Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: [1]   Go Down
  Print  
Author Topic: BCB Trap questions  (Read 3190 times)
KB1SNJ
Member

Posts: 121




Ignore
« on: November 21, 2017, 11:20:27 PM »

hi all,

I got some answers in a separate antenna matcher topic I started but this Trap is really a separate topic. I need to stop nearby 1590 AM from getting to my gear, mainly a crystal set I am playing with.

So, I built a trap with a variable capacitor with fixed caps, coil etc, which taught me about LC circuits. Then improved it with better (higher Q) ferrite and added a mica capacitor.  It stops roughly half the strength (by ear) when testing into a boat anchor radio. I'd like to block it completely.  To make the test, I attached the Trap to the antenna input, then set the RF gain low enough to where the signal was not audible. Then I connected the antenna to the trap.

Meanwhile I had also built a coax current meter with a 3mA meter, 1n4007 diode, resistors and coil and capacitor per an online schematic. It worked, couple issues but it worked. So I figured the same principle could test my trap and sent a 1590 signal from an Eico 324 to the trap and the meter. Not much happened (not enough power). So I grabbed a 1mA meter and wondered what if I simply use a diode to rectify the signal to dc, no capacitor, or resistors, I grabbed a 1n4007 diode and sent the Eico signal through the 1n4007 diode to the meter (other meter post went to the Eico chassis to complete the circuit) and it barely moved the needle.

Then somehow I decided to switch the 1n4007 diode for a germanium diode and the needle moved a lot more!? To the 8th line at max trap, min trap was the 1 line. Without the trap, the needle goes to about the 2nd line. So the trap at its peak, brings the needle down to the 1 line. That's approx half which seems in line with the receiver ear test.

Also the trap variable cap can move about 1/4 rotation and the mA meter needle stays on the 1 line, that seems like a lot. The air variable cap is 140pf

Here's my questions:

Why does the germanium diode work so much more effectively than the 1n4007?
Should a trap be able to fully null the signal rather than half?
Why can the trap BOOST the current to the meter? (isnt that like a perpetual motion machine?)


IMG_20171122_015828 by Chris P, on Flickr










Logged
VK2TIL
Member

Posts: 545




Ignore
« Reply #1 on: November 22, 2017, 01:26:51 AM »

Why does the germanium diode work so much more effectively than the 1n4007?

The 1N4007 is a power-supply frequency diode; it is not designed for RF.

The Ge diode is designed for radio frequencies.

Should a trap be able to fully null the signal rather than half?

Yes, provided that it has infinite Q; this is impossible and you must make-do with "real-world"'components.

Why can the trap BOOST the current to the meter? (isn't that like a perpetual motion machine?)

I don't quite understand this question.

The quantity that is important is power; that is the combination of voltage & current.

Any combination of those two will produce the same power; high voltage /low current or low voltage/high current give the same power.

The resistance into which the power is being dissipated is another factor.

If only one quantity, voltage or current, is being measured, it tells nothing about power unless the other quantities are known.

Logged
VK6HP
Member

Posts: 277




Ignore
« Reply #2 on: November 22, 2017, 03:07:12 AM »

You'd gain a lot by reading the introductory chapters of the ARRL or similar handbook.  There's a wealth of highly relevant information there, presented in a fairly easy to read fashion.  It's not high-level stuff but it gives you some circuit and device understanding on which to build. It's a great basis to inform your experiments.

The germanium diode current-voltage characteristics are such that it "turns on" (conducts) at a lower voltage than a silicon diode, making a more effective rectifier for the very small RF voltages in your meter circuit.  And, in your case, the Si diode is not the best RF type, as the previous poster noted.  You would do better with a Schottky diode, which are cheap these days.

You don't have an infinitely deep notch, partly because of the finite Q mentioned.  But I see you still don't have your trap in a shielded box, so all your components and connection leads are picking up the BC signals, undoubtedly giving you some "blow by" of your little filter.  Before you can really assess your filter, you need to shield it, as mentioned in your previous thread.

I think your last question asks why a passive tuned circuit can't infinitely boost your incoming signal.  Again, the basic reason is finite Q, or non-zero loss in the system. It is indeed a bit like a mechanical resonator, such as a pendulum, which will eventually stop because of losses such as friction and air resistance.  To keep it going, you need to give little pushes at the right time, just as in the tuned circuit.
« Last Edit: November 22, 2017, 03:28:37 AM by VK6HP » Logged
WB6BYU
Member

Posts: 17483




Ignore
« Reply #3 on: November 22, 2017, 08:00:33 AM »

Quote from: KB1SNJ

Why does the germanium diode work so much more effectively than the 1n4007?



Forward voltage drop is the biggest reason.  A silicon diode typically starts conducting
at 0.6V or so, while germanium is nominally 0.3V, though it can be less at low currents.
That's why germanium diodes are often used in crystal radios.

But also the 1N4007 is a 1000V diode, and may include multiple diode elements
in series to handle the high voltage. 

It is a simple test to make if your DVM doesn't have a "diode test" setting:  put
a resistor in series with a voltage source to limit the current to perhaps 10uA,
connect the diode in series and measure the voltage drop across it.  You might
try it at different currents - the voltage is less consistent for germanium types.



Quote

Should a trap be able to fully null the signal rather than half?



A trap adds impedance to a circuit.  How much the signal drops will depend on the
input impedance of the receiver, as it makes a voltage divider.  If your trap provides
a 1000 kohm impedance at resonance, and the input impedance of your receiver
is also 1000k, then the voltage would drop by half.  But in a 50 ohm circuit it will
make a much greater reduction.  (Imagine the voltage divider effect with resistors.)

That's why series-tuned traps are sometimes used as a shunt element, to lower the
input impedance at that frequency and make the voltage divider more effective.



Quote

Why can the trap BOOST the current to the meter? (isnt that like a perpetual motion machine?)



In theory it shouldn't, if the only way for signals to reach the receiver is through
the trap.  But that is rarely the case in practice when the trap is connected via
clip leads on an unshielded bench.  I've often listened to the local radio station
while testing a circuit because the wiring picks up enough signal directly that
it gets detected in the first audio stage, even with no intended antenna connected.
Any capacitive or inductive coupling around the trap will let signals through,
and if you are adding clip leads to the circuit when you add the trap, those
can add to the signal as well.

The trap can only attenuate signals that pass through it:  the more other
wiring you provide that can pick up signals and bypass the trap, the less
effective the trap will be.

It really takes a good layout to reduce such pickup.  Ground leads have to be
very short, otherwise they add reactance and pick up signals.  For the level
of reduction you are trying to achieve, you may need to build the circuit to
VHF standards, with short connections to a common ground plane ("ugly
construction" over a bare copper board is a good example).  Certainly putting
it in a shielded box will help a lot.
Logged
WA3SKN
Member

Posts: 6642




Ignore
« Reply #4 on: November 22, 2017, 09:08:37 AM »

If you are testing to block signals, you should start with using a "shielded" enclosure.
73s.

-Mike.
Logged
KB1SNJ
Member

Posts: 121




Ignore
« Reply #5 on: November 22, 2017, 02:51:00 PM »

Regarding the enclosure, I don't have one for this yet, because I dont know where to get one the right size. I was thinking to cut a piece of heavy angle, drill that for the 3/8" shaft nuts, and make the other four sides in sheet tin.  How tight does the enclosure have to be, like with the cracks at the seams?
Logged
WB6BYU
Member

Posts: 17483




Ignore
« Reply #6 on: November 22, 2017, 03:45:07 PM »

Soldering some sections of copper-clad circuit board together is a quick way to make one.

Or a couple tin cans.  I used some real metal tea tins for several projects.  A loaf pan or
cake pan on top of a cookie sheet would probably work - see what you can find at the
local thrift store.
Logged
KB1SNJ
Member

Posts: 121




Ignore
« Reply #7 on: November 22, 2017, 03:55:58 PM »

thanks dale.  how tight do the seams need to be?  I see project boxes on ebay that have overlapped edges. just wonder if that's important to stop RF

Logged
WB6BYU
Member

Posts: 17483




Ignore
« Reply #8 on: November 22, 2017, 04:50:40 PM »

Someone else may know better how to calculate the maximum slot length vs. attenuation.

But it seems to me that the old 50's / 60's vintage rigs that were buttoned up tight for TVI reduction
might have used one screw every 4" to 6" or so along the overlapping (often painted) panels.  Given
the difference in frequency, it would seem like one screw per lineal foot (or at each corner of the box) is
probably enough.

But that's only a guess.
Logged
VK2TIL
Member

Posts: 545




Ignore
« Reply #9 on: November 22, 2017, 05:35:42 PM »

The maths for the waveguide-beyond-cutoff effect are simple; the lower cutoff frequency in GHz  is 30/(2*Slot Width in cm).  A 6" slot will not pass frequencies below 1 GHz although the cutoff is not instantaneous and a good design would have gaps rather less than 6".

So gaps are relatively unimportant at around 1.5 MHz.

I have two or three AM stations a km or two from me; they have never disturbed my test bench but the TV & FM stations about 6km away have occasionally given me cause to use some ten-dollar words.

Steel would be preferable as a shield in this case for reasons of resistivity and skin effect; about 1mm/0.040" thick would do.  Ground all shields of cables entering or leaving the enclosure to the outside

The suggestion of cooking utensils is a good one.



Logged
VK6HP
Member

Posts: 277




Ignore
« Reply #10 on: November 22, 2017, 09:23:31 PM »

As with many things, the simplest implementation takes you a good way towards the goal. Just putting the filter in a small, screwed together aluminium jiffy box or similar (with good contact along the seams) will get to at least 30 dB pickup reduction, while routine design techniques involving choice of materials, better treatment of seams, etc will get to 50 dB or more. It gets more demanding beyond that, especially for large enclosures or rooms. But for your purposes at 1 MHz any small metal enclosure, including those made of double-sided circuit board, will be fine. The PCB construction approach is simple and effective but if you do you use a steel box, make the box a little bit bigger to separate the tuned circuit more from the walls to minimize loss in the steel.  (You can read about the various mechanisms involved in shielding, including reflection and transmission loss).

With the various comments about Q and voltage divider circuit theory, it might be worth noting that the impedance of your parallel circuit at resonance is not infinite, for the reasons I previously mentioned. The actual value is proportional to the product of frequency, L and Q, in one representation. When you get things to the point of being able to reliably measure a -3 dB rejection bandwidth in the signal path, you can calculate Q and, knowing the frequency and inductance, estimate the impedance if you're interested.

 

« Last Edit: November 22, 2017, 09:39:52 PM by VK6HP » Logged
WA3SKN
Member

Posts: 6642




Ignore
« Reply #11 on: November 25, 2017, 11:36:04 AM »

How "tight" does the enclosure need to be?
Well, this depends of frequency(wavelength), but for 1520 kHz, not too tight!  Now for microwaves and above you gotta care.
I mention this because in the picture every clip lead is a potential antenna!  You don't really know where the RF is getting in!
Shielding, filtering and distance (transmit antenna to receive antenna) are the solutions for every RFI problem.
73s.

-Mike.

ps: Star washers were VERY popular in the 50s rigs for a reason!
Logged
KB1SNJ
Member

Posts: 121




Ignore
« Reply #12 on: November 25, 2017, 04:23:16 PM »

Is there anything to using wire mesh / hardware cloth to enclose it, as the BCB freq wouldn't make it through the 1/4" holes?

Also, I tested the Trap by attaching the Trap to the receiver antenna lug, disconnecting antenna then turning the receiver RF gain down until no audible reception. That established (I think ) a good baseline. Connected the antenna to the input on the Trap and thats how i tested its effectiveness. I got about (by ear) a half reduction of signal strength from min to max tuning of the variable cap on the Trap. I plan to try that with my digital transceiver as well and use the S meter
Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!