Re: Harmonic performance - SSB vs CW


I discovered this problem early on and replaced the entire filter network with a QRP outboard network from another QRP radio I got off eBay (picture attached, also has other features for power and SWR and separate receive antenna port).  I use 1-of-N binary decoding like you suggest with a simple mod of the code.  I did this, though, primarily because I wanted to investigate using the uBITx on 160 and 6 meters too… but it cured spikes seen at other parts of the spectrum too. I may have the spectrum pictures from my Spec An somewhere that I can share. Using a daughter card is the way to go in my opinion because it allows experimentation beyond the 80-10 meter existing platform.  It could be made much better/ less power hungry with latching relays.


From: [] On Behalf Of Ashhar Farhan
Sent: Saturday, August 4, 2018 11:22 PM
Subject: Re: [BITX20] Harmonic performance - SSB vs CW


I have been following this thread, I have to add three things here :


1. Unike commercial products where no criticism is encouraged, ubitx is an open source radio, So, Warren, we all do want to hear and correct whatever goes wrong with this design. We all hope it gets bette with time. As Linus Travolds reminds us "Given enough eyeballs, all bugs are shallow".


2. The reason for the convoluted relay system is that we ran out of pins on the Arduino to directly switch the LPFs. Instead, we should use a 1-of-N decoders to drive the relays, increase the number of LPFs to six (from the current count of 4). Bad economics. It will also lead to a differently sized board as it will be difficult to fit two more relays AND decoder onto the same sized board. This will mean putting the people like Sunil out of business until they retool their boxes. 


3. Another possibility is to split up the design into multiple boards. It makes sense for the solder happy hams, it is  nightmare for those who'd rather ragchew and chase dx than get under their race car's hood for months.


Given that there are non-technical considerations and a few thousands of ubitx already in use, we must be careful to not break backward compatibility of these radios. Any sugggestions?


- f


On Sun, Aug 5, 2018 at 8:45 AM, Gordon Gibby <ggibby@...> wrote:

Jerry wrote:

"Looks to me like a good part of the problem is board layout.
Each LPF is fed from a relay, then has a trace from the far side of the LPF
coming back right under the LPF to get back into the other set of relay.contacts."


The photo of a dpdt relay that I enclosed was just the only internal image I could find.   I haven't found one of the actual relays used.


Enclosed is a backlit photo to show theground plane --- which covers just about everything --- and the LONG return trace  from the other end of the filter.


I don't quite know how to "read" the rigol display that Warren put up, but since some of his bands hit the -43 dBm requirement of 97.307(d) (first part):  (d) For transmitters installed after January 1, 2003, the mean power of any spurious emission from a station transmitter or external RF power amplifier transmitting on a frequency below 30 MHz must be at least 43 dB below the mean power of the fundamental emission.​....


then I'm guessing that a few dB more isolation and more bands would "pass".    The return trace comes nicely close to lots of components!   The ground plane is available almost below everything.    drilling through and adding a tin metal shield between the return trace and the remainder of the circuit with a couple of solders to the ground plane might add some dB of isolationwithout having to tear into the relay.   It is probably advisable to run the "shield" tin metal right up to the relay in hopes of possibly even improving by a tiny bit the isolation there.   dunno.   brighter people than me, and those with the nice spectrum anallyzers (or Ashar) might have to figure this one out!





From: <> on behalf of Jerry Gaffke via Groups.Io <>
Sent: Saturday, August 4, 2018 10:07 PM
Subject: Re: [BITX20] Harmonic performance - SSB vs CW


Looks to me like a good part of the problem is board layout.
Each LPF is fed from a relay, then has a trace from the far side of the LPF
coming back right under the LPF to get back into the other set of relay.contacts.

One possible solution:
Rip out that return trace on each of the LPF's.
Add a set of straps at the far end of the LPF's to select one of them for connection to the antenna jack.

I haven't been following closely, has anyone reported any numbers on just how bad
these harmonics are during SSB ops?   


On Sat, Aug 4, 2018 at 06:28 PM, ajparent1/KB1GMX wrote:

Using two of the current ones would yield a very good result as on would be literally physically
on the other side of the filter for a good layout.




…_. _._

Join to automatically receive all group messages.