Update on fixes for uBITx spurs and harmonics

The number of posts on the BITX20 IO Groups list hasn’t gone down very much over the last month,  but the ideas that swirl are all about potential fixes to the µBITx spurs and harmonics.  And a new problem has been identified … very poor intermodulation results.

The µBITx is simply not putting out the nice pure signals we would like.   We are now trapped in a pattern of waiting for the  “magic bullet” solution to our woes, but sadly this seems unlikely to result.   There are just too many issues that need to be resolved.

Somewhat understandably after months of exciting developments, new contributions from constructors with substantive mods and hacks making the whole package sing are now no longer emerging.  We had become somewhat used to regular exciting new updates on processor or screen additions, or simple fixes to niggly issues (such as low audio, or distorted audio, or the annoying pop on toggling between RX and TX), and now we are becoming a bit more despondent about those more serious problems of compromised output.

Some progress is being made in solving spurs and harmonics.  We know that the built-in LPF filters and relays in the output stage can simple be disconnected and bypassed by adding a new external filter set (e.g. using Hans G0UPL’s filter board and filter kits for LF through VHF) on the antenna output.  After doing this the rig should work just fine on CW operations on all bands.   Of course, it may need either a hack to the firmware to get automatic band switching to work, or we can just use a rotary switch.

Alternatively, there has been work on modifying the filters in situ.  By changing the relay switching logic (e.g. mounting the relays under the main board and rewiring the switch points) and at the same time adding a second set of relays on an add-in daughter board we can restore effective functioning of the existing LPFs.   They work fine, but the switching system creates unwanted blow-by.  However, this involves quite a bit of hacking on the main board that many constructors may baulk at.

And none of this solves the issues with the spurs on SSB on 15m and above, that requires yet another fix of a second filter board at the output of the 45MHz mixer, and then there is the probably unrepairable issue with intermodulation products are probably way above what would normally be expected.

So where does this all leave us?   We have had a lot of fun on the journey to date.  We all acknowledge that most of the fun has been in the journey, but what’s the point in a journey that has no end?

Should we wait a bit longer hoping for a fairy godmother and a magic bullet solution to appear?   We  could spent more money to acquire the new QRPLabs transceiver  kit when it comes out (hey, all bands all modes and around US$150!)?    However, we already know that Hans G0UPL won’t release the firmware source code – therefore, there will be no add-ons and feature enhancements unless they come from Hans.

We could now just bite the bullet and hack our µBITx to turn it into a CW rig?  But it is supposed to be a phone rig as well.

Hold in there a bit longer, the “magic bullet” may eventuate.  Who knows? And I am sure there are plenty of other projects under the workbench for you to get on with in the meantime.   At this QTH there’s a couple of mag loops under construction (Teensy controlled) and a bench power supply project …

Problems exposed on uBITx sans filters

Warren WA8TOD, who has now relocated his µBITx to the top shelf (of abandoned projects), but he has provided us with a series of shots of the output of the µBITx on different bands, with no LPF filters in place.  These show the raw product coming out of the µBITx and put in full view the problems with the 45MHz filter and the harmonic generation in the power amplifier.

80,40,20 meter CW
– Power level set on each band individually to 5 watts
– Green display line represents the -43 dBc threshold
– Second harmonic on both 80 and 40 is somewhat problematic as are odd harmonics out to 13th

80,40,20 meter SSB
– Power level set on each band individually to 5 watts CW then 1 kHz tone input level adjusted to produce the same 5 watts
– Green display line represents the -43 dBc threshold
– Second harmonic on both 80 and 40 is somewhat problematic as are odd harmonics out to 9th

17, 15, 10 meter CW
– Power level set on each band individually to 2 watts
– Green display line represents the -43 dBc threshold
– Harmonics can all be removed with a simple 34 MHz LPF. The filter included in the uBitx is more than adequate

17, 15, 10 meter SSB
– Power level set on each band individually to 2 watts CW then 1 kHz tone input level adjusted to produce the same 2 watts
– Green display line represents the -43 dBc threshold
– 17 meter spurs represented by markers 1 and 2 are problematic because only a close spaced, single band BPF can be used to remove them.
– 15 meter spurs at the low end need to be addressed.
– A single filter cannot cover these bands because of the close spaced 17 meter spurs.

Reference