EMRFD Message Archive 2060
Message Date From Subject 2060 2008-09-14 08:50:46 timshoppa BFO Feedthrough into IF I know that others have had this problem and conquered it, but I'm
wondering what steps were taken.
I've got my homebrew crystal filter in a can. No feedthrough problems
But I built my hybrid cascode (dead bug style) and on the other end of
the same board (which is only 5 inches away) I've got my BFO. I put the
mixer (a TUF-1) in between them because, well, the mixer goes in
between them :-).
No matter what attempts at shielding I make (see below), this
arrangement allows enough BFO leakage into the IF input, enough that it
is triggering some minimal AGC interaction.
Shielding I've tried so far: metal partition around BFO didn't help at
all. Feedthroughs and chokes didn't help at all. Putting my hand near
the board even with a metal partition seems to have some effect on
feedthrough, usually making it worse.
The leakage is the same whether or not the mixer is in place.
Should the mixer go behind the shielding? There's a short distance
(about half inch) of open wire from the BFO pad to the mixer, and I
don't think this is the problem because no matter where I bend it or
even if I remove it the problem seems to be unchanged.
Does the BFO need to go into it's own 100% shielded can? Does the mixer
need to go into it's own 100% shielded can? I was hoping to put the IF
amp and BFO and mixer and audio amp all on the same board.
I've built NE-602 product detector rigs with everything
deadbugged onto the same little board and never noticed any
feedthrough from the BFO there, but of course there the injection into
the NE-602 is way way less than what a TUF-1 takes. Is BFO shielding
going to be an issue on every single DBM product detector receiver
because the BFO has to put out at a higher level?
2061 2008-09-14 11:39:27 Robert Cerreto Re: BFO Feedthrough into IF Tim,
Some time back I had similiar problems with a similiar layout. I had to break out the BFO into a separate can. That turned out to be a blessing in disguise. I too wanted everything on one board. But, breaking it out gave me further physical layout possibilities that worked out better. If you put the BFO in its own home, use feedthroughs for power and either a good RF connector for the goes intos and goes outtas or use a coax whose shield can be soldered to the enclosure at the entry point. I did not have to shield the mixer or IF amp.....only the BFO. You are correct about the high level BFO making a good transmitter!
73/72, Bob WA!FXT
2062 2008-09-14 12:31:05 Rick Re: BFO Feedthrough into IF This is a classic problem, and shielding, bypassing, and decoupling are seldom well
understood even by professional engineers. Your BFO and other oscillators in the system
will have signals on the order of 1 volt, and the input to your IF amplifier will be sensitive
to signal levels well under 1 microvolt. That's over 120 dB difference, and it is very
difficult to achieve that much isolation even in the same room, let alone on the same
circuit board. Electromagnetic shielding is also difficult to simulate, so engineers who
depend on a computer to tell them what to do are at a significant disadvantage.
The most straightforward way to tackle the problem is to put each oscillator and high-gain
amplifier stage in it's own 100% shielded can. If you look inside high-performance
military receivers, NASA deep space receivers, and spectrum analyzers, that's how it is
usually done. There are a number of other tricks, like using multiple IFs and distributing
the gain, but when high performance is the goal, you will often find the receiver functions
divided up into separate, well-shielded modules.
Simply putting the BFO on its own scrap of PC board with separate power supply and
ground leads can be a big help. That's the way we often do it with prototype receivers
spread out over the bench. Then if more shielding, decoupling or physical separation is
needed, it's easy to add.
The reason I sometimes design and build radio circuits with multiple functions on a single
board is to make them easy to duplicate. Those are often the ones that are published. But
for a single experimental radio, separating the functions onto separate boards has many
advantages, and almost guarantees better performance than the same circuitry built on a
single pc board. All of my experimental prototypes and serious instrumentation receivers
are built that way. For interconnections between modules, I use either small coax or
twisted pair, with a ferrite bead slipped over the pair near each end. Wes has a number of
example radios built ugly style on separate scraps of PC board scattered through EMRFD.
NE-602 product detector radios need less shielding for a number of reasons. A well-
designed NE-602 radio will have much less IF gain than one with a stronger product
detector. If you must put all the circuitry on one board, or severely restrict the receiver
current drain, the performance limitations of the NE-602 may be an acceptable
Have fun with the design. The hybrid-cascode IF amplifier is wonderful, and deserves a
strong, high performance product detector. Separating each receiver function onto its own
ugly constructed board is the first step I'd suggest.
> Does the BFO need to go into it's own 100% shielded can? Does the mixer
> need to go into it's own 100% shielded can? I was hoping to put the IF
> amp and BFO and mixer and audio amp all on the same board.
> I've built NE-602 product detector rigs with everything
> deadbugged onto the same little board and never noticed any
> feedthrough from the BFO there, but of course there the injection into
> the NE-602 is way way less than what a TUF-1 takes. Is BFO shielding
> going to be an issue on every single DBM product detector receiver
> because the BFO has to put out at a higher level?
> Tim N3QE
2066 2008-09-14 18:44:41 Robert Cerreto Re: BFO Feedthrough into IF Hey Tim,
When you solve your troubles....be sure to share it with everyone else.
I am sure there are others out there who wanna know!!!!! <
>>> It is a common trouble.
73/72, Bob WA!FXT
2067 2008-09-15 04:55:40 Allison Parent Re: BFO Feedthrough into IF 2069 2008-09-15 07:44:54 ha5rxz Re: BFO Feedthrough into IF This is an interesting challenge and one which I encountered when I
tried to build my spectrum analyzer. I found that the only way I could
get reasonable performance was the following:
1) Each stage was enclosed in its own box made from double-sided PCB
material. All seams on the boxes were covered in copper foil.
2) Each PCB inside the box was mounted using multiple brass standoffs,
this ensures that the ground on the PCB really is ground.
3) Each RF connection inside the box was made using coax. The
connection to the socket(s) was made RF tight by wrapping each socket
in tape then wrapping in in copper foil. At the board end the
connection went right to the IC pin with the shortest pigtail on the
4) Each DC or control signal went through a feedthrough capacitor, a
1mH choke and then a second feedthrough capacitor. Inside the box two
ferrite beads were fitted.
5) All RF cables were 100% shielded. This was achieved by wrapping
each plug connection in tape then wrapping in in copper foil.
6) Lots of decoupling on the power supply in multiple places.
7) Inputs on one side of the chassis. Outputs, oscillators and high
current circuits on the other side of the chassis.
The result was an RF clean design, hope this informati
2071 2008-09-15 07:44:54 bobtbobbo Re: BFO Feedthrough into IF >Tim:
A methodology I have employed successfully on building any design is
to build separate modules defined by block diagram function, each
with its own 3 terminal voltage regulator. After testing/debugging, I
build up a PCB box around the board, with top and bottom lids held by
screws (if I ever need to get inside). Appropriate feedthrough
decoupling used as necessary. If the enclosed module retests OK, I
move on to the next one. One can build a fairly complex design by
taking "one bite at a time", testing each and then doing interconnect
Another advantage is, when a design evolves, only one particular
function module might need to be changed. The original one can be
saved and perhaps used in a future design.
2073 2008-09-15 07:44:55 cwfingertalker Re: BFO Feedthrough into IF Hello Tim
I had a similar problem while building the Progressive Receiver. I
boxed the BFO, Mixer and IF board in separate boxes. I also used
feedthru capacitors for DC lines and BNC connectors for the RF lines.
In the long run it cured the problem even though it is a little more
2075 2008-09-15 13:15:34 ve7ca2 Re: BFO Feedthrough into IF HI Tim:
Very interesting to hear of your BFO leakage problem. I experienced
the same issues when building my Homebrew Transceiver the HBR-2000. I
solved the problem by building my Product Detector, BFO and Audio
2078 2008-09-16 12:17:46 cwfingertalker Re: BFO Feedthrough into IF Hello Markus,
I am interested in your 6 meter stand alone rig. Can you give us some
details of this project? I am also penciling out a project for a 6
meter SSB single bander. I am planning to use the KK7B micro R2 and
T2 rigs. Maybe set them up on an HF band and build a transverter to 6
meters. I would be very interested in your plans. I looked at your
website. Very informative.
Sam's Valley Oregon.
2082 2008-09-17 16:48:38 Glen Leinweber Re: BFO Feedthrough into IF Engineers have always been told something like:
"if you can't get rid of it, feature it".
So was wondering if you could build something
like a VNA by featuring L.O. blow-back into
the RF port to act as a probe for impedance?
The DC output at the I.F. port should be affected
by the impedance at the RF port interacting with
the leaked L.O.
Perhaps a L.O. phase modulated with an audio-
frequency would give more useful information at
the I.F. port.
2085 2008-09-17 19:45:22 ve7ca2 Re: BFO Feedthrough into IF Hi Bill:
I have based my design on the "6.11 A Monoband SSB/CW Transceiver"
2088 2008-09-18 17:21:10 cwfingertalker Re: BFO Feedthrough into IF Hello Markus,
Thanks for the details. I did look over that design. My idea was to
recreate the "Benton Harbor Lunchbox". A small light simple box for
mountain topping. Thanks again.