Homebrew Peppermint BITX: KC1FSZ Goes Scratch-built and Sweeps a Filter

Hi Bill/Pete:

Congratulations to Bill on the award.  Very well deserved!  

I just watched the video that Bill made with Farhan and I really enjoyed it.  As usual, lots of useful information there.

Farhan’s demo of the Sweeperino was particularly timely because this past weekend I was doing something very similar.  

As I’ve mentioned to Bill before, I am working on a scratch-built BITX to accompany the Peppermint BITX-40 module rig that I’m running right now.  I took Pete’s advice and started from the audio end.  After building the audio amplifier, product detector, and first/last pair of IF amplifiers, I was able to connect the VFO where the BFO is supposed to go and enjoy a pretty reasonable direct conversion receiver.  It works great!!  My first working receiver.

So my next step was the crystal filter.  I puzzled over all of the different articles, tried to follow the G3UUR method, sorted crystals, etc. and finally put the whole thing together.  But all along I’ve had this doubt that I could have possibly followed all of the steps closely enough to achieve the precision needed for a 3.5 KC SSB filter - it just didn’t seem possible.

Last week I decided to try to “sweep” the filter.  This was a multi-day undertaking in itself.  I pieced together the microprocessor, an AD9850, a TIA amplifier, the RF power detector from the SSDRA book, and some software that generates the necessary frequency sweep and collects power readings.  The picture of the test rig is included below.  It’s not much to look at.

With low expectations (and wires that are too long), I ran a few seeps and then downloaded the data onto my computer for analysis.  No fancy software here, I’m just using Excel.  When I put the data into a line chart in Excel, what emerged from the jumble of seemingly random numbers was a thing of beauty: a 3.5 KC passband right below 12 MHz. It’s the next stage beyond the Joy of Oscillation:  The Joy Of Selectivity!

It’s even showing about the right insertion loss.  Note the setup is very primitive and noisy so the fact that I can’t measure anything below -8dB isn’t surprising, you can ignore that part.

But I’m starting to get the feeling that this might actually work.  Onto the diode mixer …

73s,

Bruce KC1FSZ https://www.qrz.com/db/kc1fsz

VU3XVR's BEAUTIFUL Morse Keys

Truly a thing of beauty.  Farhan gave me one of these.  They are crafted by VU3XVR.  It is enough to turn a phone-guy like me back to CW.  Thanks Farhan!  

Here is VU3XVR's site:

http://vu3xvr.blogspot.com/p/morse-key.html

FDIM: Listen to Bob Crane Interview Hans Summers, G0UPL

Bob Crane, our intrepid correspondent at the 2017 Four Day in May event, caught up with Homebrew Hero Hans Summers G0UPL.  Hans's amazing web sites have been the inspiration for many projects in my shacks.   My favorite so far was his QRSS transmitter using an LED based multivibrator circuit to generate a shark fin pattern on grabber screens.  That was fun.  I am now really tempted to send one of Hans's WSPR transmitters into the stratosphere using party balloons. 

Listen to the interview here:

http://soldersmoke.com/FDIM17G0UPL.mp3

Visit Hans's QRP-Labs here:

https://www.qrp-labs.com/

Thanks Bob!  Thanks Hans!

Video: Farhan in the SolderSmoke Shack! BITX, JBOTS, McDonald Straw Sig Gen, uBITX, Sweperino and more!

Thanks again to Farhan for visiting us.  It was great to see his reaction to my humble implementations of his great designs.  I got him to sign my BITX17.  This was really a fantastic day for me and for my family.

Farhan Visits SolderSmoke Headquarters

Wow, today was an important day in the history of SolderSmoke.  Our friend -- and newly minted member of the QRP Hall of Fame -- Ashhar Farhan VU2ESE came to see us.  We had a great time.  I showed him all the many devices that I built based on his designs. 

We shot a video of the shack visit. I will post it to YouTube in a few days.

Thanks for coming to see us Farhan.  And congratulations on the richly deserved Hall of Fame Induction.

Homebrew History is Made: Farhan @ W7ZOI

Thanks to Wes W7ZOI for sending me the link to his page describing the visit of our friend Farhan.  I think this visit was a historic gathering of homebrew heroes and their groundbreaking rigs. Read all about it here:  http://w7zoi.net/Farhan-visit.html

Please help me convince Farhan to visit SolderSmoke HQ before returning to India!   Send him (or me) e-mails, texts, tweets, or just post messages of support below this post.   

WA8WDQ's Automated Filter Scanner (and a tip from VK1VXG)

Bill, Pete,

Thought I'd share today's activities.  Using Bill's excellent wire stripping technique, I was able to finish assembling my 7 MHz bandpass filter.  At this point, I'd normally terminate the filter then hook up the signal generator and oscilloscope to manually scan/observe the filter's response.  But today, I took a different tack.  Using the AD9851/Arduino signal generator I built last year coupled with the W7ZOI AD8307 Power Meter I built this year, I wrote some additional Arduino code to scan a block of frequencies while using the Arduino's A/D to record a level at each scanned frequency, thus characterizing the filter's response.  I also used a W7ZOI unidirectional TIA to boost the generator's output to within the full scale range of the A/D.  This setup is similar to the popular PHSNA project that has gained attention over the last couple of years.  The Arduino formats the data as CSV then feeds it into Numbers (the Mac Excel equivalent.) for plotting (see attached but please forgive the labels/legend on the plots as I didn't take the time to format them).  Unfortunately, my Arduino UNO is low on RAM so I'm only able to save a max of 100 scanned frequencies.  Even with the lack of resolution, as you can see, the output is definitely useable to get a sense of the filter shape and roll-off.  I'm now motivated to write some plotting code so I can get a more real-time output in lieu of the spreadsheet.  The output of the AD9851 seems pretty flat within several MHz as is the TIA, so the accuracy should be pretty good within moderately sized scan blocks.  Adding additional external FLASH (or moving to a bigger Arduino like the Zero) will fix the low resolution.  FYI - Arduino behavior when RAM is low can be very erratic!  It started crashing when I got close to the max available :(.  

If you haven't as yet, definitely check out the PHSNA project.  They've added a boat load of features beyond just scanning filter response . . . and at a fraction of the cost of a spectrum analyzer with a tracking generator.

Brad  WA8WDQ

Hi Bill,

I'm having trouble posting a comment in response to Brad, WA8WDQ's

automated Filter scanner project, and I'm hoping you wouldn't mind

passing this tip along.  It might be of use to other readers too.

Recent versions of the Arduino software include a Serial Plotter feature

that lets you graph directly from your arduino.  By making use of this

feature Brad wouldn't need to save data on the uno - just send it back

to the computer and graph in realtime.  It is a little primitive, but

still usable. I believe the plot window is 500 datapoints wide.

There is a nice tutorial on how to use it here:

http://www.instructables.com/id/Ultimate-Guide-to-Adruino-Serial-Plotter/

If you're in a hurry  and can survive with a bit of inaccuracy in your

sweep measurement you can also get away with using a simple diode

detector RF probe.  This is good enough for a "yep, that filter does

what I want" measurement, but not for measuring exact loss or 3db

bandwidth etc.

Hope that helps,

73s

Greg, VK1VXG