low output only on 40m


Fortunately its only a small fraction of the v4 population that have this issue.  I carefully counted turns in the LPF - no issue there.  Also I bypassed the LPF with a wire - still stay issue that 40m power output is a bit low.  I remember my ubitx may have been 4-5 watts on 40m as delivered.  Any attempts to insert added 40m filtering into the 45 MHz transmit path further reduces power.  When 40m CW is used the transmit path seems to experience whatever is attached in the (unused) 45 MHz transmit signal path - I verified this with too much experimenting!  Could it be a faulty core material in those mixer transformers, I don't know I can't think of why it may vary otherwise between rigs. 

Since then I have decided on an added xtal filter of this configuration - retaining the 47 ohm resistor onto the mixer side, then a 0.1 uF blocking cap, a binocular transformer that feeds one side of the xtal filter.  the other side is going straight to the transistor emitter, which should be adequate given its higher impedance there.  Now my 40m CW signal is 'only' dipping to about 3 watts.  I don't possess a spectrum analyzer, but I may get to evaluate later this month.  In summary I am using Gordon's relay board for addressing harmonics (it is fully compliant now here) along with this added filtering - which I hope will improve my 17-10m band mixer spurious (I tried the v5 style LC filter it did not help much).  I do have shielded SMT inductors installed at L5 and L7 - they weren't enough by themselves to remedy this 40m challenge. 

In other news - I have the NESCAF variable bandwidth audio filter integrated inside now - wow nice to have a bandwidth control - it works nicely tuning in signals from the massive CW pileups - as well as providing appropriate selectivity for SSB. I just need to disassemble  front panel and provide the hole for the bandwidth pot. 

73 Curt


Checked the filter out with the VNA (by soldering coax leads directly to the input and output), and swept it from 6 to 14 MHz.  It has -0.3dB insertion loss at 6MHz, -0.4dB insertion loss across 7-9 MHz, and the -3dB point is at 11.1 MHz.  It looks just like a low pass filter designed to roll off around 11 MHz.  I don't think that this is the culprit.



so how low are you outputting on 40m CW?  do you have any audio oscillator to check out SSB (I am forgetting what I obtained but it may have been a bit low also)?

I am thinking that 3w CW isn't bad, and I could work this for a long time to get maybe a couple more watts.  yes a single band rig it would be easier to keep working on. 

yes I have a nice signal on 30m so it is strange, and even stranger to make any changes near the 45 MHz mixer port. 

so perhaps you have a nano VNA there or something fancier? either way you seem to have good data. 


MVS Sarma

in the code of si5351 under ubitx v5.
line 45 reads
uint8_t  si5351bx_drive[3] = {3, 3, 3}; // 0=2ma 1=4ma 2=6ma 3=8ma for CLK 0,1,2
They are set for higher output level.

while in V3 the same I remember is set to 1,1,1

In one case where similar complaint is received , we managed to change to higher current 2 or 3 depending, and the cw Tx power rose from 2 watts to 4 watts, followed by similar increase obviously on ssb.
the use is happy after that.

MVS Sarma

I meant, the user is happy athat.fter


I have been following this thread with a little confusion, and I am not sure what the resolution is or if there is one. I have a V4 uBitX and the 40m CW output has always been low. Where I see 3.5w out on 40m CW into a dummy load, I can see 12w when on SSB and "Ahhh"ing into the microphone. I see a similar disparity on 17m. Since the SSB output is reasonable, I haven't worried about it too much and assumed something about the way CW is generated isn't unbalancing the mixer enough or something. Is there a fix?

FWIW, i am using KD8CEC v 1.1 software.



Reporting Success!

First of all, my dip was down to 2-2.5 watts on 40m.  I myself would not worry much about having a less extreme dip.

Asshar gave us the valuable clue that capacitor C81, 470 pF is selected for frequency compensation of the transmit chain. 

I added 330 pF in parallel, and it solving my issue. 

approx CW power before and after this change:

1.8 MHz   8w    7.5w
3.5 MHz   6w    5w
7 MHz      2.5w   5w
10.1     5w     4.5w
14     7w       7w
21      4.8w     4.5w
28      3w        3w

i did not use my best wattmeter, but I trust this TT1202 enough that I am very happy with this change. 

caveat - I do not know how critical this adjustment is for SSB - I have another mod there to be tested. 

Thanks to Jim who brought back this thread, I hope a few hundred pF added to C81 might cure yours also

73 may your BITx efforts rock

Curt wb8yyy