Network Evaluation for the PW A10 Schematic Review Prepared by David Green; W7NE Revision: 1.0 Complete Friday, October 20, 2006

Transcription

1 Network Evaluation for the PW A Schematic Review Prepared by David Green; W7NE Revision:. Complete Friday, October 2, 26 Parametric analysis of the L network was conducted in an effort to understand the electrical properties in the system over frequency. Use of a network analyzer to achieve this goal was utilized. As referenced from the PA A schematic diagram, evaluation of the network consists of a combination of inductance, capacitance and resistive system elements. The primary function of the network serves to provide a DC bias to the plate of V while passing RF energy into the user configurable tank for VSWR matching purposes. Observation of the tank, and in particular element L reveal a structure typical of handling both low and high frequency elements. In particular, element L2 with the associated resistive elements is key to lowering the overall Q directly seen by the plate of V in an effort to ensure that VHF and/or UHF oscillation does not occur. L in conjunction with C25 shunt and C2 series makes up the primary mechanism for providing some form of impedance looking back into the +5 VDC supply source for RF isolation. Evaluation of L was not considered as part of this analysis. Observation of L reveals the impact of arcing in one section of the full inductor. In particular when addressing the section that impedes RF current flow closely associated with L2. Arcing stems from a number of factors, clearly of which excessive voltage buildup in conjunction with ample RF current leads to fusing of the wire assembly followed by open conditions rendering the amplifier useless. The experimental setup consists of sweeping the frequency response as shown in the diagrams below. Primarily single port analysis was accomplished, but dual port was used as connecting to what the V plate would see and what L (port 2) would see in the system. In reality it is highly unlikely that the 5 ohm system presented by the network analyzer is characteristic of the actual environment, but it is sufficient enough to provide a fairly accurate reading of the element response. While emphasis should be on the HF response Kc to Mc, calibration and review up through the 2M band (5Mc) was equally conducted. This is useful to see how the network responds at these frequencies that can be necessary when evaluating response to harmonic energy in the system. Section I Inductor L Evaluation

2 Figure Schematic Configuration Actual implementation of L is composed of two inductive sections where some mutual coupling must take place. My suspicion is that separating the winding into sections helps extend the frequency response. This analysis looks at L in the above configuration. With the help of a near duplicate L further analysis on one section breakout is accomplished. Due to arcing, etc., one section of L could not be electrically analyzed. However, this burnt section was unwound to ascertain number of turns and length. S Parametric Measurement.8 57 uh.5 62 uh 66 uh 7 2 uh uh 2 pf 2 pf pf pf pf. pf Table Measured Inductance response over 6 to M HF band operation The ability to report in reactance at frequency serves as one of the primary purposes of the network analyzer. Note that the inductor is designed to operate as an inductor from the 6M band up through the M band. Higher frequency input confirms the inductor simply behaves as a capacitor and would be ineffective in this function if other elements of the network were not present.

3 db -.2 :S Refl Port Log Mag.2 db/ Ref -.8 db C? Meas:Mkr 8.7 MHz.9dB : Stop MHz :Mkr (MHz) db 2:Mkr (MHz) db : :.5.5 : 7..8 > Figure 2 Log Mag Response

4 db :S Refl Port Log Mag.2 db/ Ref -.8 db C? Meas:Mkr MHz -.27dB 2 : Stop MHz :Mkr (MHz) db 2:Mkr (MHz) db : :.5.5 > : Figure Log Mag

5 25 :S Refl Port Phase 5. / Ref 5. C? Meas:Mkr MHz : Stop MHz :Mkr (MHz) Deg 2:Mkr (MHz) db : :.5.57 > : Figure Phase response Note that a resonant condition exhibits itself through the zero phase crossing at 8.6Mc.

6 k :S Refl Port Imped 5 k / Ref k C? Meas:Mkr 8.67 MHz 7.95k : Stop. MHz :Mkr (MHz) Ohm 2:Mkr (MHz) db :.8.777k 2: k > k : k Figure 5 Impedance Response Note that the resonant condition is further characterized by the rapid increase in impedance shown at 8.6Mc. This parallel resonate condition should NOT exist for any operational mode of HF band operation. As such, the coils are wound to ensure positioning of this between the M and M ham band. I don t know how stable the amplifier works on the M band due to increase choke impedance at this frequency. Detailed Evaluation of individual L Elements Inductor L is composed of two separate inductive windings on a ceramic rod. For the purposes of this evaluation, two L inductors are used. For the majority of measurements in this report the known good inductor is used for evaluation. L, being composed of two inductors respectively named LA and LB are called to conduct the following analysis. Note that mutual coupling exists between the two inductors and the results of this analysis is shown above as the L response. Inductor L section LA attaches directly to the L2 inductive network and also sees capacitor C2. The other port of LA connects to LB.

7 Inductor L section LB connects to LA on one side and connects to capacitor C25 and inductor L by referencing the schematic. Observation on the second L inductor indicates burns on LA. There is no apparent burns or shorting on LB. This leads me to believe that LA sees the RF hot side and in cases of excessive reflection in combination with RF power output results in arcing of the inductor. LA is determined to be open and therefore measurements to obtain the actual inductive response of this section cannot be determined. However, an attempt to unwind and count turns will be attempted. Number of Counted Turns on L LA 7 Turns feet wire length LB Table 2 Number of Turns Count for L Due to the destructive nature of these tests, LB has not been unwound while LA was unwound because the element was technically destroyed by arcing. Analysis of section LB S Parametric Measurement.8 76 uh.5 8 uh 86 uh 7 8 uh uh 66 uh pf pf pf pf.7 pf Table Parametric Results of LB Over frequency

8 db -.2 :S Refl Port Log Mag.2 db/ Ref -.8 db C? Meas:Mkr. MHz -.26dB : Stop 5. MHz :Mkr (MHz) db 2:Mkr (MHz) db : : : > Figure 6 LB S Log Mag

9 5 :S Refl Port Phase 5. / Ref 5. C? Meas:Mkr. MHz : Stop 5. MHz :Mkr (MHz) Deg 2:Mkr (MHz) db : : : > Figure 7 LB Phase Resonant point approximately Mc (2M band) for this section. This stands to reason as there is less inductance so an expectation of having higher resonance point can be expected.

10 k 2 :S Refl Port Imped 2 k / Ref 6 k C? Meas:Mkr.2 MHz.6k 8 : Stop 5. MHz :Mkr (MHz) Ohm 2:Mkr (MHz) db : : k : k >.2.6k Figure 8 LB Impedance Response

11 :S Refl Port Smith 2 U FS C? j5 j25-25 Meas:Mkr.267 MHz.5k k 552.fF j j -j - -j j25 -j5 Stop 5. MHz :Mkr (MHz) Ohm Ohm 2:Mkr (MHz) db : : k : k >.267.5k k Figure 9 LB S Smith analysis Section II Inductor L2 Evaluation The purpose of this inductor is to serve the high frequency portion of the HF band from M through M.

12 Note that Port of the network analyzer is hooked to the top of the tank. S Parametric Measurement.8 58 nh.5 nh 2 nh 7 5 nh 7. nh nh 29 nh nh nh nh 2 nh Table L2 Network Inductive Response Note that this inductor looks and behaves like an inductor through the full HF range. No suck-outs or peaking occurs. This is used to complement the L response especially as the frequency increases above the natural resonant point of L and serves to provide inductive isolation to V at high frequencies.

13 :S Refl Port Smith 2 U FS C j5 j25-25 Meas:Mkr2. MHz nH j 2 - j j -j j25 -j5 Stop 25. MHz :Mkr (MHz) Ohm Ohm 2:Mkr (MHz) db :.8 -.2m.79 2> : m 7.25 : Figure Smith response of L2 Network

14 db :S Refl Port Log Mag.2 db/ Ref -.8 db C? Meas:Mkr MHz -.5dB : Stop. MHz :Mkr (MHz) db 2:Mkr (MHz) db :.8.6 2> : :. -.9 Figure Log Mag of L2 Network

15 :S Refl Port Phase 5. / Ref 55. C? Meas:Mkr MHz. : Stop. MHz :Mkr (MHz) Deg 2:Mkr (MHz) db : > : :.. Figure 2 Phase Response

16 :S Refl Port Imped 5 / Ref C? Meas:Mkr MHz : Stop. MHz :Mkr (MHz) Ohm 2:Mkr (MHz) db : > : :. 2.8 Figure L2 Network Impedance Response Section III Inductor L & L2 Evaluation This section reviews the combination of L2 and L network elements. In order to minimize inductive coupling, the inductors are set perpendicular to one another. No further attempts to change coupling ratios due to non-orthogonal configurations is attempted. Capacitor C25 was hooked to ground and the same test rerun. It is determined that the inclusion of C25 does not change the overall network response as was predicted from

17 general observation of the circuit. C25 is considered fat (low impedance) when it comes to frequency evaluation. Figure L2 and L Network Measurement Configuration S Parametric Measurement.8 58 uh.5 68 uh 76 uh 7 22 uh uh.55 pf 2. pf 8.5. pf pf 2.5 pf. pf Table 5 Parametric Results

18 db :S Refl Port Log Mag.2 db/ Ref -.6 db C? Meas:Mkr. MHz -.9dB 2 : Stop. MHz :Mkr (MHz) db 2:Mkr (MHz) db : : : >. -.9 Figure 5 Log Mag Response

19 25 :S Refl Port Phase 5. / Ref 5. C? Meas:Mkr 8.67 MHz : Stop. MHz :Mkr (MHz) Deg 2:Mkr (MHz) db :.8.9 2: > : Figure 6 Phase Response

20 k 25 2 :S Refl Port Imped 5 k / Ref k C? Meas:Mkr 8.67 MHz 2.5k 5 : Stop 5. MHz :Mkr (MHz) Ohm 2:Mkr (MHz) db : k 2: k > k :..588k Figure 7 Impedance Response

21 :S Refl Port Smith 2 U FS C? j5 j25-25 Meas:Mkr.8 MHz k 57.7 H j - - j j -j j25 -j5 Stop 5. MHz :Mkr (MHz) Ohm Ohm 2:Mkr (MHz) db > k 2: k : k k : k Figure 8 Smith Response Section IV Full Two Port Response This section looks at the full network in a 5-ohm source and loading. Understand that the actual parameters of the circuit most likely has differing source and load impedance, but the story in the results is nevertheless valuable. Further, the coupling capacitor that feeds Port 2 in figure 5 will see radical range of input impedance due to the loading C/L network loading.

22 Figure 9 2-Port Configurations In this configuration, Port serves to connect where the plate of V 872 connects. Port 2 serves as the output that coupled into L and SB for the overall schematic of the L/C tank. Port and 2 a configured as 5 ohm source and load.

23 :S Refl Port Smith 2 U FS C j5 j25-25 Meas:Mkr 7. MHz nH j j -j -j j25 -j5 Stop 5. MHz :Mkr (MHz) Ohm Ohm 2:Mkr (MHz) db : : : > Figure 2 Full range S view of network S sweep of the reflected impedance on the port.

24 :S22 Refl Port2 Smith 2 U FS C j5 j25-25 Meas:Mkr 7. MHz nH j j -j -j j25 -j5 Stop 5. MHz :Mkr (MHz) Ohm Ohm 2:Mkr (MHz) db : : : > Figure 2 Full range sweep S22 sweep of the impedance on Port 2 looking back into the network.

25 data? 8 :S2 Fwd Trans Phase 2. / Ref. C? Meas:Mkr 7. MHz : Stop 5. MHz :Mkr (MHz) Deg 2:Mkr (MHz) db : : : > Figure 22 Network Phase Response

26 6 :S2 Fwd Trans Imped 2 / Ref 8 C? Meas:Mkr 7. MHz : Stop 5. MHz :Mkr (MHz) Ohm 2:Mkr (MHz) db : : : > Figure 2 Impedance Response Note the smooth response that on average runs around 2 ohms for the HF bands of operation.

27 db :S2 Fwd Trans Log Mag.5 db/ Ref -2.5 db C? Meas:Mkr 7. MHz -.57dB 2 : Stop 5. MHz :Mkr (MHz) db 2:Mkr (MHz) db : : : > Figure 2 Log Mag response This represents a smooth attenuation response. With the least impact on isolation in the network existing somewhere between the 8 and M band sections of operation. Summary Clearly operational response of the full network makes much better sense rather than looking at individual components of the network. Operation of the network is smooth and expected. Enough information is available in this analysis that winding and evaluation of the inductor, with a network analyzer can be accomplish should the network be duplicated.

28