Monday, 5 May 2014

Sweeping the BITX IF

Having brewed up the little scheme for visualising the magnitude response of filters using RF sweeps generated by the Arduino, I decided to take a look at the crystal filter I made for my SSB rig. That rig started from Farhan's famous "BITX" design and, whilst I have added a number of elaborations, it is still very much - and very proudly - a BITX.

The IF filter had been made from Farhan's schematic - but I paid no heed to the accompanying words and music, specifically making no attempt to match or select crystals. Instead, I just took the first four crystals out of the bag and worked with them...


The results were less than perfect - so I made some adjustments "by ear" - you can see evidence of some of these modifications in the capacitors still hanging off the filter in the photo above. It had been sounding and working FB for the past few months - but now I was in a position to measure it...

I started by disconnecting the filter from the surrounding rig and measuring the IF response according to the scheme described in the previous post. This yielded the response shown below...


OK - it might not win any awards - but it is consistent with the acceptable performance my rig has been turning in over the past few months - and consistent with the BFO frequencies I've been using in my Double DDS Code (as described in the current edition of SPRAT)...


[My rig can be used in either LSB or USB mode, according to the band it is working - the two numbers above are the BFO frequencies for LSB and USB operation, respectively.]

However, the filter isn't intended to be used in some artificial test environment - it is intended to be operated in the very specific loading conditions in the middle of my rig - so I really ought to be measuring its performance there. Before I could do that, I decided some housekeeping was in order...

The IF section of the rig was squeezed into the corner of the PCB (as shown by the dashed red box)...


I took a few moments of redistribute things and give myself a little more elbow room, giving the whole IF section a slightly bigger red box to work in (and opening up the whole dashed blue box pregnant with possibilities for future developments)...


Now, it was easy to measure the response of the entire IF section - not only of the filter, but of the amplifiers on either side of it.

The signal from the DDS module was attenuated through a PI pad (actually a cascade of such sections, embodied in my switched attenuators), to both attenuate the signal and ensure a 50 Ohms source impedance and the amplifier after the filter was loaded by the 50 Ohm termination at the 'scope input, as shown by the overall scheme in this graphic...


Readers familiar with the BITX will recognize the segment of the original system under test.

The response is actually similar to that measured with the filter out-of-circuit (albeit with some gain - associated with the rather random combination of the loss from my attenuator and the gain from the two amplifiers)...


This little sweeper actually seems capable of useful work! Perhaps I should use it to try to refine the passband response of the SSB rig's IF filter - then again, life is quite short and what's 2.5 dB between friends?

...-.- de m0xpd

No comments:

Post a comment