RF Ham Radio‎ > ‎

Regen: Aamazing Regen Receiver

2011/9/21

Regen is ever popular at around 1950s/1960s in ham circle.



(this boxing is built @2012)


I ever built a super regenerative receiver for FM and Air Band, it's amazing hear something never did before.  A Regen , different type receiver capable shortwave/CW , interesting thing to play with.  ARRL Regen for beginner, www.arrl.org/files/file/Technology/tis/info/pdf/0009061.pdf ,  N1TEV, suggest this receiver: you need a Regen built experiment before you move on. it's worthy to do.



below pictures show the final Receiver built via Manhattan style PCB with a panel. Panel make it easy to tune for station,plus cool view , highly recommend to  built one.

Fig 2 (click for enlarge)



(this is the original 'box' version @2011)


big wheel is the main tune from 3Mhz to 10Mhz (10-356pF C1 with 8uH L1), top mid knob is the Main Regen control, right bottom conner is fine Regen control use a multi-turn POT, seems over engineering. Top right is a fine tuner. it's cool to have additional knob to play with and
easily lock a station .
I use a very common transistor in China , 2sc9018, might with poor Gain,  sensitiveness is not so good for CW receiver, at least without a  Antenna is impossible (40M/80M). I try receive the 40M/80M ammeter band signal by a whip antenna(1.5Meter, with load at middle) but failed.  then i built a very very simple DP antenna to play with : a wipe spread out from window and a 2-3Meter wire roll down to window as another arm, length is not critical for receiver. Refer the Fig.2  left image .

after equipment a DP Antenna, CW receive is OK, poor performance(low voice, could improve by add a audio amp stage) at 80M/40M, but amazing at 30M band.

to receive AM broadcast, set Regen to the detector just not oscillate, to play with CW, make it just oscillate.  Additional NOTES: so wide tune range , 3Mhz to 10Mhz,  the right Regen knob position is variable: more positive feed back needed for lower frequency. 


updated this section @2012/7/24

a year later, i want a more permanent built, boxing it so it's will be easy to carry with.  and i also want to make some change.

This time, built the NA5N pipsqueak Regen receiver.  but with some modify.
1) replace lm386 with a 100x feedback pre amplifier plus a 8x power amplifier
2) replace one 1N941 with a RED led, make the work point higher, shift the correct Regen point higher.[the point just substantiating the oscillate]
3) the 2N2222 replace with a 9018
4) original coil is big, can not built in a small box (120mmx75mmx55mm). i'm going to built a FCZ coil to replace it. FCZ coil: 25T tap @5T, 3uH




My version Pipsqueak  @2012/7/26

i built the box first, this is the first mistake.  a pre-drill and mounted capacitor and pot limited the circuit board design, and must use a long wire to connect together. this lead so much Parasitic capacitance even make it totally does not work. plus signal path chaos.

first coil i made use 10S FCZ coil 10 turn tap at 5 turn. but there is no the point which it just begin oscillate.  carefully check the connection, one of  the tune capacitor  lead contact the the aluminum box, unfortunately i connect this one to L1.

my built had seems chaos, i want sensitive so tune more Regen but at the highest sensitives point, it start squeal, and CW receive is definitely impossible, reasons might be :
1) too long wire to var-cap C1
2) wrong wire type: i use the shield cable, witch had much more capacitance: at least 60pF.
3) the most important reason: bad arrange the circuit , signal path not optimized. even with more decouple does not make it better: i.e insert RC decouple every stage.


several days play around, here is some experience:
* the point that just substantiating the oscillate,  spread too wide in whole band it covered, one of reason is a wide band cover from 5Mhz to 15Mhz.
* Regen control make detune , even with ss9018(1pf internal capacitance, better than 2N2222)
* use 3 diode vs 1 diode + one red led as voltage regulator:  the work voltage affect the Regen control, work voltage shift the oscillate point up/down. for given Regen POT,  should find a work voltage make the oscillate point just the mid[@mid point of the whole band].
* don't know why the audio seems loss so much high frequency spectrum, very dull.[might root cause: it just start squeal ] 
 
built suggestion:
* if you are the new guy for RF, please make  exactly copy as you can,  according the N1TEV's version.
* if you try to replace the coil, tap it at 1/5 even 1/10

* if it does not oscillate at all,check your built first, then check your tunning cap and the coil, don't use long wire, ensure high Q.
* don't make it cover too wide band, that will make hardly tune to station and un-smooth Regen control.
*built the circuit first, then considerate a suitable box. here is the fisrt version box, you can see, there is no room for the real receiver!!!


several  typical problem scenario  when debug the Regen
updated @2012/7/29


* can find the oscillation point: check the wire , i connect capacitor's COMMON END to hot end of the L1, the capacitor's common end get in touch with GND, the aluminum box. still failed? check the voltage of each point to find where is wrong.
*can not find any station:  one cause is my receiver cover 12Mhz to 22Mhz,  impossible tune it to a station without a fine tune. I limited the range via replacing C3 to 100pF fix capacitor, then many station pop up.
*motorboat sound when put all circuit to box:  my box is small, signal patch chaos when push all of them to a small box. a small receiver take you much more time to arrange the position of all kind of thing, it's not worthy to over minimize it for  almost all home-brew device,especially your circuit is built to  separated PCB blocksRegen  is so high gain for one stage, and my 2 audio stage given left 1000 times gain, make it easy to be a motor boat
Final successfully box it @7/30 2012

i finally found my version Pipsqueak's detector had severely  pitfall to prevent the CW receiver and lead to severely motorboat noise:
Oscilloscope show me the detector almost don't have a carrier bypass capacitor, then carrier feed to audio stage make it oscillation so if you try receive the CW signal, the audio stage is overloaded by carrier signal.

* the C16,C9, C15 provide sufficient RF bypass for detector
* C8 adding to T4 provide high frequency negative feedback, prevent high frequncy oscillating.
* C3 change to 47p so it cover 8.7Mhz to 11Mhz make it easy to tune.
* use BB910 as the band set capacitor, so make it easy to arrange box's tune Knob.
* C17 decouple the POT noise, make it quiet while tuning.
* change R7 to 4.7k is also get enough gain if you connect a 20db Gain audio power amplifier.


The final box for this receiver:



Final though

I remenber the first version, use LM386 as amplifier for ear phone hearing. ear phone is very sensitive than the speaker, make it easy to catch the signal.  and the fine tune to re-gen and band-set is always worthy to have.  finnally, the high quality COIL is prefer, i wonder that's the reason why the 2011 version is better than the boxing version. 


Comments