20V.

FREQUENCY (GHz)

1

6

2

4

5

3

0

-0.2

-0.4

-0.6

-0.8

-1.0

Figure 20. Isolation at Different Frequencies

with Forward Current as a Parameter.

FREQUENCY (GHz)

-10

-14

-18

-22

-26

-30

1

6

2

0.15 mA

0.25 mA

0.5 mA

1 mA

1.5 mA

20 mA

4

5

3

Figure 18. Return loss of HMPP-389T mounted on test line at 0V and -20V

bias.

Normalization was used to remove the pcb’s and connec-

tors’ losses from the measurement of the shunt switch’s

loss. The active trace was divided by the memorized

trace (Data/Memory) to produce the normalized data.

At zero bias, the insertion loss was under 0.6 dB up to 6

GHz (Figure 23). Applying a reverse bias to the PIN diode

has the effect of reducing its parasitic capacitance. With a

reverse bias of -20V, the insertion loss improved to better

than 0.5 dB (Figure 24).

Figure 18. Insertion loss of HMPP-389T at 0V.

Figure 20. Isolation at different frequencies with forward current as a

parameter.

The combination of the HMPP-389T and the demo-board

allows a high performance shunt switch to be constructed

swiftly and economically. The extremely low parasitic

inductance of the package allows the switch to operate

over a very wide frequency range.

Figure 16. Insertion Loss of Reference

Line.

FREQUENCY (GHz)

0

-0.2

-0.4

-0.6

-0.8

-1.0

1

6

2

4

5

3

Figure 16. Insertion loss of reference line.

Insertion loss of the reference was very low and generally,

increased with frequency (Figure 21). If the demo-board

has been constructed carefully, there should not be any

evidence of resonance. The reference line’s insertion loss

trace can be stored in the VNA’s display memory and

used to correct for the insertion loss of the test line in the

subsequent measurements.