TRF II: Infinity 0V2

Mar/3 2016

It's been 4 years since the first time I trying TRF.  I want to play with the JFET infinity detector, seems attractive.  At the same time,  want to know if the TRF must equip with big antenna,   how bad the ferrite antenna could be?  

I'm not that far from the broadcast station(suburbs areas of the big city). I ever built a reflex receiver works fine with ferrite antenna(Rflex: with voltage doubler detector).  And there are several other TRFs, perform baddly. 

after pickup a suitable variable capacitor and ferrite rod, i trying to adjust that LC coverage from 550 khz to 1.6 Mhz, finally i get 650 khz to 1.5 Mhz, i satisfy of this result as a experimental radio. 

just for a quick check i connect it to LM386, directly. wow, the sound is so loud, how it possible with a 40dB gain LM386? 

several guys remind me, does it oscillating? then i check the output of LM386, god, the waveform totally mess. 

tuned LC connect to the LM386 via a several tap,  if i connect 2 copper clap board showing in the photo, oscillating stop, stations going away also. while it working received something the body effect is huge, the feedback path of this casual regen receiver not well constructed that's why i get what i have.

A simple interesting Audio PA with ancient transistor

Mar/8 2016

AA1TJ in it's 'Talking Doll 40M regen receiver' introduce a interesting audio PA, a tuned Audio class A amplifier.  i build my version use 2 small common mode inductor, total inductance 245 mH, with 0.33 uF serial with 0.68 uF capacitor, it tuned to 700 HZ. 

this little beast bias with 5mA  current, given a max max output 4 Vpp to my 450 ohm earphone, VCC 7.4V to 9V.

It's voltage gain @690Hz about 26,  output 4 Vpp with input 15mV.  while frequency jump to 1.8 Khz, output only 0.5 Vpp,  voltage gain only 4.  now i notice this is a awesome PA for CW reception, not for me!!

Here is some other character of this little PA:

Input impedance 1.2 K, i measure the input voltage is 30 mV without connect signal source to the amplifier, my signal source output impedance is 600 R. while connect it to the amplifier, signal dropped to 20 mV, so i calculate the input resistance is 1.2k.  output impedance is 600 ohm(same calculation process: output  3.5V without load,  connected 450 ohm earphone, output 1.5V).

some feature about my 450 ohm earphone:

with 0.2 V driven( 90 uW), i get the clear audio voice, guess it would be 20 dB SBL. driven with 1.6 V signal get very loud audio.  i don't know it's sensitivity but feel it very bad from this rough estimation. compare to normal modern 32 ohm earphone for mobile, this is too bad.

Audio PA  Testing 

 4:1 transformer, 450-ohm earphone

• 0.6v/20mV = 600/20=30  @690HZ

• 0.8v/20mV =  40@1.8khz

4:1 transformer,  64 ohm earphone:

• 20mV is enough for hearing

• 120mV output with 20 mV input, voltage gain 6 @690HZ

Infinity detector

@Mar/10-13  2016

there is a good article to reading about the Infinity impedance detector:  http://sound.westhost.com/articles/am-radio.htm

the following tis the basic form of infinite impedance detector.  the idea of the infinite impedance detector is bias the JFET to cut-off range, then only in the positive half of the signal conduct the jfet, charging the small 1000p capacitor, and this capacitor can only discharge to the 100k resistor, so this is a peak detector.

casualty shortwave receiver

With a 15-meter long wire antenna, I'm pretty confident it will help to make a receiver. Bad news is,  It has no stations at all.  TRF frustrating me as before, even with the big antenna (works very well on 40 Meter SSB, CW).

My ferrite inductor has 2 taps,  I found only connect the tap to JFET,  stations pop out.  Nothing there if connect to the top of the tuner. This confuses me one week.   I finally notice,  the LC tank is not working on MW band,  it's Shortwave band Receiver! 

Tap1 is connected to JFET, tap2 is grounded, top of the inductor does not use.  Because mess builds,  I wrongly think tap 2 is another coil on the ferrite.

Some other thing still confuses me. why there is no medium station? even with my long wire antenna?    shortwave with long wire antenna works well, the detector showing good selectivity and good sensitivity stations( i think the long wire antenna contributes the sensitivity mostly).

In another way,  I test several detector diodes too.  My 1N60 seems bad, I bought it at a very low price, so makes sense it's bad.  An old 2AP9 glass encapsulated diode seems better.  

How about infinity detector:   much much better than 2AP9, worthing to trying, definitely. 

For medium wave, no station with a long antenna, outdoor or near the window, still does not work.

RF stage is the key for sensitivity. It's time to add one stage RF amplifier.  With a 5.6uH RFC,  the RF stage start oscillation(2N5486, UHF JFET).   4.7mH  RFC could cease the oscillate, but still no station outside or with the long antenna.

Finally, I replace the RFC with a resistor, this works, even though wired, I hear one station whispering near the windows. With long wire, still all noise.

My 1V1 seems a failure. I suspect the audio stage, its gain is so low, actually,  for my sensitive earphone, with impedance is 64 ohm,  the voltage gain only 6, that's is 15dB.

Cause I  did not know exactly my earphone sensitivity, no way to know the power needed to get enough SPL. The experiment about the earphone sensitivity scheduled to experiments. 

My audio generator spec:

Audio signal generator to oscilloscope high impedance 1X probe :  0dB,  1Vpp

Audio signal generator output impedance: 600Ω

Step Attenuator ( output to high enough load, 1 MΩ ):

earphone impedance testing (64Ω) :

-10 dB output to the earphone, 0.3 Vpp,  got 28 mV to deliver to the earphone,  it's could be calculated by output impedance vs earphone impedance, i.e, 0.3* 64/(600+64) = 28.9 mV

the sensitivity: [ ear phone about 100 ohm @ 1khz]

hear-able audio:          1.8 khz,  Signal Attenuation -50 dB ,  Signal amplitude 3 mVpp, earphone get around 420 uVpp

whisper audio:             1.8 khz   Signal Attenuation -30 dB,   Signal amplitude 30 mV, earphone get 4.28 mV

moderate loud audio   1.8khz,   Signal Attenuation -20dB,     Signal amplitude 0.1Vpp,   earphone get around 14mVpp (less than normal conversation volume)

A bad sensitivity earphone:

 hear-able audio:    1.8 kHz,  Signal Attenuation -30 dB ,  Signal amplitude 33 mVpp, earphone get around 20 mVpp

moderate loud audio   1.8 khz,   Signal Attenuation  0 dB,   Signal amplitude 1 Vpp,   earphone get around  1 Vpp (normal conversation volume)

Now, it's clear, 3 mV at least, delivery to the earphone,  we can hear something.  Another question is how to know the signal level on the aerial....

TRF/Rflex... sensitivity?

let's get some actual data about sensitivity, the TA7642, according to its datasheet, max sensitivity is 600uV, and it's delivered to the TA7642's input.

Back to value age, Reflex receiver could have a sensitivity of about 15uV on MW band and 30 uV on the shortwave band. On transistor age, the reflex and TRF is rarely been tested for sensitivity, and hard to find any reference.

Signal pick up on ferrite rod LC tank

refer to Weak_signals-mag_loop,  a ferrite rod antenna is actually a small magnetic loop antenna.  signal voltage appearing at the resonant tank related to several factor, hate math but this relatively easy to understand.

sensitivity of a receiver normally expressed in the uV/m, this is the intensity of electromagnetic wave. a antenna, pick up the signal, to voltage, relationship is fairly simple:

Е = h e                                              (1)

Е – is the voltage between antenna terminals in uV ( transforms the electromagnetic wave energy into voltage source)

е – is the intensity of electromagnetic wave in  uV/m.

h - effective height (a dimension in meters)

most of antenna is resonant, so, the resonant voltage we get is:

U= Q Е                       (8)             where    Q = 2пfL/ RL 

if we know the h, we can calculate the U.

h=2*pi*A/λ 

А = w μR S 

w  - the number of ML turns;

μR is the effective magnetic permeability of the ferrite rod SML(for my 5cm length, 1cm diameter,permeability 800, the effective permeability only 30)

S – is the area of the wingdings in m^2;

My rod:

A=120 x 30 x 0.000095 = 0.285

h=2*pi*0.285/300  @1Mhz  ~= 0.007

The Q factor of my rod, based on calculated, around 150.  

typical field strength on different area 

88 dBu: 25 mV/m (urban)

74 dBu: 5 mV/m (residential)

54 dBu: 0.5 mV/m (rural)

0.15 mV/m (43.5 dBu long rural)

Assume it's 0.5mV/m,  My location must be worse than this cause of high buildings everywhere around me . 

U=QE=Qhe=150*0.007*0.5= 0.5 mV, this indicate one stage audio might work. But in fact, It did not.  Assume I'm in 0.15 mV level area,  then I need about 100 times voltage gain.  This estimation based on the sensitivity of earphones.  It's quite easy the signal vary in one decade range.

So audio stage of an infinity detector 0V-1/2/3 receiver might need 40 dB to 46 dB voltage gain,  this indicate a  2 stages audio amplifier needed or a LM386.

SPL material:

https://diyaudioheaven.files.wordpress.com/2013/03/headphone-power-and-amplifiers.pdf

http://www.arrl.org/files/file/Technology/tis/info/pdf/culter.pdf

Relex receiver sensitivity:

http://educypedia.karadimov.info/library/chapte28.pdf

Voltage levels that can be obtained from SM:

http://www.lz1aq.signacor.com/docs/fa-eng/Weak_signals-mag_loop_engl.htm

typical broadcast field strength

http://radio-timetraveller.blogspot.com/2015/02/the-db-versus-dbu-mystery-signal.html

It need another audio stage. The plan is to use the same audio amplifier cascaded together.  Bad news is the 2 stage amplifier going oscillating while use transformer output. Finally resolved by replacing the 125 mH inductor with a 2.2 Kohm resistor. 

The result is really amazing! best straight receiver ever got, i can hear 2 strong station near the window,  outside receiver 8-13 stations.  i plan to stop here. won't add RF stage for it. this kind of receiver is a crystal radio with a powerful audio stage.

0V2(zero RF amplifier + infinity detector + 2 stage audio) schematic:

i'm tried to test the gain of detector + audio stages, by connect the RF generator to detector. but the signal is very weak and oscilloscope introduce huge interference, so given up.

2 stage audio stage amplifer character:

Gain : 220  (raw test data:  -50 dbm audio(3.3 mVpp), output 700mV)

Input impedance:  3.6k  (raw  test data: 600ohm 35 mVpp, after connect to amplifier 30mV)

The weak signal pick up from the ferrite rod antenna is interesting, but I have no way to measure it directly.  Base on the audio stage gain,  with the effective height of my rod antenna,  it might be deduced the field strength(or guess).

There is  one strong station,  the sound volume almost same as the earphone driven by 20 mVpp signal. 

• Audio input get  20mVpp/220 = ~ 0.09mV   (audio stage voltage amplifier gain)

• Considerate the 5.6k resistor and the input impedance of the audio stage, the detector output around 0.23 mVpp Audio signal

• Assume the signal had 60% modulation, the RF signal should be (0.23/0.6)x2= 0.77 mVpp

• Assume detector have 3dB loss, the detector should get 1 mVpp RF signal

• At 1Mhz,  Rod antenna :  Q <= 150,  h=0.007 (effective height compared to dipole antenna),  the field strength should be: (1/150)/0.007 ~=  1 mV/m

Above calculation  is kind of guess.  The typical field strength on different area() :

88 dBu: 25 mV/m (urban)

74 dBu: 5 mV/m (residential)

54 dBu: 0.5 mV/m (rural)   *** I'm here

0.15 mV/m (43.5 dBu long rural)

Listen Environment

I carry with this radio into downtown, result is, the audio volume is painful, there, the field strength is so strong, i suspect all radio you throw together will work well in down town.