Test drive; pi network sim + measurement

What a fun night!

I decided to hook up the existing NE602/LM386 circuitry as a “test drive” of my VFO (the circuitry that was in the old version of the MRad-40, on 2 pieces of perfboard).  It worked!  Although with the extraneous wires there’s a lot of AC hum being picked up.  I’m not too worried about that.  I was surprised it even worked given how ad-hoc it was tossed together.  Signals were a little weak even with the amplified speaker but I can worry about that battle another time (and when things are properly built inside the box with short wire spans).  Still it was pretty stable and I tuned the whole 40m band, even listened to a net for a bit (the “brothers net”).

After hearing how loud the harmonics were on 20m, 15m, 10m and even above, I decided I might go ahead and build the suggested pi filter to low-pass filter the VFO output.  I later realized that I think the pi was meant to transform the high output impedance of the JFET down to 50 ohms. However, in my application, the NE602, from the datasheets I read, also has a fairly high input impedance, maybe 1,000 ohms or so.  So I don’t really want to transform to 50 ohms.  But in the design process I just assumed 50 ohms on either end, which is also not right (probably more like a few kohm on each side).  I may need to tweak this further.

Anyhow, I went to my trusty pal Puff, the old DOS-based software package.  In about 5 minutes I had a filter design that looked pretty nice (for 50 ohms at least):

IMG_0539Unfortunately I forgot to put the marker on 7 MHz before snapping the shot, but you can see from the scale that at 7 the S21 is at essentially 0 dB.  At the first harmonic, I’m down 12 dB.  The text suggested lowering the Q by placing a resistor across L (broadening the response).  For my purpose on 40m the response is already more than broad enough so I left well enough alone.

I decided to build the series coil using my new T50-7 toroid core.  For the prescribed 1 uH it took about 18 turns or so.  I didn’t have 550 pF in the junkbox so I made do with 400-something pF.  I put it on my little TAPR VNA and got a measured result which wasn’t as good as on Puff but still quite nice:

mr40 pi filter for oscillator

Again apologies for not the greatest clarity here, but the marker #2 is at 14 MHz and shows S21 at 6 dB down!  So, not the 12 from Puff, but 6 dB is a pretty decent knock-down.  (maybe with proper impedance matching I can get a better roll-off — but does it even really matter for this?)

Listening to my oscillator with my Kenwood on the various harmonics showed a marked change in S-meter reading after filter insertion.  Before, all the harmonics sounded just as loud as the fundamental.  Now, I could barely hear them.

If I wanted to knock the level down more I could add additional stages.  But I think that’s overkill.  We’ll see.

The other thing I really need to do is install a bandpass filter coming in off the antenna.  I’m getting bad broadcast breakthrough.  I think it’s a local AM station but it might also be a strong out of band shortwave station (I couldn’t get an ID).  They were far louder than the 40m ham signals!  Not acceptable!  My old version of this radio had a bandpass filter on it and so I’m going to rebuild that but do it in a neater/tighter package and not with wires strewn around (I hope).  I don’t expect amazing performance from this simple rig but I can certainly work to exclude as much noise as possible.

I also replaced a somewhat questionable JFET with one that was still in it’s original packaging (and in a piece of anti-static foam).  And my gain did creep up a bit.  The buffer, from what I can tell, is unity gain.  The second stage has some gain but not a lot, maybe 3 dB.  I didn’t bother trying to boost the buffer gain because I wasn’t sure it was necessary.  The output signal from the VFO is about 3 mW.  That doesn’t seem like much but the 602 chip seemed ok with it.  If I find out I should push for more power I can try to increase gain.

So, bandpass filter, then the (for me) painstaking process of trying to neatly Manhattanize the 602/386 stuff into the box.  I’ve never done this style before but I’ve got these nice pre-made pads to try out.  Hopefully I don’t end up with a gluey mess!  I hope to have the rig done this weekend.

This entry was posted in Mrad-40, pi network, Puff, TAPR VNA. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s