EMRFD Message Archive 4455
Message Date From Subject 4455 2010-03-14 17:45:14 Phil Sittner EMRFD Fig 6.79 I'm in the process of building the 14Mhz receiver as detailed in chapter 6. The 4mhz filter shown in the above figure was built and the insertion loss was measured at 7.5dB yet the figure implies of loss of 2dB or so. I then used the AADE filter design program, entered the filter as shown and the results predicted about 5dB loss. I then used Ladpac 2008 and the DTC mixed form option and arrived at an insertion loss of approx 2.35 dB but the circuit configuration differs from fig 6.79. Does anyone have an idea of what I've done wrong here? I'm pretty confident that Wes had a specific reason for the design but don't understand why.
[Non-text portions of this message have been removed]
4465 2010-03-16 20:13:27 w7zoi Re: EMRFD Fig 6.79 Hi Phil,
I took a look at Fig 6.79 and am confused with your results. That figure is a simulation done in Eagleware's Superstar Professional, which was a pretty good linear simulator in its day. It's still good, but outdated. The IL predicted with it is 1.35 dB.
The same result, exactly, comes up with GPLA.
LT SPICE predicted 1.34 dB. All three of these simulators differ in the methods used for analysis. In all cases, I assumed the capacitors were perfect and the inductor Q was 200. This is a reasonable number for a T68-2 core at 3.95 MHz.
I played with some of the other components in the circuit and nothing was very sensitive. Dropping the Q of all three shunt capacitors to a mere 100, and that is a profoundly bad capacitor, still didn't push the IL over 2 dB.
So what Qu did you assume for L in your simulations? Did you happen to read some of the values wrong for those shunt capacitors? While I used Silver Mica shunt capacitors for the versi
4467 2010-03-17 19:48:49 Phil Sittner Re: EMRFD Fig 6.79 Wes-
It took some doing but I finally fixed my problem. The initial thought was I had a capacitor or inductor problem and that was partially correct. The large trimmers you had shown weren't available from my junk box so I padded some smaller units and then tuned the circuit for max response using an 8640B signal generator and a homemade AD8307-based power meter; that's where I got the 7.5dB insertion loss. Well after spending some time at it the realization was I still wasn't within tuning range of the padded trimmer and made the appropriate correction and you can see the results. Insertion loss is down to 2.5 dB but I'm using a T50-2 core with 40 turns of #24--too much for the core. Bandwidth is good and the match is also. I'm truly sorry for the false alarm but I guess ignorance gives rise to such things. Now I'll spend a little time with GPLA and figure out how to duplicate the circuit and evaluate it. Thanks again for taking the time to lead me down the proper path. The first attachment shows just how far off I was with the second attempt at padding the trimmers.
----- Original Message -----