Ashhar Farhan VU2ESE, the designer of the µBITx has been experimenting with solutions to the spurs on SSB mode above 18MHz. He has posted to the list that one spur fix that is now consistently working for him, and that is easy to apply, is to fix the distortion in the 45 MHz amp.
He inserts a 0.3µH inductor in series with a 10 pf capacitor across TP13 on the v4 board. This is on the output of the 45 MHz IF amp going to the front end of the mixer. The inductor comprises 8 turns on a T30-6 toroid (to give 0.3µH or 300nH).
UPDATE But wait … he proposes a better spurs fix a day or two later
1. Change the resistors R26 and R46 to 220 ohms (from 470 ohms). this increases the darlington pair’s standing current and decreases the Harmonic distortions.
2. The 90 MHz trap was difficult to tune unless you had a spectrum analyzer. Instead, we can use a Low Pass Filter instead. A ‘T’, rather than a Pi is used. This helps improve isolation on both the RX/TX path.
3. The LPF consists of an L-C-L of 300nH, 47pf, 300nH.
Removing the BFO leak into the local oscillator
Ashhar also found that there is cross talk inside the Si5351, whereby the 12 MHz output from the BFO leaks into the local oscillator.
Ashhar used a 5µH inductor in series with a 47 pF capacitor placed across C202 of the LPF suppresses this. The inductor comprises 35 turns on a T30-6 toroid (to give 5µH). However, another constructor found this blanked out the 30m band, and calculations reveal that there may be an error with the calculation of inductance or capacitance that is waiting to be confirmed. The filter should be tuned to 12MHz suggesting a slightly small capacitance value of around 35pF
uBITx.net will be looking for verification from constructors that these mods represent a definitive fix for the spurs and carrier leak problems.
Ashhar VU2ESE and Raj VU2ZAP have nailed the cause of the spurs. The 45 MHz IF amplifier is distorting and producing a second harmonic at 90 MHz, which mixes with the local oscillator to produce a spur at 90 MHz.
Here is an example of how it happens :
the spur moves down as you tune up.
when the radio is tuned to 28 MHz, the spur is at 17 MHz. When the uBITx is tuned to 28.150 MHz, the spur is on 16.850 MHz.
At 28 MHz, the local oscillator is at 73 MHz.
There is a signal X such that :
X – local oscillator = 17 mhz
Local oscillator = 73Mhz.
X – 73 = 17, so X must be 90 MHz.
After checking the spur on a number of frequencies between 15 MHz and 30 MHz, it was confirmed that the above formula works consistently to predict the spur.
Now to confirm a simple solution … if the problem is in the 45MHz IF, then it may also be possible to address the IMD at the same time.
Jim W0EB has finally received his >500 mW Axicom relays, after they got misrouted in the mail.
H plugged these into his V4 test bed built up on an aluminium open chassis. Sockets are mica filled 2mm thick with silver plated flat contacts. This leaves the relays 2mm above the PC board.
With the >500mW coils (i.e. with coils that have fewer turns on them) the hope was that this would lessen the extent of inductive coupling between the switched portion of the circuit and the relay coils. Unfortunately this does not appear to be the case. Jim got almost exactly the same results as with the 400 mW relays mounted the same way.
40 meters is still just barely in spec with the 3rd harmonic being -43.8 dBC. The others are also in spec but not much better with 3rd harmonic varying from -46 dBC on 10 meters, -44.5 dBC on 12 & 15 meters , -44.7 dBC on 20 meters and -57.3 dBC on 80 meters. 60 meters wasn’t tested.
Raj VU2ZAP notes that it makes quite a difference between socketed an unsocketed replacements. It is best not to use sockets for the relays, in order to get the relays as close as possible to the PCB.
Before and after shots of relay replacement as observed by Raj follow:
Raj and Jim both note that there is variation between boards in terms of improvement in performance. The reasons for this are not known.
The conclusions are:
relays help, and ubitx.net recommends them, but replacing relays with AXICOM relays may not always be sufficient to reduce “blow by” on some bands or modes
there is no point in spending more on Axicom relays with a higher coil current
the best option would be to replace the filter section altogether, and there are a number of ways of doing this.
One of the last challenges to be conquered with the µBITx is the unacceptably high intermodulation distortion produced in the IF stages on transmit.
The IMD seems to be sourced in the Termination Insensitive Amplifiers used as bi-directional IF amps at both 45 and 12 MHz.
Warren WA8TOD is experimenting with a prototype MMIC amplifier utilising the ERA-3SM+ (80 cents each on eBay) mounted on an RF prototyping board ($2.50 each from SV1AFN) that is the proper size and form factor to replace both transmit TIAs. The MMIC/board measure flat response from 300 Hz to well over 200 MHz, and will provide around 22dB of gain.
This test quantifies the IMD performance of the combo. The reference point on the SA is shifted +6 dB so that the readings reflect the power level that would be achieved with a single tone for convenience of interpretation.
Measurements were taken at Vcc of 3.3 volts and current draw was the recommended 35 ma.
The blue trace shows performance at 0 dBm out and is an acceptable -42 dBc (-36 dB minus 6 dB for a single tone carrier). Performance at – 10 dBm out, the level needed to drive the uBitx driver/PA chain, is an excellent -51 dBc (-45 dB minus 6 dB for a single tone carrier).
Yet to be determined is how to add the LM1117-3.3 regulator to hold the Vcc at the required level.
The graph above provided by Gary AG5TX shows the resultant effects of using Axicom relays (using Jim’s data points).
“Your shared data maybe useful as you have shown the same board and relay with different measured results. For that particular board and the measurement data you gathered, I would say the results ARE encouraging from this viewpoint:
You have 4.5 dBc more margin to FCC spec on your worst case data point (40m 3rd harmonic) with the supplied data set.”
These relays can be obtained from a number of sources. If you live in the US they can be found at Arrow, Digikey and Mouser for less than US$3 each. They are also available from Aliexpress for under US$2 each in a set of 10 and from Component HK in Hong Kong for under US$1 each.
More details on using Axicom relays for RF switching
For those interested in the benefits of Axicom relays see this webpage for a switched Bandpass filter bank:
There are wealth of tips on this page for those building switched filter banks.
It appears that the Axicom relays are the answer, and Ashhar Farhan has already committed to replacing these relays in production µBITx, presumably when the current µBITx stocks have been depleted. Hopefully Raj’s mod for removing spurs will also feature in a v4 board upgrade shortly.
Raj VU2ZAP has found good reduction of spurs with the following mods:
1. Replace L5 with an SMD inductor 680nH or 681nH. Mount at right angles to old toroid
2. Replace L7 as above. Note that C211 is bypassed with this mod.
Don’t be distracted by the missing C216 and C210. These parts were never populated.
An additional 45MHz filter is not required and, therefore, you should get almost the same power out as stock boards. If you ALSO use the second 45MHz filter the spurs get further reduced, but with some loss of output power. Raj used Murata inductors 1210 size. 1206 size will do.
With Mikes W0MNE relay fix and this mod, Raj has the uBitxv3 and v4 boards now operating within normal emission standards.
Raj thinks that a combination of high Q, small size and possibly internal shielding contribute to the reason why. Another hypothesis is that the original inductors caused ringing. Raj has also replaced L1-4 with 330nH SMD and that helped a little on one test board. His interest was to fix the problem of spurs with the simplest solution and without major surgery.
Mike W0MNE has come up with a fix for the harmonics problem with the µBITX that is simpler than other fixes …
He has two Ver 3 and one Ver 4 uBiTX rigs. He simply replaced K3, KT1, KT2, and KT3 relays with Axicom relays.
This has knocked down the 3rd, 5th, and 7th harmonic levels from -35dB to below -50dB on all 3 rigs. No board modifications are required, and no changes to the LPF circuitry is required. This may not be reproducible on all rigs, but with 3 rigs showing the same results it may be something constructors should consider.
Mike purchased the relays from Mouser – part # 655-V23105A5403A201. Cost was US$2.89 per relay plus shipping.
Raj VU2ZAP found a picture of the “innards” of the Axicom relay, and suspects the improvement in isolation is the result of the contacts being closer to the ground plane (as the relay coil is on top in this relay, whereas the standard relays supplied with the main board have the relay coil underneath the contacts).
This is a replacement for the 45 MHz transmit amp comprised of Q20 – Q22 on the µBITx main board. Warren removed C20 and C22 and used two short lengths of miniature coax to take the signal off-board to a 7 dB attenuator and then on to the MMIC amplifier, and then back onto the board.
The resulting MMIC amplifier gain was +16 dB to match that of the BI-DI amp on the stock µBITx.
The yellow trace is before the amplifier board was inserted and the purple trace after. 3rd order IMD was reduced by nearly 10 dB over the stock µBITx by using the MMIC amp.
Warren restored the Q20 – Q22 amplifier and moved the MMIC amp to the Q40 – 42 amp with similar results. Here he found that he required the full 23 dB of gain provided by the MMIC to achieve the same level of main signal.
Yellow and purple traces are as before with the new measurement indicated on the blue trace. Results are almost identical indicating replacing these two amps together would provide 19 – 20 dB of IMD improvement which would make the transmitter completely viable and, in fact, better than some commercial radios in terms of IMD.
In both cases the indicated power out from the two tone test was a little over 4 watts.
Warren also tested this MMIC board for its susceptibility to oscillation. He connected a 60 dB attenuator between the input and output while feeding a signal into the input. He then gradually reduced the attenuation one dB at a time until oscillation was visible on a 1.5 GHz spectrum.
The amplifier broke into oscillation very reliably when the attenuation was stepped below -14 dB. Higher than that and it was completely stable.
In the course of these two tests the board was hanging unshielded about 3 inches from the PA heat sinks and the output was a little over 4 watts. In both cases he saw no indication of oscillation.