Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
Re: Changing out L5 & L7 in UBitX for Spur Fix
We've got a bunch of different spurs.
toggle quoted message
Show quoted text
For example, the 12mhz spur due to clk0 crosstalk into clk2 is not going to be cured by swapping out L5,L7. Also, I would not expect that swapping out L5,L7 would clean up a 45mhz-DialFreq spur due to a second harmonic at Q20,21,22.? My best guess had been that L5,L7 prevent energy from the power amp from finding a path back into the 45mhz IF amp and taking another trip through the D1,D2 mixer creating all sorts of unexpected spurs. Perhaps Raj's L5,L7 mod reduces magnetic field coupling from toroids in the PA? Can be problematic to have a 10W amp operating unshielded in the same box with the low level signals going through the IF amps. In post 61645, Farhan states:? "With the relays that are now used for ubitx,? the harmonics at 3.5, 7 and 10 are in under -45 dbc. we don't need the axicom relays." In post 62307, he says: "The boards continued to ship with the same relays as some have also reported no difference with the new relay.? On the other hand, a new board layout with spur fix, allions suggested lpf layout and an lm386 af amp is in the mail. I will prototype it as soon as it arrives and share the results" My reading is that the v4 uBitx now ships the same as it always has.? ?There have been many many reports here of trouble with harmonics on v4 boards, though perhaps not at the frequencies Farhan was testing at (not 14mhz and up?), and perhaps with toroids oriented a bit differently or something.? So I'd take that "don't need the axicom relays" with a very large grain of salt.? The good news is that post 62307 also suggests that v5 is being actively worked on to reduce spurs and harmonics. > I have read that Raj has agreed that either the LCL filter OR his replacement of L5 and L7 do reduce the spurs. I think Raj said his extra 45mhz crystal filter was as good as some of the other fixes, not so sure he has recently claimed the L5,L7 fix to be a cure-all. Would be good if Raj and/or Farhan could give us an update on the current thinking of all these various spur fixes. My take: The extra 45mhz crystal filter might well take care of most of these spurs, though hard to obtain those 45mhz filters, and hard to verify that the center freq of the new one matches the one at Y1.? If adding a 45mhz crystal filter,? I'd want to sweep the IF to verify all is well. Farhan's LPF filter for the 90mhz harmonic takes care of the 45mhz-DialFreq spur, but won't touch all the lower frequency stuff we get from coupling between PA and IF. The L5,L7 fix would take care of that coupling, but not the 45mhz-DialFreq spur. The 12mhz trap fixes a spur that none of the above address. I don't have a way to easily measure harmonics and spurs? But I'd guess the best path for most is: ? replace kt1, kt2, kt3 with the axicom relays (or add a set of outboard low pass filters) ? do Raj's mod of replacing L5,L7 with shielded 1210 surface mount inductors (other SM inductors did not work) ? do Farhan's LPF mod to knock out that 90mhz harmonic of 45mhz going into D1,D2 ? somehow knock down the 12mhz spur due to clk0-to-clk2 crosstalk, perhaps using Farhan's 12mhz trap This stuff is not trivial to get right. Would be good to arrive at and easy and simple recipe to clean up the v3 and v4 rigs. In many cases the spurs and harmonics may not meet regs, but if operating at 5 or 10 watts are not too big an issue.? Operating the uBitx with a 100W amp could indeed be a big issue, and may attract unwanted attention. Jerry On Mon, Dec 24, 2018 at 11:20 AM, Evan Hand wrote: I have been following all of the threads you have listed above.? Note that the low pass filter by Ashhar is the newest.? I have read that Raj has agreed that either the LCL filter OR his replacement of L5 and L7 do reduce the spurs.? The solution from Ashhar is more complex, in that there are bias resistors that need to be replaced as well. |
to navigate to use esc to dismiss